opentrep-svn Mailing List for Open Travel Request Parser (Page 5)
Status: Beta
Brought to you by:
denis_arnaud
Menu
▾
▴
opentrep-svn — Subversion Activity Notification List
You can subscribe to this list here.
| 2009 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
(52) |
Aug
(19) |
Sep
(4) |
Oct
(10) |
Nov
(2) |
Dec
(4) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2010 |
Jan
(3) |
Feb
|
Mar
(3) |
Apr
|
May
|
Jun
(1) |
Jul
(2) |
Aug
(1) |
Sep
(9) |
Oct
|
Nov
(1) |
Dec
|
| 2011 |
Jan
|
Feb
(8) |
Mar
|
Apr
|
May
|
Jun
|
Jul
(2) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: <den...@us...> - 2009-07-18 20:30:12
|
Revision: 139
http://opentrep.svn.sourceforge.net/opentrep/?rev=139&view=rev
Author: denis_arnaud
Date: 2009-07-18 20:30:08 +0000 (Sat, 18 Jul 2009)
Log Message:
-----------
[Dev] Fixed a bug causing an infinite loop when no-matching words were part of the query string.
Modified Paths:
--------------
trunk/opentrep/opentrep/batches/searcher.cpp
trunk/opentrep/opentrep/bom/ResultHolder.cpp
trunk/opentrep/opentrep/bom/StringMatcher.cpp
trunk/opentrep/opentrep/bom/StringMatcher.hpp
trunk/opentrep/opentrep/command/RequestInterpreter.cpp
Modified: trunk/opentrep/opentrep/batches/searcher.cpp
===================================================================
--- trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-18 18:32:04 UTC (rev 138)
+++ trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-18 20:30:08 UTC (rev 139)
@@ -232,11 +232,11 @@
const OPENTREP::NbOfMatches_T nbOfMatches =
opentrepService.interpretTravelRequest (lTravelQuery, lLocationList);
- if (nbOfMatches != 0) {
std::cout << nbOfMatches << " (geographical) location(s) have been found "
<< "matching your query (`" << lTravelQuery << "´)."
<< std::endl;
+ if (nbOfMatches != 0) {
OPENTREP::NbOfMatches_T idx = 1;
for (OPENTREP::LocationList_T::const_iterator itLocation =
lLocationList.begin();
Modified: trunk/opentrep/opentrep/bom/ResultHolder.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/ResultHolder.cpp 2009-07-18 18:32:04 UTC (rev 138)
+++ trunk/opentrep/opentrep/bom/ResultHolder.cpp 2009-07-18 20:30:08 UTC (rev 139)
@@ -96,8 +96,8 @@
break;
}
- // Remove a word from the query string
- StringMatcher::removeOneWord (ioPartialQueryString);
+ // Remove the furthest right word from the query string
+ StringMatcher::removeFurthestRightWord (ioPartialQueryString);
// Stop when the resulting string gets empty.
if (ioPartialQueryString.empty() == true) {
@@ -150,22 +150,40 @@
// DEBUG
OPENTREP_LOG_DEBUG ("==> Matching of the query string: `"
<< lQueryString << "'");
+
+ /**
+ Remove, from the lRemainingQueryString string, the part
+ which has been already successfully parsed. <br>For
+ instance, when 'sna francisco rio de janeiro' is the
+ initial full clean query string, the searchString()
+ method first reduce the query string to 'sna francisco',
+ which successfully matches against SFO (San Francisco
+ airport). <br>Then, the remaining part of the query
+ string to be parsed is 'rio de janeiro'. So, the already
+ parsed part, namely 'sna francisco', must be subtracted
+ from the initial query string.
+ */
+ StringMatcher::subtractParsedToRemaining (lQueryString,
+ lRemainingQueryString);
+ } else {
+ // DEBUG
+ OPENTREP_LOG_DEBUG ("==> No matching of the query string: `"
+ << lRemainingQueryString
+ << "'. Skip the beginning word.");
+ assert (lQueryString.empty() == true);
+
+ /**
+ We must suppress (at least) the furthest left word, as it
+ hinders the remaining of the query string to be
+ matched. If that furthest left word is the only word of
+ the query string, the remaining query string will
+ therefore be empty, and the loop will therefore be exited
+ in the next step below.
+ */
+ // Remove the furthest right word from the query string
+ StringMatcher::removeFurthestLeftWord (lRemainingQueryString);
}
- /**
- Remove, from the lRemainingQueryString string, the part which
- has been already successfully parsed.
- <br>For instance, when 'sna francisco rio de janeiro' is the
- initial full clean query string, the searchString() method
- first reduce the query string to 'sna francisco', which
- successfully matches against SFO (San Francisco airport).
- <br>Then, the remaining part of the query string to be parsed is
- 'rio de janeiro'. So, the already parsed part, namely
- 'sna francisco', must be subtracted from the initial query string.
- */
- StringMatcher::subtractParsedToRemaining (lQueryString,
- lRemainingQueryString);
-
// If there is nothing left to be parsed, we have then finished
// to parse the initial string.
if (lRemainingQueryString.empty() == true) {
Modified: trunk/opentrep/opentrep/bom/StringMatcher.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/StringMatcher.cpp 2009-07-18 18:32:04 UTC (rev 138)
+++ trunk/opentrep/opentrep/bom/StringMatcher.cpp 2009-07-18 20:30:08 UTC (rev 139)
@@ -312,7 +312,7 @@
}
// //////////////////////////////////////////////////////////////////////
- void StringMatcher::removeOneWord (std::string& ioQueryString) {
+ void StringMatcher::removeFurthestRightWord (std::string& ioQueryString) {
assert (ioQueryString.empty() == false);
WordList_T lWordList;
@@ -328,6 +328,22 @@
}
// //////////////////////////////////////////////////////////////////////
+ void StringMatcher::removeFurthestLeftWord (std::string& ioQueryString) {
+ assert (ioQueryString.empty() == false);
+
+ WordList_T lWordList;
+ WordHolder::tokeniseStringIntoWordList (ioQueryString, lWordList);
+ assert (lWordList.empty() == false);
+
+ // Remove the furthest left word
+ lWordList.pop_front();
+
+ const std::string& lReducedString =
+ WordHolder::createStringFromWordList (lWordList);
+ ioQueryString = lReducedString;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
void StringMatcher::
subtractParsedToRemaining (const std::string& iAlreadyParsedQueryString,
std::string& ioRemainingQueryString) {
Modified: trunk/opentrep/opentrep/bom/StringMatcher.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/StringMatcher.hpp 2009-07-18 18:32:04 UTC (rev 138)
+++ trunk/opentrep/opentrep/bom/StringMatcher.hpp 2009-07-18 20:30:08 UTC (rev 139)
@@ -41,8 +41,11 @@
extractBestMatchingDocumentFromMSet (const Xapian::MSet&, Document&);
/** Remove the word furthest at right. */
- static void removeOneWord (std::string& ioQueryString);
+ static void removeFurthestRightWord (std::string& ioQueryString);
+ /** Remove the word furthest at left. */
+ static void removeFurthestLeftWord (std::string& ioQueryString);
+
/** Remove, from a string, the part corresponding to the one given
as parameter. */
static void
Modified: trunk/opentrep/opentrep/command/RequestInterpreter.cpp
===================================================================
--- trunk/opentrep/opentrep/command/RequestInterpreter.cpp 2009-07-18 18:32:04 UTC (rev 138)
+++ trunk/opentrep/opentrep/command/RequestInterpreter.cpp 2009-07-18 20:30:08 UTC (rev 139)
@@ -159,6 +159,7 @@
of the Place objects. Those (Location) structures are passed
back to the caller of the service. */
lPlaceHolder.createLocations (ioLocationList);
+ oNbOfMatches = ioLocationList.size();
return oNbOfMatches;
}
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-18 18:32:11
|
Revision: 138
http://opentrep.svn.sourceforge.net/opentrep/?rev=138&view=rev
Author: denis_arnaud
Date: 2009-07-18 18:32:04 +0000 (Sat, 18 Jul 2009)
Log Message:
-----------
[Dev] Improved the interface, so that the travel search now returns a list of Location structures, that the caller can then benefit from.
Modified Paths:
--------------
trunk/opentrep/opentrep/OPENTREP_Service.hpp
trunk/opentrep/opentrep/batches/indexer.cpp
trunk/opentrep/opentrep/batches/searcher.cpp
trunk/opentrep/opentrep/bom/BomAbstract.hpp
trunk/opentrep/opentrep/bom/Document.cpp
trunk/opentrep/opentrep/bom/Document.hpp
trunk/opentrep/opentrep/bom/Names.cpp
trunk/opentrep/opentrep/bom/Names.hpp
trunk/opentrep/opentrep/bom/Place.cpp
trunk/opentrep/opentrep/bom/Place.hpp
trunk/opentrep/opentrep/bom/PlaceHolder.cpp
trunk/opentrep/opentrep/bom/PlaceHolder.hpp
trunk/opentrep/opentrep/bom/Result.cpp
trunk/opentrep/opentrep/bom/Result.hpp
trunk/opentrep/opentrep/bom/ResultHolder.cpp
trunk/opentrep/opentrep/bom/ResultHolder.hpp
trunk/opentrep/opentrep/bom/World.cpp
trunk/opentrep/opentrep/bom/World.hpp
trunk/opentrep/opentrep/command/RequestInterpreter.cpp
trunk/opentrep/opentrep/command/RequestInterpreter.hpp
trunk/opentrep/opentrep/service/OPENTREP_Service.cpp
trunk/opentrep/opentrep/sources.mk
Added Paths:
-----------
trunk/opentrep/opentrep/Location.hpp
trunk/opentrep/opentrep/OPENTREP_Abstract.hpp
Added: trunk/opentrep/opentrep/Location.hpp
===================================================================
--- trunk/opentrep/opentrep/Location.hpp (rev 0)
+++ trunk/opentrep/opentrep/Location.hpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -0,0 +1,217 @@
+#ifndef __OPENTREP_LOCATION_HPP
+#define __OPENTREP_LOCATION_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// STL
+#include <istream>
+#include <ostream>
+#include <string>
+#include <list>
+// OpenTrep
+#include <opentrep/OPENTREP_Types.hpp>
+#include <opentrep/OPENTREP_Abstract.hpp>
+
+namespace OPENTREP {
+
+ /** List of names for a given (geographical) location. */
+ typedef std::list<std::string> LocationNameList_T;
+
+
+ /** Structure modelling a (geographical) location. */
+ struct Location : public OPENTREP_Abstract {
+ public:
+ // ///////// Getters ////////
+ /** Get the Location code. */
+ std::string getLocationCode() const {
+ return _locationCode;
+ }
+
+ /** Get the City code. */
+ std::string getCityCode() const {
+ return _cityCode;
+ }
+
+ /** Get the State code. */
+ std::string getStateCode() const {
+ return _stateCode;
+ }
+
+ /** Get the Country code. */
+ std::string getCountryCode() const {
+ return _countryCode;
+ }
+
+ /** Get the Region code. */
+ std::string getRegionCode() const {
+ return _regionCode;
+ }
+
+ /** Get the Continent code. */
+ std::string getContinentCode() const {
+ return _continentCode;
+ }
+
+ /** Get the Time-zone group. */
+ std::string getTimeZoneGroup() const {
+ return _timeZoneGroup;
+ }
+
+ /** Get the Longitude. */
+ double getLongitude() const {
+ return _longitude;
+ }
+
+ /** Get the Latitude. */
+ double getLatitude() const {
+ return _latitude;
+ }
+
+ /** Get the list of (American) English names for that location. */
+ const LocationNameList_T& getNameList () const {
+ return _nameList;
+ }
+
+
+ // ///////// Setters //////////
+ /** Set the Location code. */
+ void setLocationCode (const std::string& iLocationCode) {
+ _locationCode = iLocationCode;
+ }
+
+ /** Set the City code. */
+ void setCityCode (const std::string& iCityCode) {
+ _cityCode = iCityCode;
+ }
+
+ /** Set the State code. */
+ void setStateCode (const std::string& iStateCode) {
+ _stateCode = iStateCode;
+ }
+
+ /** Set the Country code. */
+ void setCountryCode (const std::string& iCountryCode) {
+ _countryCode = iCountryCode;
+ }
+
+ /** Set the Region code. */
+ void setRegionCode (const std::string& iRegionCode) {
+ _regionCode = iRegionCode;
+ }
+
+ /** Set the Continent code. */
+ void setContinentCode (const std::string& iContinentCode) {
+ _continentCode = iContinentCode;
+ }
+
+ /** Set the Time-zone group. */
+ void setTimeZoneGroup (const std::string& iTimeZoneGroup) {
+ _timeZoneGroup = iTimeZoneGroup;
+ }
+
+ /** Set the Longitude. */
+ void setLongitude (const double& iLongitude) {
+ _longitude = iLongitude;
+ }
+
+ /** Set the Latitude. */
+ void setLatitude (const double& iLatitude) {
+ _latitude = iLatitude;
+ }
+
+ /** Set the list of (American) English names for that location. */
+ void getNameList (const LocationNameList_T& iNameList) {
+ _nameList = iNameList;
+ }
+
+
+ public:
+ // ///////// Display methods ////////
+ /** Dump a structure into an output stream.
+ @param ostream& the output stream. */
+ void toStream (std::ostream& ioOut) const {
+ ioOut << toString();
+ }
+
+ /** Read a structure from an input stream.
+ @param istream& the input stream. */
+ void fromStream (std::istream&) {
+ }
+
+ /** Get a short display of the Location structure. */
+ std::string toShortString() const {
+ std::ostringstream oStr;
+ oStr << _locationCode << ", " << _cityCode << ", " << _stateCode
+ << ", " << _countryCode << ", " << _regionCode
+ << ", " << _continentCode << ", " << _timeZoneGroup
+ << ", " << _longitude << ", " << _latitude;
+ return oStr.str();
+ }
+
+ /** Get the serialised version of the Location structure. */
+ std::string toString() const {
+ std::ostringstream oStr;
+ oStr << toShortString();
+ for (LocationNameList_T::const_iterator itName = _nameList.begin();
+ itName != _nameList.end(); ++itName) {
+ oStr << ", " << *itName;
+ }
+ return oStr.str();
+ }
+
+
+ public:
+ /** Main Constructor. */
+ Location (const std::string& iPlaceCode, const std::string& iCityCode,
+ const std::string& iStateCode, const std::string& iCountryCode,
+ const std::string& iRegionCode, const std::string& iContinentCode,
+ const std::string& iTimeZoneGroup,
+ const double iLongitude, const double iLatitude,
+ const LocationNameList_T& iNameList)
+ : _locationCode (iPlaceCode), _cityCode (iCityCode),
+ _stateCode (iStateCode), _countryCode (iCountryCode),
+ _regionCode (iRegionCode), _continentCode (iContinentCode),
+ _timeZoneGroup (iTimeZoneGroup), _longitude (iLongitude),
+ _latitude (iLatitude), _nameList (iNameList) {
+ }
+
+ /** Default Constructor. */
+ // Location ();
+ /** Default copy constructor. */
+ // Location (const Location&);
+
+ /** Destructor. */
+ virtual ~Location() {}
+
+
+ private:
+ // /////// Attributes /////////
+ /** Location code. */
+ std::string _locationCode;
+ /** City code. */
+ std::string _cityCode;
+ /** State code. */
+ std::string _stateCode;
+ /** Country code. */
+ std::string _countryCode;
+ /** Region code. */
+ std::string _regionCode;
+ /** Continent code. */
+ std::string _continentCode;
+ /** Time-zone group. */
+ std::string _timeZoneGroup;
+ /** Longitude. */
+ double _longitude;
+ /** Latitude. */
+ double _latitude;
+ /** List of (American) English names. */
+ LocationNameList_T _nameList;
+ };
+
+
+ /** List of (geographical) location structures. */
+ typedef std::list<Location> LocationList_T;
+
+}
+#endif // __OPENTREP_LOCATION_HPP
Added: trunk/opentrep/opentrep/OPENTREP_Abstract.hpp
===================================================================
--- trunk/opentrep/opentrep/OPENTREP_Abstract.hpp (rev 0)
+++ trunk/opentrep/opentrep/OPENTREP_Abstract.hpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -0,0 +1,84 @@
+#ifndef __OPENTREP_OPENTREP_ABSTRACT_HPP
+#define __OPENTREP_OPENTREP_ABSTRACT_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// STL
+#include <istream>
+#include <ostream>
+#include <sstream>
+#include <string>
+
+namespace OPENTREP {
+
+ /** Base class for the OPENTREP interface structures. */
+ struct OPENTREP_Abstract {
+ public:
+ // /////////// Display support methods /////////
+ /** Dump a structure into an output stream.
+ @param ostream& the output stream. */
+ virtual void toStream (std::ostream& ioOut) const = 0;
+
+ /** Read a structure from an input stream.
+ @param istream& the input stream. */
+ virtual void fromStream (std::istream& ioIn) = 0;
+
+ /** Get the serialised version of the structure. */
+ virtual std::string toString() const = 0;
+
+
+ protected:
+ /** Protected Default Constructor to ensure this class is abtract. */
+ OPENTREP_Abstract () {}
+ OPENTREP_Abstract (const OPENTREP_Abstract&) {}
+
+ /** Destructor. */
+ virtual ~OPENTREP_Abstract() {}
+ };
+}
+
+/**
+ Piece of code given by Nicolai M. Josuttis, Section 13.12.1 "Implementing
+ Output Operators" (p653) of his book "The C++ Standard Library: A Tutorial
+ and Reference", published by Addison-Wesley.
+ */
+template <class charT, class traits>
+inline
+std::basic_ostream<charT, traits>&
+operator<< (std::basic_ostream<charT, traits>& ioOut,
+ const OPENTREP::OPENTREP_Abstract& iStructure) {
+ /**
+ string stream:
+ - with same format
+ - without special field width
+ */
+ std::basic_ostringstream<charT,traits> ostr;
+ ostr.copyfmt (ioOut);
+ ostr.width (0);
+
+ // Fill string stream
+ iStructure.toStream (ostr);
+
+ // Print string stream
+ ioOut << ostr.str();
+
+ return ioOut;
+}
+
+/**
+ Piece of code given by Nicolai M. Josuttis, Section 13.12.1 "Implementing
+ Output Operators" (pp655-657) of his book "The C++ Standard Library:
+ A Tutorial and Reference", published by Addison-Wesley.
+ */
+template <class charT, class traits>
+inline
+std::basic_istream<charT, traits>&
+operator>> (std::basic_istream<charT, traits>& ioIn,
+ OPENTREP::OPENTREP_Abstract& ioStucture) {
+ // Fill Bom object with input stream
+ ioStucture.fromStream (ioIn);
+ return ioIn;
+}
+
+#endif // __OPENTREP_OPENTREP_ABSTRACT_HPP
Modified: trunk/opentrep/opentrep/OPENTREP_Service.hpp
===================================================================
--- trunk/opentrep/opentrep/OPENTREP_Service.hpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/OPENTREP_Service.hpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -7,30 +7,44 @@
// STL
#include <ostream>
#include <string>
-// OPENTREP
+// OpenTREP
#include <opentrep/OPENTREP_Types.hpp>
+#include <opentrep/Location.hpp>
namespace OPENTREP {
- /** Forward declaration. */
+ // Forward declaration
class OPENTREP_ServiceContext;
+
/** Interface for the OPENTREP Services. */
class OPENTREP_Service {
public:
- /** Constructor. */
+ // /////////// Business Methods /////////////
+ /** Build the Xapian database (index) on the BOM held in memory. */
+ void buildSearchIndex ();
+
+ /** Match the given string, thanks to a full-text search on the
+ underlying Xapian index (named "database").
+ @param const std::string& (Travel-related) query string (e.g.,
+ "sna francicso rio de janero lso angles reykyavki nce iev mow").
+ @param LocationList_T& List of (geographical) locations, if any,
+ matching the given query string.
+ @return NbOfMatches_T Number of matches. */
+ NbOfMatches_T interpretTravelRequest (const std::string& iTravelQuery,
+ LocationList_T&);
+
+
+ // ////////// Constructors and destructors //////////
+ /** Constructor.
+ @param std::ostream& Output log stream (for instance, std::cout).
+ @param const std::string& Filepath of the Xapian index/database. */
OPENTREP_Service (std::ostream& ioLogStream,
const std::string& iXapianDatabaseFilepath);
/** Destructor. */
~OPENTREP_Service();
- /** Build the Xapian database (index) on the BOM held in memory. */
- void buildSearchIndex ();
-
- /** Perform the query, thanks to the underlying Xapian database
- (index) name. */
- void interpretTravelRequest (const std::string& iTravelQuery);
private:
// /////// Construction and Destruction helper methods ///////
@@ -43,12 +57,10 @@
void init (std::ostream& ioLogStream,
const std::string& iXapianDatabaseFilepath);
- /** Initilise the log. */
- void logInit (const LOG::EN_LogLevel iLogLevel, std::ostream& ioLogStream);
-
- /** Finaliser. */
+ /** Finalise. */
void finalise ();
+
private:
// ///////// Service Context /////////
/** Opentrep context. */
Modified: trunk/opentrep/opentrep/batches/indexer.cpp
===================================================================
--- trunk/opentrep/opentrep/batches/indexer.cpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/batches/indexer.cpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -139,6 +139,11 @@
logOutputFile.open (lLogFilename.c_str());
logOutputFile.clear();
+ //
+ std::cout << "Creating the Xapian index/database may take a few minutes "
+ << "on some architectures (and a few seconds on fastest ones)..."
+ << std::endl;
+
// Initialise the context
OPENTREP::OPENTREP_Service opentrepService (logOutputFile,
lXapianDatabaseName);
Modified: trunk/opentrep/opentrep/batches/searcher.cpp
===================================================================
--- trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -228,7 +228,23 @@
lXapianDatabaseName);
// Query the Xapian database (index)
- opentrepService.interpretTravelRequest (lTravelQuery);
+ OPENTREP::LocationList_T lLocationList;
+ const OPENTREP::NbOfMatches_T nbOfMatches =
+ opentrepService.interpretTravelRequest (lTravelQuery, lLocationList);
+
+ if (nbOfMatches != 0) {
+ std::cout << nbOfMatches << " (geographical) location(s) have been found "
+ << "matching your query (`" << lTravelQuery << "´)."
+ << std::endl;
+
+ OPENTREP::NbOfMatches_T idx = 1;
+ for (OPENTREP::LocationList_T::const_iterator itLocation =
+ lLocationList.begin();
+ itLocation != lLocationList.end(); ++itLocation, ++idx) {
+ const OPENTREP::Location& lLocation = *itLocation;
+ std::cout << " [" << idx << "]: " << lLocation << std::endl;
+ }
+ }
// Close the Log outputFile
logOutputFile.close();
Modified: trunk/opentrep/opentrep/bom/BomAbstract.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/BomAbstract.hpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/BomAbstract.hpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -9,8 +9,6 @@
#include <ostream>
#include <sstream>
#include <string>
-// OpenTrep
-#include <opentrep/bom/Language.hpp>
namespace OPENTREP {
@@ -32,11 +30,11 @@
/** Get a string describing the whole key (differentiating two objects
at any level). */
- virtual const std::string describeKey() const = 0;
+ virtual std::string describeKey() const = 0;
/** Get a string describing the short key (differentiating two objects
at the same level). */
- virtual const std::string describeShortKey() const = 0;
+ virtual std::string describeShortKey() const = 0;
protected:
Modified: trunk/opentrep/opentrep/bom/Document.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/Document.cpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/Document.cpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -24,14 +24,14 @@
}
// //////////////////////////////////////////////////////////////////////
- const std::string Document::describeShortKey() const {
+ std::string Document::describeShortKey() const {
std::ostringstream oStr;
oStr << _queryString;
return oStr.str();
}
// //////////////////////////////////////////////////////////////////////
- const std::string Document::describeKey() const {
+ std::string Document::describeKey() const {
return describeShortKey();
}
Modified: trunk/opentrep/opentrep/bom/Document.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/Document.hpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/Document.hpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -84,11 +84,11 @@
/** Get a string describing the whole key (differentiating two objects
at any level). */
- const std::string describeKey() const;
+ std::string describeKey() const;
/** Get a string describing the short key (differentiating two objects
at the same level). */
- const std::string describeShortKey() const;
+ std::string describeShortKey() const;
public:
Modified: trunk/opentrep/opentrep/bom/Names.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/Names.cpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/Names.cpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -2,7 +2,7 @@
// Import section
// //////////////////////////////////////////////////////////////////////
// C
-#include <assert.h>
+#include <cassert>
// STL
#include <sstream>
// OpenTrep
@@ -12,6 +12,7 @@
// //////////////////////////////////////////////////////////////////////
Names::Names() : _languageCode (Language::en_US) {
+ assert (false);
}
// //////////////////////////////////////////////////////////////////////
@@ -41,14 +42,14 @@
}
// //////////////////////////////////////////////////////////////////////
- const std::string Names::describeShortKey() const {
+ std::string Names::describeShortKey() const {
std::ostringstream oStr;
oStr << "[" << Language::getLongLabel (_languageCode) << "]: ";
return oStr.str();
}
// //////////////////////////////////////////////////////////////////////
- const std::string Names::describeKey() const {
+ std::string Names::describeKey() const {
return describeShortKey();
}
Modified: trunk/opentrep/opentrep/bom/Names.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/Names.hpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/Names.hpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -33,6 +33,7 @@
private:
/** Default constructor: should not be used. */
Names();
+
public:
// /////////// Getters ///////////////
@@ -78,14 +79,15 @@
/** Get a string describing the whole key (differentiating two objects
at any level). */
- const std::string describeKey() const;
+ std::string describeKey() const;
/** Get a string describing the short key (differentiating two objects
at the same level). */
- const std::string describeShortKey() const;
+ std::string describeShortKey() const;
- // Attributes
+ private:
+ // //////////// Attributes //////////////
/** Language code (e.g., en_US, fr_FR, etc.). */
Language::EN_Language _languageCode;
@@ -93,7 +95,11 @@
NameList_T _nameList;
};
- // Type definitions
+ // ////////////// Type definitions ////////////
+ /** Matrix of place names: for each of the language, the place gets a
+ corresponding list of names.
+ <br>For instance, MUC corresponds to Munich in English, München
+ in German, Munique in French, Мюнхен in Russian, etc. */
typedef std::map<Language::EN_Language, Names> NameMatrix_T;
}
Modified: trunk/opentrep/opentrep/bom/Place.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/Place.cpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/Place.cpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -27,16 +27,40 @@
// //////////////////////////////////////////////////////////////////////
Place::~Place () {
}
+
+ // //////////////////////////////////////////////////////////////////////
+ std::string Place::getCityCode() const {
+ std::string oCityCode (_cityCode);
+ if (oCityCode.empty() == true) {
+ oCityCode = _placeCode;
+ }
+ return oCityCode;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ bool Place::getNameList (const Language::EN_Language& iLanguageCode,
+ NameList_T& ioNameList) const {
+ bool oFoundNameList = false;
+
+ NameMatrix_T::const_iterator itNameList = _nameMatrix.find (iLanguageCode);
+ if (itNameList != _nameMatrix.end()) {
+ const Names& lNameList = itNameList->second;
+ ioNameList = lNameList.getNameList();
+ oFoundNameList = true;
+ }
+
+ return oFoundNameList;
+ }
// //////////////////////////////////////////////////////////////////////
- const std::string Place::describeShortKey() const {
+ std::string Place::describeShortKey() const {
std::ostringstream oStr;
oStr << _placeCode;
return oStr.str();
}
// //////////////////////////////////////////////////////////////////////
- const std::string Place::describeKey() const {
+ std::string Place::describeKey() const {
return describeShortKey();
}
@@ -46,13 +70,10 @@
not an airport. The city code is thus the same as the place code
itself. */
std::ostringstream oStr;
- oStr << describeShortKey() << ", ";
- if (_cityCode.empty()) {
- oStr << _placeCode << ", ";
- } else {
- oStr << _cityCode << ", ";
- }
- oStr << _stateCode
+ oStr << describeShortKey();
+
+ const std::string& lCityCode = getCityCode();
+ oStr << ", " << lCityCode << ", " << _stateCode
<< ", " << _countryCode << ", " << _regionCode
<< ", " << _continentCode << ", " << _timeZoneGroup
<< ", " << _longitude << ", " << _latitude << ", " << _docID << ". ";
@@ -72,22 +93,23 @@
not an airport. The city code is thus the same as the place code
itself. */
std::ostringstream oStr;
- oStr << describeShortKey() << ", ";
- if (_cityCode.empty()) {
- oStr << _placeCode << ", ";
- } else {
- oStr << _cityCode << ", ";
- }
- oStr << _stateCode
+ oStr << describeShortKey();
+
+ const std::string& lCityCode = getCityCode();
+ oStr << ", " << lCityCode << ", " << _stateCode
<< ", " << _countryCode << ", " << _regionCode
<< ", " << _continentCode << ", " << _timeZoneGroup
<< ", " << _longitude << ", " << _latitude << ", " << _docID;
NameMatrix_T::const_iterator itNameHolder = _nameMatrix.begin();
- const Names& lNameHolder = itNameHolder->second;
- const std::string& lFirstName = lNameHolder.getFirstName();
- if (lFirstName.empty() == false) {
- oStr << ", " << lFirstName << ".";
+ if (itNameHolder != _nameMatrix.end()) {
+
+ const Names& lNameHolder = itNameHolder->second;
+ const std::string& lFirstName = lNameHolder.getFirstName();
+
+ if (lFirstName.empty() == false) {
+ oStr << ", " << lFirstName << ".";
+ }
}
return oStr.str();
@@ -103,18 +125,16 @@
}
// //////////////////////////////////////////////////////////////////////
- const std::string Place::shortDisplay() const {
+ std::string Place::shortDisplay() const {
/* When the city code is empty, it means that the place is a city and
not an airport. The city code is thus the same as the place code
itself. */
std::ostringstream oStr;
oStr << describeKey();
- if (_cityCode.empty()) {
- oStr << ", city code = " << _placeCode;
- } else {
- oStr << ", city code = " << _cityCode;
- }
- oStr << ", state code = " << _stateCode
+
+ const std::string& lCityCode = getCityCode();
+ oStr << ", city code = " << lCityCode
+ << ", state code = " << _stateCode
<< ", country code = " << _countryCode
<< ", region code = " << _regionCode
<< ", continent code = " << _continentCode
@@ -127,7 +147,7 @@
}
// //////////////////////////////////////////////////////////////////////
- const std::string Place::display() const {
+ std::string Place::display() const {
std::ostringstream oStr;
oStr << shortDisplay();
for (NameMatrix_T::const_iterator itNameList = _nameMatrix.begin();
@@ -174,4 +194,27 @@
_nameMatrix.clear();
}
+ // //////////////////////////////////////////////////////////////////////
+ Location Place::createLocation() const {
+
+ const std::string& lCityCode = getCityCode();
+
+ NameList_T lNameList;
+ const bool hasFoundNameList = getNameList (Language::en_US, lNameList);
+
+ if (hasFoundNameList == false) {
+ //
+ OPENTREP_LOG_ERROR ("No list of names in (American) English (en_US "
+ << "locale) can be found for the following place: "
+ << toShortString());
+ throw LanguageCodeNotDefinedInNameTableException();
+ }
+ assert (hasFoundNameList == true);
+
+ // Copy the parameters from the Place object to the Location structure
+ Location oLocation (_placeCode, lCityCode, _stateCode, _countryCode,
+ _regionCode, _continentCode, _timeZoneGroup,
+ _longitude, _latitude, lNameList);
+ return oLocation;
+ }
}
Modified: trunk/opentrep/opentrep/bom/Place.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/Place.hpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/Place.hpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -11,6 +11,7 @@
#include <map>
// OpenTrep Bom
#include <opentrep/OPENTREP_Types.hpp>
+#include <opentrep/Location.hpp>
#include <opentrep/bom/BomAbstract.hpp>
#include <opentrep/bom/Names.hpp>
@@ -20,7 +21,7 @@
class World;
class PlaceHolder;
- /** Structure modelling an place. */
+ /** Class modelling a place. */
class Place : public BomAbstract {
friend class FacWorld;
friend class FacPlace;
@@ -29,37 +30,38 @@
public:
// ///////// Getters ////////
/** Get the Place code. */
- std::string getPlaceCode() const {
+ const std::string& getPlaceCode() const {
return _placeCode;
}
- /** Get the City code. */
- std::string getCityCode() const {
- return _cityCode;
- }
-
+ /** Get the City code.
+ <br>When the city code is empty, it means that the place is a
+ city and not an airport. The city code is thus the same as the
+ place code itself. */
+ std::string getCityCode() const;
+
/** Get the State code. */
- std::string getStateCode() const {
+ const std::string& getStateCode() const {
return _stateCode;
}
/** Get the Country code. */
- std::string getCountryCode() const {
+ const std::string& getCountryCode() const {
return _countryCode;
}
/** Get the Region code. */
- std::string getRegionCode() const {
+ const std::string& getRegionCode() const {
return _regionCode;
}
/** Get the Continent code. */
- std::string getContinentCode() const {
+ const std::string& getContinentCode() const {
return _continentCode;
}
/** Get the Time-zone group. */
- std::string getTimeZoneGroup() const {
+ const std::string& getTimeZoneGroup() const {
return _timeZoneGroup;
}
@@ -83,6 +85,14 @@
return _nameMatrix;
}
+ /** Get, for a given language (code), the corresponding list of names.
+ @param const Language::EN_Language& Language code.
+ @param NameList_T& Empty list of names, which will be filled by the
+ method if a list exists for that language code.
+ @return bool Whether or not such a list exists for the given
+ language. */
+ bool getNameList (const Language::EN_Language&, NameList_T&) const;
+
// ///////// Setters ////////
/** Set the Place code. */
@@ -146,6 +156,14 @@
public:
+ // /////////// Business methods /////////
+ /** Create a Location structure, which is a light copy
+ of the Place object. That (Location) structure is passed
+ back to the caller of the service. */
+ Location createLocation() const;
+
+
+ public:
// ///////// Display methods ////////
/** Dump a Business Object into an output stream.
@param ostream& the output stream. */
@@ -163,17 +181,17 @@
/** Get a string describing the whole key (differentiating two objects
at any level). */
- const std::string describeKey() const;
+ std::string describeKey() const;
/** Get a string describing the short key (differentiating two objects
at the same level). */
- const std::string describeShortKey() const;
+ std::string describeShortKey() const;
/** Display the full Place context. */
- const std::string display() const;
+ std::string display() const;
/** Display a short Place context. */
- const std::string shortDisplay() const;
+ std::string shortDisplay() const;
private:
Modified: trunk/opentrep/opentrep/bom/PlaceHolder.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/PlaceHolder.cpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/PlaceHolder.cpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -27,13 +27,13 @@
}
// //////////////////////////////////////////////////////////////////////
- const std::string PlaceHolder::describeShortKey() const {
+ std::string PlaceHolder::describeShortKey() const {
std::ostringstream oStr;
return oStr.str();
}
// //////////////////////////////////////////////////////////////////////
- const std::string PlaceHolder::describeKey() const {
+ std::string PlaceHolder::describeKey() const {
return describeShortKey();
}
@@ -78,4 +78,18 @@
void PlaceHolder::fromStream (std::istream& ioIn) {
}
+ // //////////////////////////////////////////////////////////////////////
+ void PlaceHolder::createLocations (LocationList_T& ioLocationList) const {
+
+ for (PlaceOrderedList_T::const_iterator itPlace = _placeOrderedList.begin();
+ itPlace != _placeOrderedList.end(); ++itPlace) {
+ const Place* lPlace_ptr = *itPlace;
+ assert (lPlace_ptr != NULL);
+
+ const Location& lLocation = lPlace_ptr->createLocation();
+ ioLocationList.push_back (lLocation);
+ }
+ }
+
+
}
Modified: trunk/opentrep/opentrep/bom/PlaceHolder.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/PlaceHolder.hpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/PlaceHolder.hpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -6,6 +6,7 @@
// //////////////////////////////////////////////////////////////////////
// OpenTREP
#include <opentrep/OPENTREP_Types.hpp>
+#include <opentrep/Location.hpp>
#include <opentrep/bom/BomAbstract.hpp>
#include <opentrep/bom/PlaceList.hpp>
@@ -27,8 +28,12 @@
public:
// /////////// Business methods /////////
-
+ /** Create the list of Location structures, which are light copies
+ of the Place objects. Those (Location) structures are passed
+ back to the caller of the service. */
+ void createLocations (LocationList_T&) const;
+
public:
// /////////// Display support methods /////////
/** Dump a Business Object into an output stream.
@@ -47,11 +52,11 @@
/** Get a string describing the whole key (differentiating two objects
at any level). */
- const std::string describeKey() const;
+ std::string describeKey() const;
/** Get a string describing the short key (differentiating two objects
at the same level). */
- const std::string describeShortKey() const;
+ std::string describeShortKey() const;
private:
Modified: trunk/opentrep/opentrep/bom/Result.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/Result.cpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/Result.cpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -27,14 +27,14 @@
}
// //////////////////////////////////////////////////////////////////////
- const std::string Result::describeShortKey() const {
+ std::string Result::describeShortKey() const {
std::ostringstream oStr;
oStr << _queryString;
return oStr.str();
}
// //////////////////////////////////////////////////////////////////////
- const std::string Result::describeKey() const {
+ std::string Result::describeKey() const {
return describeShortKey();
}
Modified: trunk/opentrep/opentrep/bom/Result.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/Result.hpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/Result.hpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -71,11 +71,11 @@
/** Get a string describing the whole key (differentiating two objects
at any level). */
- const std::string describeKey() const;
+ std::string describeKey() const;
/** Get a string describing the short key (differentiating two objects
at the same level). */
- const std::string describeShortKey() const;
+ std::string describeShortKey() const;
private:
Modified: trunk/opentrep/opentrep/bom/ResultHolder.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/ResultHolder.cpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/ResultHolder.cpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -31,14 +31,14 @@
}
// //////////////////////////////////////////////////////////////////////
- const std::string ResultHolder::describeShortKey() const {
+ std::string ResultHolder::describeShortKey() const {
std::ostringstream oStr;
oStr << _queryString;
return oStr.str();
}
// //////////////////////////////////////////////////////////////////////
- const std::string ResultHolder::describeKey() const {
+ std::string ResultHolder::describeKey() const {
return describeShortKey();
}
Modified: trunk/opentrep/opentrep/bom/ResultHolder.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/ResultHolder.hpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/ResultHolder.hpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -65,11 +65,11 @@
/** Get a string describing the whole key (differentiating two objects
at any level). */
- const std::string describeKey() const;
+ std::string describeKey() const;
/** Get a string describing the short key (differentiating two objects
at the same level). */
- const std::string describeShortKey() const;
+ std::string describeShortKey() const;
private:
Modified: trunk/opentrep/opentrep/bom/World.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/World.cpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/World.cpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -25,13 +25,13 @@
}
// //////////////////////////////////////////////////////////////////////
- const std::string World::describeShortKey() const {
+ std::string World::describeShortKey() const {
std::ostringstream oStr;
return oStr.str();
}
// //////////////////////////////////////////////////////////////////////
- const std::string World::describeKey() const {
+ std::string World::describeKey() const {
return describeShortKey();
}
@@ -50,14 +50,14 @@
}
// //////////////////////////////////////////////////////////////////////
- const std::string World::shortDisplay() const {
+ std::string World::shortDisplay() const {
std::ostringstream oStr;
oStr << describeKey() << " one world " << std::endl;
return oStr.str();
}
// //////////////////////////////////////////////////////////////////////
- const std::string World::display() const {
+ std::string World::display() const {
std::ostringstream oStr;
oStr << shortDisplay();
Modified: trunk/opentrep/opentrep/bom/World.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/World.hpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/bom/World.hpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -50,17 +50,17 @@
/** Get a string describing the whole key (differentiating two objects
at any level). */
- const std::string describeKey() const;
+ std::string describeKey() const;
/** Get a string describing the short key (differentiating two objects
at the same level). */
- const std::string describeShortKey() const;
+ std::string describeShortKey() const;
/** Display the full World context. */
- const std::string display() const;
+ std::string display() const;
/** Display a short World context. */
- const std::string shortDisplay() const;
+ std::string shortDisplay() const;
/** Retrieve a generic BOM object from the dedicated list. */
GenericBom_T getGenericBom (const XapianDocID_T& iDocID) const;
Modified: trunk/opentrep/opentrep/command/RequestInterpreter.cpp
===================================================================
--- trunk/opentrep/opentrep/command/RequestInterpreter.cpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/command/RequestInterpreter.cpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -1,6 +1,8 @@
// //////////////////////////////////////////////////////////////////////
// Import section
// //////////////////////////////////////////////////////////////////////
+// C
+#include <cassert>
// STL
#include <istream>
#include <ostream>
@@ -23,63 +25,47 @@
#include <xapian.h>
namespace OPENTREP {
-
+
// //////////////////////////////////////////////////////////////////////
- void RequestInterpreter::
- interpretTravelRequest (soci::session& ioSociSession,
- const TravelDatabaseName_T& iTravelDatabaseName,
- const TravelQuery_T& iTravelQuery) {
+ void createResults (const DocumentList_T& iDocumentList,
+ const Xapian::Database& iXapianDatabase,
+ ResultHolder& ioResultHolder) {
- try {
-
- // Make the database
- Xapian::Database lXapianDatabase (iTravelDatabaseName);
-
- // Create a ResultHolder object
- ResultHolder& lResultHolder =
- FacResultHolder::instance().create (iTravelQuery, lXapianDatabase);
-
- // DEBUG
- OPENTREP_LOG_DEBUG (std::endl
- << "=========================================");
+ // Back-up the (retrieved) matching Xapian documents into still
+ // to-be-created Result objects.
+ for (DocumentList_T::const_iterator itDoc = iDocumentList.begin();
+ itDoc != iDocumentList.end(); ++itDoc) {
+ // Retrieve both the Xapian document object and the corresponding
+ // matching percentage (most of the time, it is 100%)
+ const Document& lMatchingDocument = *itDoc;
- // Main algorithm
- DocumentList_T lDocumentList;
- lResultHolder.searchString (lDocumentList);
-
- // Back-up the (retrieved) matching Xapian documents into still
- // to-be-created Result objects.
- for (DocumentList_T::const_iterator itDoc = lDocumentList.begin();
- itDoc != lDocumentList.end(); ++itDoc) {
- // Retrieve both the Xapian document object and the corresponding
- // matching percentage (most of the time, it is 100%)
- const Document& lMatchingDocument = *itDoc;
-
- // Create a Result object
- Result& lResult = FacResult::instance().create (lXapianDatabase);
-
- // Fill the Result object with both the corresponding Document object
- // and its associated query string
- lResult.setMatchingDocument (lMatchingDocument);
-
- // Add the Result object (holding the list of matching
- // documents) to the dedicated list.
- FacResultHolder::initLinkWithResult (lResultHolder, lResult);
- }
-
- // DEBUG
- OPENTREP_LOG_DEBUG (std::endl
- << "========================================="
- << std::endl << "Matching list: " << std::endl
- << lResultHolder.toString()
- << "========================================="
- << std::endl << std::endl);
-
- // Create a PlaceHolder object, to collect the matching Place objects
- PlaceHolder& lPlaceHolder = FacPlaceHolder::instance().create();
+ // Create a Result object
+ Result& lResult = FacResult::instance().create (iXapianDatabase);
+
+ // Fill the Result object with both the corresponding Document object
+ // and its associated query string
+ lResult.setMatchingDocument (lMatchingDocument);
+
+ // Add the Result object (holding the list of matching
+ // documents) to the dedicated list.
+ FacResultHolder::initLinkWithResult (ioResultHolder, lResult);
+ }
+ // DEBUG
+ OPENTREP_LOG_DEBUG (std::endl
+ << "========================================="
+ << std::endl << "Matching list: " << std::endl
+ << ioResultHolder.toString()
+ << "========================================="
+ << std::endl << std::endl);
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void createPlaces (const ResultHolder& iResultHolder,
+ soci::session& ioSociSession, PlaceHolder& ioPlaceHolder) {
+
// Browse the list of result objects
- const ResultList_T& lResultList = lResultHolder.getResultList();
+ const ResultList_T& lResultList = iResultHolder.getResultList();
for (ResultList_T::const_iterator itResult = lResultList.begin();
itResult != lResultList.end(); ++itResult) {
// Retrieve the result object
@@ -108,7 +94,7 @@
if (hasRetrievedPlace == true) {
// Insert the Place object within the PlaceHolder object
- FacPlaceHolder::initLinkWithPlace (lPlaceHolder, lPlace);
+ FacPlaceHolder::initLinkWithPlace (ioPlaceHolder, lPlace);
// DEBUG
OPENTREP_LOG_DEBUG ("Retrieved Document: " << lPlace.toString());
@@ -118,18 +104,63 @@
OPENTREP_LOG_DEBUG ("No retrieved Document for ID = " << lDocID);
}
}
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ NbOfMatches_T RequestInterpreter::
+ interpretTravelRequest (soci::session& ioSociSession,
+ const TravelDatabaseName_T& iTravelDatabaseName,
+ const TravelQuery_T& iTravelQuery,
+ LocationList_T& ioLocationList) {
+ NbOfMatches_T oNbOfMatches = 0;
+ // Create a PlaceHolder object, to collect the matching Place objects
+ PlaceHolder& lPlaceHolder = FacPlaceHolder::instance().create();
+
+ try {
+
+ // Make the database
+ Xapian::Database lXapianDatabase (iTravelDatabaseName);
+
+ // Create a ResultHolder object
+ ResultHolder& lResultHolder =
+ FacResultHolder::instance().create (iTravelQuery, lXapianDatabase);
+
// DEBUG
- OPENTREP_LOG_NOTIFICATION (std::endl
- << "========================================="
- << std::endl << "Summary:" << std::endl
- << lPlaceHolder.toShortString() << std::endl
- << "========================================="
- << std::endl);
+ OPENTREP_LOG_DEBUG (std::endl
+ << "=========================================");
+ // Main algorithm
+ DocumentList_T lDocumentList;
+ lResultHolder.searchString (lDocumentList);
+
+ /** Create the list of Result objects corresponding to the list
+ of documents. */
+ createResults (lDocumentList, lXapianDatabase, lResultHolder);
+
+ /** Create the list of Place objects, for each of which a
+ look-up is made in the SQL database (e.g., MySQL or Oracle)
+ to retrieve complementary data. */
+ createPlaces (lResultHolder, ioSociSession, lPlaceHolder);
+
} catch (const Xapian::Error& error) {
OPENTREP_LOG_ERROR ("Exception: " << error.get_msg());
}
+
+ // DEBUG
+ OPENTREP_LOG_DEBUG (std::endl
+ << "========================================="
+ << std::endl << "Summary:" << std::endl
+ << lPlaceHolder.toShortString() << std::endl
+ << "========================================="
+ << std::endl);
+
+ /** Create the list of Location structures, which are light copies
+ of the Place objects. Those (Location) structures are passed
+ back to the caller of the service. */
+ lPlaceHolder.createLocations (ioLocationList);
+
+ return oNbOfMatches;
}
}
Modified: trunk/opentrep/opentrep/command/RequestInterpreter.hpp
===================================================================
--- trunk/opentrep/opentrep/command/RequestInterpreter.hpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/command/RequestInterpreter.hpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -8,6 +8,7 @@
#include <string>
// OpenTrep
#include <opentrep/OPENTREP_Types.hpp>
+#include <opentrep/Location.hpp>
// Forward declarations
namespace soci {
@@ -23,10 +24,19 @@
class RequestInterpreter {
friend class OPENTREP_Service;
private:
- /** Interpret a search query. */
- static void interpretTravelRequest (soci::session&,
- const TravelDatabaseName_T&,
- const TravelQuery_T&);
+ /** Match the given string, thanks to a full-text search on the
+ underlying Xapian index (named "database").
+ @param soci::session& SQL Database (e.g., MySQL, Oracle) session.
+ @param const TravelDatabaseName_T& Filepath to the Xapian database.
+ @param const std::string& (Travel-related) query string (e.g.,
+ "sna francicso rio de janero lso angles reykyavki nce iev mow").
+ @param LocationList_T& List of (geographical) locations, if any,
+ matching the given query string.
+ @return NbOfMatches_T Number of matches. */
+ static NbOfMatches_T interpretTravelRequest (soci::session&,
+ const TravelDatabaseName_T&,
+ const TravelQuery_T&,
+ LocationList_T&);
private:
/** Constructors. */
Modified: trunk/opentrep/opentrep/service/OPENTREP_Service.cpp
===================================================================
--- trunk/opentrep/opentrep/service/OPENTREP_Service.cpp 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/service/OPENTREP_Service.cpp 2009-07-18 18:32:04 UTC (rev 138)
@@ -19,8 +19,9 @@
namespace OPENTREP {
// //////////////////////////////////////////////////////////////////////
- OPENTREP_Service::OPENTREP_Service (std::ostream& ioLogStream,
- const std::string& iXapianDatabaseFilepath)
+ OPENTREP_Service::
+ OPENTREP_Service (std::ostream& ioLogStream,
+ const std::string& iXapianDatabaseFilepath)
: _opentrepServiceContext (NULL) {
init (ioLogStream, iXapianDatabaseFilepath);
}
@@ -43,6 +44,12 @@
}
// //////////////////////////////////////////////////////////////////////
+ void logInit (const LOG::EN_LogLevel iLogLevel,
+ std::ostream& ioLogOutputFile) {
+ Logger::instance().setLogParameters (iLogLevel, ioLogOutputFile);
+ }
+
+ // //////////////////////////////////////////////////////////////////////
void OPENTREP_Service::init (std::ostream& ioLogStream,
const std::string& iTravelDatabaseName) {
// Set the log file
@@ -65,12 +72,6 @@
}
// //////////////////////////////////////////////////////////////////////
- void OPENTREP_Service::logInit (const LOG::EN_LogLevel iLogLevel,
- std::ostream& ioLogOutputFile) {
- Logger::instance().setLogParameters (iLogLevel, ioLogOutputFile);
- }
-
- // //////////////////////////////////////////////////////////////////////
void OPENTREP_Service::finalise () {
assert (_opentrepServiceContext != NULL);
@@ -109,8 +110,9 @@
}
// //////////////////////////////////////////////////////////////////////
- void OPENTREP_Service::
- interpretTravelRequest (const std::string& iTravelQuery) {
+ NbOfMatches_T OPENTREP_Service::
+ interpretTravelRequest (const std::string& iTravelQuery,
+ LocationList_T& ioLocationList) {
if (_opentrepServiceContext == NULL) {
throw NonInitialisedServiceException();
}
@@ -128,9 +130,10 @@
// Delegate the query execution to the dedicated command
BasChronometer lRequestInterpreterChronometer;
lRequestInterpreterChronometer.start();
- RequestInterpreter::interpretTravelRequest (lSociSession,
- lTravelDatabaseName,
- iTravelQuery);
+ const NbOfMatches_T nbOfMatches =
+ RequestInterpreter::interpretTravelRequest (lSociSession,
+ lTravelDatabaseName,
+ iTravelQuery, ioLocationList);
const double lRequestInterpreterMeasure =
lRequestInterpreterChronometer.elapsed();
@@ -138,6 +141,8 @@
OPENTREP_LOG_DEBUG ("Match query on Xapian database (index): "
<< lRequestInterpreterMeasure << " - "
<< lOPENTREP_ServiceContext.display());
+
+ return nbOfMatches;
}
}
Modified: trunk/opentrep/opentrep/sources.mk
===================================================================
--- trunk/opentrep/opentrep/sources.mk 2009-07-18 15:15:40 UTC (rev 137)
+++ trunk/opentrep/opentrep/sources.mk 2009-07-18 18:32:04 UTC (rev 138)
@@ -1,4 +1,6 @@
service_h_sources = \
$(top_srcdir)/opentrep/OPENTREP_Types.hpp \
+ $(top_srcdir)/opentrep/OPENTREP_Abstract.hpp \
+ $(top_srcdir)/opentrep/Location.hpp \
$(top_srcdir)/opentrep/OPENTREP_Service.hpp
service_cc_sources =
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-18 15:15:47
|
Revision: 137
http://opentrep.svn.sourceforge.net/opentrep/?rev=137&view=rev
Author: denis_arnaud
Date: 2009-07-18 15:15:40 +0000 (Sat, 18 Jul 2009)
Log Message:
-----------
[Dev] Better debugging logs.
Modified Paths:
--------------
trunk/opentrep/opentrep/bom/DocumentList.hpp
trunk/opentrep/opentrep/bom/PlaceList.hpp
trunk/opentrep/opentrep/bom/Result.cpp
trunk/opentrep/opentrep/bom/Result.hpp
trunk/opentrep/opentrep/bom/ResultHolder.cpp
trunk/opentrep/opentrep/bom/ResultHolder.hpp
trunk/opentrep/opentrep/bom/StringMatcher.cpp
trunk/opentrep/opentrep/bom/StringMatcher.hpp
trunk/opentrep/opentrep/bom/sources.mk
trunk/opentrep/opentrep/command/RequestInterpreter.cpp
trunk/opentrep/opentrep/factory/FacPlaceHolder.cpp
Added Paths:
-----------
trunk/opentrep/opentrep/bom/Document.cpp
trunk/opentrep/opentrep/bom/Document.hpp
Added: trunk/opentrep/opentrep/bom/Document.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/Document.cpp (rev 0)
+++ trunk/opentrep/opentrep/bom/Document.cpp 2009-07-18 15:15:40 UTC (rev 137)
@@ -0,0 +1,59 @@
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// C
+#include <cassert>
+// OpenTREP
+#include <opentrep/bom/Document.hpp>
+
+namespace OPENTREP {
+
+ // //////////////////////////////////////////////////////////////////////
+ Document::Document () {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ Document::Document (const Document& iDocument)
+ : _queryString (iDocument._queryString),
+ _document (iDocument._document),
+ _documentList (iDocument._documentList) {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ Document::~Document () {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ const std::string Document::describeShortKey() const {
+ std::ostringstream oStr;
+ oStr << _queryString;
+ return oStr.str();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ const std::string Document::describeKey() const {
+ return describeShortKey();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ std::string Document::toString() const {
+ std::ostringstream oStr;
+ oStr << describeShortKey() << std::endl;
+
+ const Xapian::docid& lDocID = _document.get_docid();
+ oStr << "Document ID " << lDocID << "\t" << _percentage
+ << "% [" << _document.get_data() << "]";
+
+ return oStr.str();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void Document::toStream (std::ostream& ioOut) const {
+ ioOut << toString();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void Document::fromStream (std::istream& ioIn) {
+ }
+
+}
Added: trunk/opentrep/opentrep/bom/Document.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/Document.hpp (rev 0)
+++ trunk/opentrep/opentrep/bom/Document.hpp 2009-07-18 15:15:40 UTC (rev 137)
@@ -0,0 +1,123 @@
+#ifndef __OPENTREP_BOM_DOCUMENT_HPP
+#define __OPENTREP_BOM_DOCUMENT_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// STL
+#include <list>
+// OpenTREP
+#include <opentrep/OPENTREP_Types.hpp>
+#include <opentrep/bom/BomAbstract.hpp>
+// Xapian
+#include <xapian.h>
+
+namespace OPENTREP {
+
+ // //////////////// Type definitions /////////////////
+ /** List of Xapian documents. */
+ typedef std::list<Xapian::Document> XapianDocumentList_T;
+
+
+ // //////////////// Main Class /////////////////
+ /** Structure wrapping a Xapian document having matched part of a
+ given query string.
+ <br>It is a structure, as it is aimed to be temporary, the time
+ a Result object be created with the corresponding content. */
+ struct Document : public BomAbstract {
+ public:
+ // ////////////////// Getters ////////////////
+ /** Get the query string. */
+ const TravelQuery_T& getTravelQuery() {
+ return _queryString;
+ }
+
+ /** Get the matching Xapian document. */
+ const Xapian::Document& getXapianDocument() const {
+ return _document;
+ }
+
+ /** Get the matching percentage associated to the Xapian document. */
+ const Xapian::percent& getXapianPercentage() const {
+ return _percentage;
+ }
+
+ /** Get the extra list of matching Xapian documents. */
+ const XapianDocumentList_T& getExtraDocumentList() const {
+ return _documentList;
+ }
+
+
+ // ////////////////// Setters ////////////////
+ void setQueryString (const TravelQuery_T& iQueryString) {
+ _queryString = iQueryString;
+ }
+
+ /** Set the matching Xapian document. */
+ void setXapianDocument (const Xapian::Document& iMatchingDocument) {
+ _document = iMatchingDocument;
+ }
+
+ /** Set the matching percentage associated to the Xapian document. */
+ void setXapianPercentage (const Xapian::percent& iPercentage) {
+ _percentage = iPercentage;
+ }
+
+ /** Add a matching Xapian document (having the same matching percentage). */
+ void addExtraDocument (const Xapian::Document& iMatchingDocument) {
+ _documentList.push_back (iMatchingDocument);
+ }
+
+
+ public:
+ // /////////// Display support methods /////////
+ /** Dump a Business Object into an output stream.
+ @param ostream& the output stream. */
+ void toStream (std::ostream& ioOut) const;
+
+ /** Read a Business Object from an input stream.
+ @param istream& the input stream. */
+ void fromStream (std::istream& ioIn);
+
+ /** Get the serialised version of the Business Object. */
+ std::string toString() const;
+
+ /** Get a string describing the whole key (differentiating two objects
+ at any level). */
+ const std::string describeKey() const;
+
+ /** Get a string describing the short key (differentiating two objects
+ at the same level). */
+ const std::string describeShortKey() const;
+
+
+ public:
+ // //////////////// Constructors and Destructors /////////////
+ /** Default constructor. */
+ Document ();
+ /** Default copy constructor. */
+ Document (const Document&);
+ /** Default destructor. */
+ ~Document ();
+
+
+ private:
+ // ///////////////// Attributes //////////////////
+ /** Query string with which a Xapian full text search is done. */
+ TravelQuery_T _queryString;
+
+ /** Matching percentage, as returned by the Xapian full text search.
+ <br>Generally, that percentage is equal to, or close to, 100%. */
+ Xapian::percent _percentage;
+
+ /** Matching document, as returned by the Xapian full text search. */
+ Xapian::Document _document;
+
+ /** List of Xapian documents having the same matching percentage.
+ <br>Hence, any of those other Xapian documents could have been
+ chosen, instead of the main one. */
+ XapianDocumentList_T _documentList;
+ };
+
+}
+#endif // __OPENTREP_BOM_DOCUMENT_HPP
Modified: trunk/opentrep/opentrep/bom/DocumentList.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/DocumentList.hpp 2009-07-18 09:33:50 UTC (rev 136)
+++ trunk/opentrep/opentrep/bom/DocumentList.hpp 2009-07-18 15:15:40 UTC (rev 137)
@@ -8,20 +8,14 @@
#include <list>
// OpenTREP
#include <opentrep/OPENTREP_Types.hpp>
+#include <opentrep/bom/Document.hpp>
// Xapian
#include <xapian.h>
namespace OPENTREP {
- /** Xapian document and its associated matching percentage. */
- typedef std::pair<Xapian::percent, Xapian::Document> MatchingDocument_T;
-
- /** A matching Xapian document, along with the query string which it
- matches. */
- typedef std::pair<TravelQuery_T, MatchingDocument_T> QueryAndDocument_T;
-
/** List of matching Xapian documents. */
- typedef std::list<QueryAndDocument_T> DocumentList_T;
+ typedef std::list<Document> DocumentList_T;
}
#endif // __OPENTREP_BOM_DOCUMENTLIST_HPP
Modified: trunk/opentrep/opentrep/bom/PlaceList.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/PlaceList.hpp 2009-07-18 09:33:50 UTC (rev 136)
+++ trunk/opentrep/opentrep/bom/PlaceList.hpp 2009-07-18 15:15:40 UTC (rev 137)
@@ -14,9 +14,10 @@
class Place;
// ///////////// Type definitions ////////////////////
- typedef std::size_t PlaceID_T;
+ // typedef std::size_t PlaceID_T;
// typedef std::map<PlaceID_T, Place*> PlaceDirectList_T;
- typedef std::map<std::string, Place*> PlaceList_T;
+
+ typedef std::multimap<std::string, Place*> PlaceList_T;
typedef std::list<Place*> PlaceOrderedList_T;
}
Modified: trunk/opentrep/opentrep/bom/Result.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/Result.cpp 2009-07-18 09:33:50 UTC (rev 136)
+++ trunk/opentrep/opentrep/bom/Result.cpp 2009-07-18 15:15:40 UTC (rev 137)
@@ -41,14 +41,7 @@
// //////////////////////////////////////////////////////////////////////
std::string Result::toString() const {
std::ostringstream oStr;
- oStr << describeShortKey() << std::endl;
-
- const Xapian::percent& lPercentage = _matchingDocument.first;
- const Xapian::Document& lDocument = _matchingDocument.second;
- const Xapian::docid& lDocID = lDocument.get_docid();
- oStr << "Document ID " << lDocID << "\t" << lPercentage
- << "% [" << lDocument.get_data() << "]" << std::endl;
-
+ oStr << _matchingDocument.toString();
return oStr.str();
}
Modified: trunk/opentrep/opentrep/bom/Result.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/Result.hpp 2009-07-18 09:33:50 UTC (rev 136)
+++ trunk/opentrep/opentrep/bom/Result.hpp 2009-07-18 15:15:40 UTC (rev 137)
@@ -7,7 +7,7 @@
// OpenTREP
#include <opentrep/OPENTREP_Types.hpp>
#include <opentrep/bom/BomAbstract.hpp>
-#include <opentrep/bom/DocumentList.hpp>
+#include <opentrep/bom/Document.hpp>
namespace OPENTREP {
@@ -27,19 +27,19 @@
/** Get the Matching Xapian document object, along with its
corresponding matching percentage. */
- const MatchingDocument_T& getMatchingDocument() const {
+ const Document& getMatchingDocument() const {
return _matchingDocument;
}
/** Retrieve the percentage corresponding to the matching Xapian
document object. */
- const Xapian::percent& getPercentage() const {
- return _matchingDocument.first;
+ const Xapian::percent& getXapianPercentage() const {
+ return _matchingDocument.getXapianPercentage();
}
/** Retrieve the matching Xapian document object. */
- const Xapian::Document& getDocument() const {
- return _matchingDocument.second;
+ const Xapian::Document& getXapianDocument() const {
+ return _matchingDocument.getXapianDocument();
}
@@ -51,17 +51,10 @@
/** Set the matching Xapian document object and its corresponding
matching percentage. */
- void setMatchingDocument (const MatchingDocument_T& iMatchingDocument) {
+ void setMatchingDocument (const Document& iMatchingDocument) {
_matchingDocument = iMatchingDocument;
}
- /** Set the matching Xapian document object and its corresponding
- matching percentage. */
- void setQueryAndDocument (const QueryAndDocument_T& iQueryAndDocument) {
- _queryString = iQueryAndDocument.first;
- _matchingDocument = iQueryAndDocument.second;
- }
-
public:
// /////////// Display support methods /////////
@@ -112,7 +105,7 @@
/** Matching Xapian document object, along with its corresponding
matching percentage. */
- MatchingDocument_T _matchingDocument;
+ Document _matchingDocument;
};
}
Modified: trunk/opentrep/opentrep/bom/ResultHolder.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/ResultHolder.cpp 2009-07-18 09:33:50 UTC (rev 136)
+++ trunk/opentrep/opentrep/bom/ResultHolder.cpp 2009-07-18 15:15:40 UTC (rev 137)
@@ -69,7 +69,7 @@
// //////////////////////////////////////////////////////////////////////
bool ResultHolder::searchString (TravelQuery_T& ioPartialQueryString,
- MatchingDocument_T& ioMatchingDocument) {
+ Document& ioMatchingDocument) {
bool oFoundDocument = false;
// Catch any Xapian::Error exceptions thrown
@@ -78,11 +78,8 @@
bool shouldStop = false;
while (shouldStop == false) {
// DEBUG
- /*
- OPENTREP_LOG_DEBUG (std::endl << "--------------------------------"
- << std::endl << "Current query string: `" << ioPartialQueryString
- << "'");
- */
+ OPENTREP_LOG_DEBUG ("Current query string: `"
+ << ioPartialQueryString << "'");
// Retrieve the list of documents matching the query string
Xapian::MSet lMatchingSet;
@@ -125,11 +122,10 @@
bool shouldStop = false;
while (shouldStop == false) {
// DEBUG
- /*
- OPENTREP_LOG_DEBUG (std::endl
- << "================================" << std::endl
- << "Current query string: `" << lRemainingQueryString << "'");
- */
+ OPENTREP_LOG_DEBUG ("---------------------")
+ OPENTREP_LOG_DEBUG ("Remaining part of the query string: `"
+ << lRemainingQueryString << "'");
+
/**
Search with the initial full string, then by removing a word if
there was no result, then by removing another word if there was
@@ -143,14 +139,17 @@
furthest right words, so that the remaining left part be matched
against the Xapian database).
*/
- MatchingDocument_T lMatchingDocument;
+ Document lMatchingDocument;
const bool hasFoundDocument = searchString (lQueryString,
lMatchingDocument);
if (hasFoundDocument == true) {
- const QueryAndDocument_T lQueryAndDocument (lQueryString,
- lMatchingDocument);
- ioDocumentList.push_back (lQueryAndDocument);
+ lMatchingDocument.setQueryString (lQueryString);
+ ioDocumentList.push_back (lMatchingDocument);
+
+ // DEBUG
+ OPENTREP_LOG_DEBUG ("==> Matching of the query string: `"
+ << lQueryString << "'");
}
/**
Modified: trunk/opentrep/opentrep/bom/ResultHolder.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/ResultHolder.hpp 2009-07-18 09:33:50 UTC (rev 136)
+++ trunk/opentrep/opentrep/bom/ResultHolder.hpp 2009-07-18 15:15:40 UTC (rev 137)
@@ -47,7 +47,7 @@
@param TravelQuery_T& The partial query string.
@param MatchingDocument_T& The best matching Xapian document (if found).
@return bool Whether such a best matching document has been found. */
- bool searchString(TravelQuery_T& ioPartialQueryString, MatchingDocument_T&);
+ bool searchString (TravelQuery_T& ioPartialQueryString, Document&);
public:
Modified: trunk/opentrep/opentrep/bom/StringMatcher.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/StringMatcher.cpp 2009-07-18 09:33:50 UTC (rev 136)
+++ trunk/opentrep/opentrep/bom/StringMatcher.cpp 2009-07-18 15:15:40 UTC (rev 137)
@@ -4,7 +4,8 @@
// C
#include <cassert>
// STL
-#include <iostream>
+#include <istream>
+#include <ostream>
#include <sstream>
#include <string>
#include <list>
@@ -73,7 +74,7 @@
}
} catch (const Xapian::Error& error) {
- std::cerr << "Exception: " << error.get_msg() << std::endl;
+ OPENTREP_LOG_ERROR ("Exception: " << error.get_msg());
}
}
@@ -267,14 +268,14 @@
*/
} catch (const Xapian::Error& error) {
- std::cerr << "Exception: " << error.get_msg() << std::endl;
+ OPENTREP_LOG_ERROR ("Exception: " << error.get_msg());
}
}
// //////////////////////////////////////////////////////////////////////
bool StringMatcher::
extractBestMatchingDocumentFromMSet (const Xapian::MSet& iMatchingSet,
- MatchingDocument_T& ioMatchingDocument) {
+ Document& ioMatchingDocument) {
bool oFoundDocument = false;
if (iMatchingSet.empty() == true) {
@@ -290,9 +291,23 @@
same: it appears random).
*/
Xapian::MSetIterator itDoc = iMatchingSet.begin();
- ioMatchingDocument.first = itDoc.get_percent();
- ioMatchingDocument.second = itDoc.get_document();
+ const Xapian::percent& lBestPercentage = itDoc.get_percent();
+ ioMatchingDocument.setXapianPercentage (lBestPercentage);
+ ioMatchingDocument.setXapianDocument (itDoc.get_document());
+ /** Add all the Xapian documents having reached the same matching
+ percentage. */
+ for ( ; itDoc != iMatchingSet.end(); ++itDoc) {
+ const Xapian::percent& lPercentage = itDoc.get_percent();
+
+ if (lPercentage == lBestPercentage) {
+ ioMatchingDocument.addExtraDocument (itDoc.get_document());
+
+ } else {
+ break;
+ }
+ }
+
return oFoundDocument;
}
Modified: trunk/opentrep/opentrep/bom/StringMatcher.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/StringMatcher.hpp 2009-07-18 09:33:50 UTC (rev 136)
+++ trunk/opentrep/opentrep/bom/StringMatcher.hpp 2009-07-18 15:15:40 UTC (rev 137)
@@ -9,7 +9,7 @@
// OpenTREP
#include <opentrep/bom/BomAbstract.hpp>
#include <opentrep/bom/WordList.hpp>
-#include <opentrep/bom/DocumentList.hpp>
+#include <opentrep/bom/Document.hpp>
// Forward declarations
namespace Xapian {
@@ -38,8 +38,7 @@
@return bool Whether or not there was a matching document.
*/
static bool
- extractBestMatchingDocumentFromMSet (const Xapian::MSet&,
- MatchingDocument_T&);
+ extractBestMatchingDocumentFromMSet (const Xapian::MSet&, Document&);
/** Remove the word furthest at right. */
static void removeOneWord (std::string& ioQueryString);
Modified: trunk/opentrep/opentrep/bom/sources.mk
===================================================================
--- trunk/opentrep/opentrep/bom/sources.mk 2009-07-18 09:33:50 UTC (rev 136)
+++ trunk/opentrep/opentrep/bom/sources.mk 2009-07-18 15:15:40 UTC (rev 137)
@@ -9,6 +9,7 @@
$(top_srcdir)/opentrep/bom/Place.hpp \
$(top_srcdir)/opentrep/bom/PlaceList.hpp \
$(top_srcdir)/opentrep/bom/PlaceHolder.hpp \
+ $(top_srcdir)/opentrep/bom/Document.hpp \
$(top_srcdir)/opentrep/bom/DocumentList.hpp \
$(top_srcdir)/opentrep/bom/Result.hpp \
$(top_srcdir)/opentrep/bom/ResultList.hpp \
@@ -22,6 +23,7 @@
$(top_srcdir)/opentrep/bom/Names.cpp \
$(top_srcdir)/opentrep/bom/Place.cpp \
$(top_srcdir)/opentrep/bom/PlaceHolder.cpp \
+ $(top_srcdir)/opentrep/bom/Document.cpp \
$(top_srcdir)/opentrep/bom/Result.cpp \
$(top_srcdir)/opentrep/bom/ResultHolder.cpp \
$(top_srcdir)/opentrep/bom/StringMatcher.cpp
Modified: trunk/opentrep/opentrep/command/RequestInterpreter.cpp
===================================================================
--- trunk/opentrep/opentrep/command/RequestInterpreter.cpp 2009-07-18 09:33:50 UTC (rev 136)
+++ trunk/opentrep/opentrep/command/RequestInterpreter.cpp 2009-07-18 15:15:40 UTC (rev 137)
@@ -39,6 +39,10 @@
ResultHolder& lResultHolder =
FacResultHolder::instance().create (iTravelQuery, lXapianDatabase);
+ // DEBUG
+ OPENTREP_LOG_DEBUG (std::endl
+ << "=========================================");
+
// Main algorithm
DocumentList_T lDocumentList;
lResultHolder.searchString (lDocumentList);
@@ -49,14 +53,14 @@
itDoc != lDocumentList.end(); ++itDoc) {
// Retrieve both the Xapian document object and the corresponding
// matching percentage (most of the time, it is 100%)
- const QueryAndDocument_T& lQueryAndDocument = *itDoc;
+ const Document& lMatchingDocument = *itDoc;
// Create a Result object
Result& lResult = FacResult::instance().create (lXapianDatabase);
// Fill the Result object with both the corresponding Document object
// and its associated query string
- lResult.setQueryAndDocument (lQueryAndDocument);
+ lResult.setMatchingDocument (lMatchingDocument);
// Add the Result object (holding the list of matching
// documents) to the dedicated list.
@@ -83,8 +87,9 @@
assert (lResult_ptr != NULL);
// Retrieve the parameters of the best matching document
- const Xapian::Document& lDocument = lResult_ptr->getDocument();
- const Xapian::percent& lDocPercentage = lResult_ptr->getPercentage();
+ const Xapian::Document& lDocument = lResult_ptr->getXapianDocument();
+ const Xapian::percent& lDocPercentage =
+ lResult_ptr->getXapianPercentage();
const Xapian::docid& lDocID = lDocument.get_docid();
const std::string& lDocData = lDocument.get_data();
Modified: trunk/opentrep/opentrep/factory/FacPlaceHolder.cpp
===================================================================
--- trunk/opentrep/opentrep/factory/FacPlaceHolder.cpp 2009-07-18 09:33:50 UTC (rev 136)
+++ trunk/opentrep/opentrep/factory/FacPlaceHolder.cpp 2009-07-18 15:15:40 UTC (rev 137)
@@ -59,15 +59,16 @@
ioPlace._placeHolder = &ioPlaceHolder;
// Add the Place to the PlaceHolder internal map (of Place objects)
- const bool insertSucceeded = ioPlaceHolder._placeList.
- insert (PlaceList_T::value_type (ioPlace.describeShortKey(),
- &ioPlace)).second;
- if (insertSucceeded == false) {
- OPENTREP_LOG_ERROR ("Insertion failed for "
- << ioPlaceHolder.describeKey()
- << " and " << ioPlace.describeShortKey());
- assert (insertSucceeded == true);
- }
+ // const bool insertSucceeded =
+ ioPlaceHolder._placeList.
+ insert (PlaceList_T::value_type (ioPlace.describeShortKey(), &ioPlace));
+
+// if (insertSucceeded == false) {
+// OPENTREP_LOG_ERROR ("Insertion failed for "
+// << ioPlaceHolder.describeKey()
+// << " and " << ioPlace.describeShortKey());
+// assert (insertSucceeded == true);
+// }
// Add the Place to the PlaceHolder internal list (of Place objects)
ioPlaceHolder._placeOrderedList.push_back (&ioPlace);
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-18 09:33:57
|
Revision: 136
http://opentrep.svn.sourceforge.net/opentrep/?rev=136&view=rev
Author: denis_arnaud
Date: 2009-07-18 09:33:50 +0000 (Sat, 18 Jul 2009)
Log Message:
-----------
[DB] Fixed the issue with location of the CSV files.
Modified Paths:
--------------
trunk/opentrep/db/data/ref_place_names.csv
trunk/opentrep/db/maintenance/create_and_fill_mysql_db.sh
trunk/opentrep/db/maintenance/tables/create_and_fill_mysql_db.sql
trunk/opentrep/opentrep/batches/searcher.cpp
trunk/opentrep/opentrep/bom/StringMatcher.cpp
Modified: trunk/opentrep/db/data/ref_place_names.csv
===================================================================
--- trunk/opentrep/db/data/ref_place_names.csv 2009-07-18 00:50:00 UTC (rev 135)
+++ trunk/opentrep/db/data/ref_place_names.csv 2009-07-18 09:33:50 UTC (rev 136)
@@ -5057,7 +5057,7 @@
en,reg,reggio calabria,reggio calabria,reggio calabria/it:t menniti
en,reh,rehoboth beach,rehoboth beach,rehoboth beach/de/us
en,rei,regina,regina,regina/gf
-en,rek,reykjavik,reykjavik,reykjavik/is
+en,rek,reykjavik,reykjavik,reykjavik/is,reykjavik main,reykjavik city
en,rel,trelew,trelew,trelew/cb/ar
en,ren,orenburg,orenburg,orenburg/ru
en,reo,rome,rome,rome/or/us:state
Modified: trunk/opentrep/db/maintenance/create_and_fill_mysql_db.sh
===================================================================
--- trunk/opentrep/db/maintenance/create_and_fill_mysql_db.sh 2009-07-18 00:50:00 UTC (rev 135)
+++ trunk/opentrep/db/maintenance/create_and_fill_mysql_db.sh 2009-07-18 09:33:50 UTC (rev 136)
@@ -39,20 +39,34 @@
# Database Name
DB_NAME="opentrep"
+# Check file existence
+function checkSQLFile() {
+ if [ ! -r ${SQL_FILE} ]; then
+ echo
+ echo "The ${SQL_FILE} SQL file can not be found"
+ echo
+ exit -1;
+ fi
+}
+
# Create the database
function createDatabase() {
+ checkSQLFile
echo "The '${DB_NAME}' database will be created:"
mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} mysql < ${SQL_FILE}
}
# Scan a SQL script for the names of (database) tables
function createTable() {
+ checkSQLFile
echo "The ref_place_details and ref_place_names tables will be created:"
mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} < ${SQL_FILE}
}
#
function loadData() {
+ SQL_FILE=${SQL_LOADER_FILE}
+ checkSQLFile
echo "The ref_place_details and ref_place_names tables will be filled from ../data/*.csv files:"
mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} < ${SQL_LOADER_FILE}
echo "Done"
@@ -74,15 +88,15 @@
}
# Database
-SQL_FILE="../data_structure/ref_db.sql"
+SQL_FILE="tables/ref_db.sql"
createDatabase
# Table: Airport and City
-SQL_FILE="../data_structure/ref_city.sql"
+SQL_FILE="tables/ref_city.sql"
createTable
# Load data into the table
-SQL_LOADER_FILE="create_and_fill_mysql_db.sql"
+SQL_LOADER_FILE="tables/create_and_fill_mysql_db.sql"
loadData
# Trim the spaces from the state_code field of the ref_place_details table
Modified: trunk/opentrep/db/maintenance/tables/create_and_fill_mysql_db.sql
===================================================================
--- trunk/opentrep/db/maintenance/tables/create_and_fill_mysql_db.sql 2009-07-18 00:50:00 UTC (rev 135)
+++ trunk/opentrep/db/maintenance/tables/create_and_fill_mysql_db.sql 2009-07-18 09:33:50 UTC (rev 136)
@@ -1,3 +1,8 @@
+--
+-- Note: that file is expected to be launched from the
+-- $(top_srcdir)/db/maintenance sub-directory, as the CSV files are
+-- to be found in $(top_srcdir)/db/data sub-directory
+--
--
-- Load the Airport and City geographical details into the MySQL table
Modified: trunk/opentrep/opentrep/batches/searcher.cpp
===================================================================
--- trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-18 00:50:00 UTC (rev 135)
+++ trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-18 09:33:50 UTC (rev 136)
@@ -28,7 +28,7 @@
const std::string K_OPENTREP_DEFAULT_DATABSE_FILEPATH("/tmp/opentrep/traveldb");
/** Default travel query string, to be seached against the Xapian database. */
-const std::string K_OPENTREP_DEFAULT_QUERY_STRING ("sna francicso rio de janero lso anglese reykyavki");
+const std::string K_OPENTREP_DEFAULT_QUERY_STRING ("sna francicso rio de janero lso angles reykyavki");
/** Default error distance for spelling corrections. */
const unsigned short K_OPENTREP_DEFAULT_SPELLING_ERROR_DISTANCE = 3;
Modified: trunk/opentrep/opentrep/bom/StringMatcher.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/StringMatcher.cpp 2009-07-18 00:50:00 UTC (rev 135)
+++ trunk/opentrep/opentrep/bom/StringMatcher.cpp 2009-07-18 09:33:50 UTC (rev 136)
@@ -34,7 +34,7 @@
const EditDistance_T lQueryStringSize = iPhrase.size();
- oEditDistance = lQueryStringSize / 3;
+ oEditDistance = lQueryStringSize / 4;
return oEditDistance;
}
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-18 00:50:11
|
Revision: 135
http://opentrep.svn.sourceforge.net/opentrep/?rev=135&view=rev
Author: denis_arnaud
Date: 2009-07-18 00:50:00 +0000 (Sat, 18 Jul 2009)
Log Message:
-----------
1. Re-worked the searcher, so that the Result objects be created by the command (and no longer by the BOM itself). 2. Added building makefiles for database handling.
Modified Paths:
--------------
trunk/opentrep/Makefile.am
trunk/opentrep/configure.ac
trunk/opentrep/opentrep/OPENTREP_Types.hpp
trunk/opentrep/opentrep/batches/indexer.cpp
trunk/opentrep/opentrep/batches/searcher.cpp
trunk/opentrep/opentrep/bom/DocumentList.hpp
trunk/opentrep/opentrep/bom/Result.cpp
trunk/opentrep/opentrep/bom/Result.hpp
trunk/opentrep/opentrep/bom/ResultHolder.cpp
trunk/opentrep/opentrep/bom/ResultHolder.hpp
trunk/opentrep/opentrep/bom/StringMatcher.cpp
trunk/opentrep/opentrep/bom/StringMatcher.hpp
trunk/opentrep/opentrep/command/RequestInterpreter.cpp
trunk/opentrep/opentrep/command/RequestInterpreter.hpp
trunk/opentrep/opentrep/factory/FacResult.cpp
trunk/opentrep/opentrep/factory/FacResult.hpp
trunk/opentrep/opentrep/service/OPENTREP_Service.cpp
Added Paths:
-----------
trunk/opentrep/db/
trunk/opentrep/db/Makefile.am
trunk/opentrep/db/admin/
trunk/opentrep/db/admin/Makefile.am
trunk/opentrep/db/admin/create_opentrep_user.sh
trunk/opentrep/db/admin/create_opentrep_user.sql
trunk/opentrep/db/data/Makefile.am
trunk/opentrep/db/data/sources.mk
trunk/opentrep/db/maintenance/
trunk/opentrep/db/maintenance/Makefile.am
trunk/opentrep/db/maintenance/create_and_fill_mysql_db.sh
trunk/opentrep/db/maintenance/drop_tables_from_mysql_db.sh
trunk/opentrep/db/maintenance/tables/
trunk/opentrep/db/maintenance/tables/Makefile.am
trunk/opentrep/db/maintenance/tables/create_and_fill_mysql_db.sql
trunk/opentrep/db/maintenance/tables/ref_city.sql
trunk/opentrep/db/maintenance/tables/ref_db.sql
trunk/opentrep/db/maintenance/tables/sources.mk
Removed Paths:
-------------
trunk/opentrep/db/data_structure/
trunk/opentrep/db/maintenance/ref_city.sql
trunk/opentrep/db/maintenance/ref_db.sql
trunk/opentrep/db/mysql/create_and_fill_mysql_db.sh
trunk/opentrep/db/mysql/create_and_fill_mysql_db.sql
trunk/opentrep/db/mysql/create_opentrep_user.sh
trunk/opentrep/db/mysql/create_opentrep_user.sql
trunk/opentrep/db/mysql/drop_tables_from_mysql_db.sh
trunk/opentrep/refdata/
Property Changed:
----------------
trunk/opentrep/db/data/
trunk/opentrep/opentrep/batches/
Modified: trunk/opentrep/Makefile.am
===================================================================
--- trunk/opentrep/Makefile.am 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/Makefile.am 2009-07-18 00:50:00 UTC (rev 135)
@@ -24,7 +24,7 @@
EXTRA_DIST = @PACKAGE@.spec @PACKAGE@.m4 @PACKAGE@.pc Makefile.common
# Build in these directories:
-SUBDIRS = opentrep win32 po man $(INFO_DOC_DIR) $(HTML_DOC_DIR) $(TEST_DIR)
+SUBDIRS = opentrep win32 po man $(INFO_DOC_DIR) $(HTML_DOC_DIR) db $(TEST_DIR)
# Configuration helpers
Modified: trunk/opentrep/configure.ac
===================================================================
--- trunk/opentrep/configure.ac 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/configure.ac 2009-07-18 00:50:00 UTC (rev 135)
@@ -233,6 +233,11 @@
doc/doxygen_html.cfg
doc/sourceforge/howto_release_opentrep.html
po/Makefile.in
+ db/Makefile
+ db/admin/Makefile
+ db/maintenance/Makefile
+ db/maintenance/tables/Makefile
+ db/data/Makefile
test/com/Makefile
test/parsers/Makefile
test/Makefile
Property changes on: trunk/opentrep/db
___________________________________________________________________
Added: svn:ignore
+ Makefile
Makefile.in
Added: trunk/opentrep/db/Makefile.am
===================================================================
--- trunk/opentrep/db/Makefile.am (rev 0)
+++ trunk/opentrep/db/Makefile.am 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,6 @@
+## db sub-directory
+include $(top_srcdir)/Makefile.common
+
+MAINTAINERCLEANFILES = Makefile.in
+
+SUBDIRS = admin data maintenance
Property changes on: trunk/opentrep/db/admin
___________________________________________________________________
Added: svn:ignore
+ Makefile
Makefile.in
Added: trunk/opentrep/db/admin/Makefile.am
===================================================================
--- trunk/opentrep/db/admin/Makefile.am (rev 0)
+++ trunk/opentrep/db/admin/Makefile.am 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,8 @@
+## db sub-directory
+include $(top_srcdir)/Makefile.common
+
+MAINTAINERCLEANFILES = Makefile.in
+
+MYSQL_ADMIN_FILES = create_opentrep_user.sh create_opentrep_user.sql
+
+EXTRA_DIST = $(MYSQL_ADMIN_FILES)
Copied: trunk/opentrep/db/admin/create_opentrep_user.sh (from rev 134, trunk/opentrep/refdata/mysql/create_opentrep_user.sh)
===================================================================
--- trunk/opentrep/db/admin/create_opentrep_user.sh (rev 0)
+++ trunk/opentrep/db/admin/create_opentrep_user.sh 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,52 @@
+#!/bin/sh
+#
+# Two parameters are required for this script:
+# - the administrator username
+# - the administrator password
+#
+# Two parameters are optional:
+# - the host server of the database
+# - the port of the database
+#
+
+if [ "$1" = "" -o "$2" = "" -o "$1" = "-h" -o "$1" = "--help" ];
+then
+ echo "Usage: $0 <Admin Username> <Admin password> [<Database Server Hostname> [<Database Server Port>]]"
+ echo ""
+ exit -1
+fi
+
+##
+# Database Server Hostname
+DB_HOST="localhost"
+if [ "$3" != "" ];
+then
+ DB_HOST="$3"
+fi
+
+# Database Server Port
+DB_PORT="3306"
+if [ "$4" != "" ];
+then
+ DB_PORT="$4"
+fi
+
+# Database User
+DB_USER="$1"
+
+# Database Password
+DB_PASSWD="$2"
+
+# Database Name
+DB_NAME="mysql"
+
+function createOpenTrepUser() {
+ echo "Creating the opentrep user within the database:"
+ mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} < ${SQL_FILE}
+ mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} -e "flush privileges"
+}
+
+# Creating the opentrep user
+SQL_FILE="create_opentrep_user.sql"
+createOpenTrepUser
+
Copied: trunk/opentrep/db/admin/create_opentrep_user.sql (from rev 134, trunk/opentrep/refdata/mysql/create_opentrep_user.sql)
===================================================================
--- trunk/opentrep/db/admin/create_opentrep_user.sql (rev 0)
+++ trunk/opentrep/db/admin/create_opentrep_user.sql 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,7 @@
+
+insert into `user` (`Host`, `User`, `Password`, `Select_priv`, `Insert_priv`, `Update_priv`, `Delete_priv`, `Create_priv`, `Drop_priv`, `Reload_priv`, `Shutdown_priv`, `Process_priv`, `File_priv`, `Grant_priv`, `References_priv`, `Index_priv`, `Alter_priv`, `Show_db_priv`, `Super_priv`, `Create_tmp_table_priv`, `Lock_tables_priv`, `Execute_priv`, `Repl_slave_priv`, `Repl_client_priv`, `Create_view_priv`, `Show_view_priv`, `Create_routine_priv`, `Alter_routine_priv`, `Create_user_priv`, `ssl_type`, `ssl_cipher`, `x509_issuer`, `x509_subject`, `max_questions`, `max_updates`, `max_connections`, `max_user_connections`) values
+('%', 'opentrep', '*C21B5F0DB6BBABAA20B5496E75D652982A6AC65C', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'N', 'N', 'N', 'Y', 'N', 'N', 'Y', 'Y', 'N', 'N', 'Y', 'N', 'Y', 'N', 'N', 'Y', 'Y', 'Y', 'Y', 'N', '', '', '', '', 0, 0, 0, 0),
+('localhost', 'opentrep', '*C21B5F0DB6BBABAA20B5496E75D652982A6AC65C', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'N', 'N', 'N', 'Y', 'N', 'N', 'Y', 'Y', 'N', 'N', 'Y', 'N', 'Y', 'N', 'N', 'Y', 'Y', 'Y', 'Y', 'N', '', '', '', '', 0, 0, 0, 0);
+
+flush privileges;
+
Property changes on: trunk/opentrep/db/data
___________________________________________________________________
Added: svn:ignore
+ Makefile
Makefile.in
Added: trunk/opentrep/db/data/Makefile.am
===================================================================
--- trunk/opentrep/db/data/Makefile.am (rev 0)
+++ trunk/opentrep/db/data/Makefile.am 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,26 @@
+## db/data sub-directory
+include $(top_srcdir)/Makefile.common
+include $(top_srcdir)/db/data/sources.mk
+
+datadir = @datadir@
+pkgdatadir = $(datadir)/@PACKAGE@
+dbscriptsdir = $(pkgdatadir)/db/mysql/fill_tables
+
+MAINTAINERCLEANFILES = Makefile.in Makefile
+
+noinst_DATA = $(data_mysql_sources)
+EXTRA_DIST = $(noinst_DATA)
+
+
+# Targets
+install-data-local:
+ $(mkinstalldirs) $(DESTDIR)$(dbscriptsdir); \
+ for f in $(data_mysql_sources); do \
+ $(INSTALL_DATA) $$f $(DESTDIR)$(dbscriptsdir); \
+ done
+
+uninstall-local:
+ rm -rf $(DESTDIR)$(dbscriptsdir)
+
+clean-local:
+ rm -rf *.log *.tag
Added: trunk/opentrep/db/data/sources.mk
===================================================================
--- trunk/opentrep/db/data/sources.mk (rev 0)
+++ trunk/opentrep/db/data/sources.mk 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,4 @@
+data_mysql_sources = \
+ $(top_srcdir)/db/data/ref_city.csv \
+ $(top_srcdir)/db/data/ref_place_details.csv \
+ $(top_srcdir)/db/data/ref_place_names.csv
Property changes on: trunk/opentrep/db/maintenance
___________________________________________________________________
Added: svn:ignore
+ Makefile
Makefile.in
Added: trunk/opentrep/db/maintenance/Makefile.am
===================================================================
--- trunk/opentrep/db/maintenance/Makefile.am (rev 0)
+++ trunk/opentrep/db/maintenance/Makefile.am 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,10 @@
+## db/maintenance sub-directory
+include $(top_srcdir)/Makefile.common
+
+MAINTAINERCLEANFILES = Makefile.in
+
+SUBDIRS = tables
+
+TABLE_MAINT_FILES = create_and_fill_mysql_db.sh drop_tables_from_mysql_db.sh
+
+EXTRA_DIST = $(TABLE_MAINT_FILES)
Copied: trunk/opentrep/db/maintenance/create_and_fill_mysql_db.sh (from rev 134, trunk/opentrep/refdata/mysql/create_and_fill_mysql_db.sh)
===================================================================
--- trunk/opentrep/db/maintenance/create_and_fill_mysql_db.sh (rev 0)
+++ trunk/opentrep/db/maintenance/create_and_fill_mysql_db.sh 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,98 @@
+#!/bin/sh
+#
+# One parameter is required for this script:
+# - the username
+#
+# Two parameters are optional:
+# - the host server of the database
+# - the port of the database
+#
+
+if [ "$1" = "" -o "$1" = "-h" -o "$1" = "--help" ];
+then
+ echo "Usage: $0 <Database Username> [<Database Server Hostname> [<Database Server Port>]]"
+ echo ""
+ exit -1
+fi
+
+##
+# Database Server Hostname
+DB_HOST="localhost"
+if [ "$2" != "" ];
+then
+ DB_HOST="$2"
+fi
+
+# Database Server Port
+DB_PORT="3306"
+if [ "$3" != "" ];
+then
+ DB_PORT="$3"
+fi
+
+# Database User
+DB_USER="$1"
+
+# Database Password
+DB_PASSWD="${DB_USER}"
+
+# Database Name
+DB_NAME="opentrep"
+
+# Create the database
+function createDatabase() {
+ echo "The '${DB_NAME}' database will be created:"
+ mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} mysql < ${SQL_FILE}
+}
+
+# Scan a SQL script for the names of (database) tables
+function createTable() {
+ echo "The ref_place_details and ref_place_names tables will be created:"
+ mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} < ${SQL_FILE}
+}
+
+#
+function loadData() {
+ echo "The ref_place_details and ref_place_names tables will be filled from ../data/*.csv files:"
+ mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} < ${SQL_LOADER_FILE}
+ echo "Done"
+}
+
+#
+function trimStateCode() {
+ echo "Triming the spaces from the state_code field of the ${TABLE} table:"
+ mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} -e "update ${TABLE} set city_code=NULL where city_code='';"
+ mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} -e "update ${TABLE} set state_code=NULL where state_code like '%null%';"
+ mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} -e "update ${TABLE} set state_code=NULL where length(state_code)=2;"
+ mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} -e "update ${TABLE} set state_code=substring(state_code,2,2) where length(state_code)=4;"
+}
+
+#
+function countRows() {
+ echo "Counting the rows from the ${TABLE} table:"
+ mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} -e "select count(*) from ${TABLE};"
+}
+
+# Database
+SQL_FILE="../data_structure/ref_db.sql"
+createDatabase
+
+# Table: Airport and City
+SQL_FILE="../data_structure/ref_city.sql"
+createTable
+
+# Load data into the table
+SQL_LOADER_FILE="create_and_fill_mysql_db.sql"
+loadData
+
+# Trim the spaces from the state_code field of the ref_place_details table
+TABLE=ref_place_details
+trimStateCode
+
+# Count the rows from the ref_place_details table
+TABLE=ref_place_details
+countRows
+
+# Count the rows from the ref_place_names table
+TABLE=ref_place_names
+countRows
Copied: trunk/opentrep/db/maintenance/drop_tables_from_mysql_db.sh (from rev 134, trunk/opentrep/refdata/mysql/drop_tables_from_mysql_db.sh)
===================================================================
--- trunk/opentrep/db/maintenance/drop_tables_from_mysql_db.sh (rev 0)
+++ trunk/opentrep/db/maintenance/drop_tables_from_mysql_db.sh 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,54 @@
+#!/bin/sh
+#
+# One parameter is required for this script:
+# - the username
+#
+# Two parameters are optional:
+# - the host server of the database
+# - the port of the database
+#
+
+if [ "$1" = "" -o "$1" = "-h" -o "$1" = "--help" ];
+then
+ echo "Usage: $0 <Database Username> [<Database Server Hostname> [<Database Server Port>]]"
+ echo ""
+ exit -1
+fi
+
+##
+# Database Server Hostname
+DB_HOST="localhost"
+if [ "$2" != "" ];
+then
+ DB_HOST="$2"
+fi
+
+# Database Server Port
+DB_PORT="3306"
+if [ "$3" != "" ];
+then
+ DB_PORT="$3"
+fi
+
+# Database User
+DB_USER="$1"
+
+# Database Password
+DB_PASSWD="${DB_USER}"
+
+# Database Name
+DB_NAME="opentrep"
+
+# Drop a table
+function dropTable() {
+ echo "The ${TABLE} table will be dropped:"
+ mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} -e "drop table ${DB_NAME}.${TABLE}"
+}
+
+# Table: drop the ref_place_details table
+TABLE=ref_place_details
+dropTable
+
+# Table: drop the ref_place_names table
+TABLE=ref_place_names
+dropTable
Deleted: trunk/opentrep/db/maintenance/ref_city.sql
===================================================================
--- trunk/opentrep/refdata/data_structure/ref_city.sql 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/db/maintenance/ref_city.sql 2009-07-18 00:50:00 UTC (rev 135)
@@ -1,46 +0,0 @@
---
--- Place details
--- Does not depend on language
---
-create table if not exists ref_place_details (
- code char(3) collate utf8_unicode_ci not null,
- city_code char(3) collate utf8_unicode_ci,
- xapian_docid integer,
- is_airport char(1) collate utf8_unicode_ci not null,
- is_city char(1) collate utf8_unicode_ci not null,
- is_main char(1) collate utf8_unicode_ci not null default 'N',
- is_commercial char(1) collate utf8_unicode_ci not null,
- state_code varchar(5) collate utf8_unicode_ci,
- country_code char(2) collate utf8_unicode_ci not null,
- region_code varchar(5) collate utf8_unicode_ci not null,
- continent_code varchar(4) collate utf8_unicode_ci not null,
- time_zone_grp varchar(5) collate utf8_unicode_ci not null,
- longitude float(20),
- latitude float(20),
- primary key (code),
- key `geographical codes`(city_code, continent_code, country_code, region_code, time_zone_grp)
-) engine=myisam default charset=utf8 collate=utf8_unicode_ci;
-
---
--- Place names
--- Depends on language
---
-create table if not exists ref_place_names (
- language_code char(2) collate utf8_unicode_ci not null,
- code char(3) collate utf8_unicode_ci not null,
- classical_name varchar(30) collate utf8_unicode_ci not null,
- classical_name2 varchar(50) collate utf8_unicode_ci not null,
- extended_name varchar(100) collate utf8_unicode_ci not null,
- alternate_name1 varchar(60) collate utf8_unicode_ci,
- alternate_name2 varchar(60) collate utf8_unicode_ci,
- alternate_name3 varchar(60) collate utf8_unicode_ci,
- alternate_name4 varchar(60) collate utf8_unicode_ci,
- alternate_name5 varchar(60) collate utf8_unicode_ci,
- alternate_name6 varchar(60) collate utf8_unicode_ci,
- alternate_name7 varchar(60) collate utf8_unicode_ci,
- alternate_name8 varchar(60) collate utf8_unicode_ci,
- alternate_name9 varchar(60) collate utf8_unicode_ci,
- alternate_name10 varchar(60) collate utf8_unicode_ci,
- primary key (language_code, code)
-) engine=myisam default charset=utf8 collate=utf8_unicode_ci;
-
Deleted: trunk/opentrep/db/maintenance/ref_db.sql
===================================================================
--- trunk/opentrep/refdata/data_structure/ref_db.sql 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/db/maintenance/ref_db.sql 2009-07-18 00:50:00 UTC (rev 135)
@@ -1,6 +0,0 @@
-
---
--- Create the opentrep database
---
-create database if not exists opentrep
-default character set utf8 collate utf8_unicode_ci;
Property changes on: trunk/opentrep/db/maintenance/tables
___________________________________________________________________
Added: svn:ignore
+ Makefile
Makefile.in
Added: trunk/opentrep/db/maintenance/tables/Makefile.am
===================================================================
--- trunk/opentrep/db/maintenance/tables/Makefile.am (rev 0)
+++ trunk/opentrep/db/maintenance/tables/Makefile.am 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,26 @@
+## db/maintenance/tables sub-directory
+include $(top_srcdir)/Makefile.common
+include $(top_srcdir)/db/maintenance/tables/sources.mk
+
+datadir = @datadir@
+pkgdatadir = $(datadir)/@PACKAGE@
+dbscriptsdir = $(pkgdatadir)/db/mysql/create_tables
+
+MAINTAINERCLEANFILES = Makefile.in Makefile
+
+noinst_DATA = $(dbscript_mysql_sources)
+EXTRA_DIST = $(noinst_DATA)
+
+
+# Targets
+install-data-local:
+ $(mkinstalldirs) $(DESTDIR)$(dbscriptsdir); \
+ for f in $(dbscript_mysql_sources); do \
+ $(INSTALL_DATA) $$f $(DESTDIR)$(dbscriptsdir); \
+ done
+
+uninstall-local:
+ rm -rf $(DESTDIR)$(dbscriptsdir)
+
+clean-local:
+ rm -rf *.log *.tag
Copied: trunk/opentrep/db/maintenance/tables/create_and_fill_mysql_db.sql (from rev 134, trunk/opentrep/refdata/mysql/create_and_fill_mysql_db.sql)
===================================================================
--- trunk/opentrep/db/maintenance/tables/create_and_fill_mysql_db.sql (rev 0)
+++ trunk/opentrep/db/maintenance/tables/create_and_fill_mysql_db.sql 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,16 @@
+
+--
+-- Load the Airport and City geographical details into the MySQL table
+--
+load data local infile '../data/ref_place_details.csv' ignore
+into table ref_place_details
+fields terminated by ',' enclosed by '' escaped by '\\'
+ignore 1 lines;
+
+--
+-- Load the Airport and City names into the MySQL table
+--
+load data local infile '../data/ref_place_names.csv' ignore
+into table ref_place_names
+fields terminated by ',' enclosed by '' escaped by '\\'
+ignore 1 lines;
Copied: trunk/opentrep/db/maintenance/tables/ref_city.sql (from rev 134, trunk/opentrep/refdata/data_structure/ref_city.sql)
===================================================================
--- trunk/opentrep/db/maintenance/tables/ref_city.sql (rev 0)
+++ trunk/opentrep/db/maintenance/tables/ref_city.sql 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,46 @@
+--
+-- Place details
+-- Does not depend on language
+--
+create table if not exists ref_place_details (
+ code char(3) collate utf8_unicode_ci not null,
+ city_code char(3) collate utf8_unicode_ci,
+ xapian_docid integer,
+ is_airport char(1) collate utf8_unicode_ci not null,
+ is_city char(1) collate utf8_unicode_ci not null,
+ is_main char(1) collate utf8_unicode_ci not null default 'N',
+ is_commercial char(1) collate utf8_unicode_ci not null,
+ state_code varchar(5) collate utf8_unicode_ci,
+ country_code char(2) collate utf8_unicode_ci not null,
+ region_code varchar(5) collate utf8_unicode_ci not null,
+ continent_code varchar(4) collate utf8_unicode_ci not null,
+ time_zone_grp varchar(5) collate utf8_unicode_ci not null,
+ longitude float(20),
+ latitude float(20),
+ primary key (code),
+ key `geographical codes`(city_code, continent_code, country_code, region_code, time_zone_grp)
+) engine=myisam default charset=utf8 collate=utf8_unicode_ci;
+
+--
+-- Place names
+-- Depends on language
+--
+create table if not exists ref_place_names (
+ language_code char(2) collate utf8_unicode_ci not null,
+ code char(3) collate utf8_unicode_ci not null,
+ classical_name varchar(30) collate utf8_unicode_ci not null,
+ classical_name2 varchar(50) collate utf8_unicode_ci not null,
+ extended_name varchar(100) collate utf8_unicode_ci not null,
+ alternate_name1 varchar(60) collate utf8_unicode_ci,
+ alternate_name2 varchar(60) collate utf8_unicode_ci,
+ alternate_name3 varchar(60) collate utf8_unicode_ci,
+ alternate_name4 varchar(60) collate utf8_unicode_ci,
+ alternate_name5 varchar(60) collate utf8_unicode_ci,
+ alternate_name6 varchar(60) collate utf8_unicode_ci,
+ alternate_name7 varchar(60) collate utf8_unicode_ci,
+ alternate_name8 varchar(60) collate utf8_unicode_ci,
+ alternate_name9 varchar(60) collate utf8_unicode_ci,
+ alternate_name10 varchar(60) collate utf8_unicode_ci,
+ primary key (language_code, code)
+) engine=myisam default charset=utf8 collate=utf8_unicode_ci;
+
Copied: trunk/opentrep/db/maintenance/tables/ref_db.sql (from rev 134, trunk/opentrep/refdata/data_structure/ref_db.sql)
===================================================================
--- trunk/opentrep/db/maintenance/tables/ref_db.sql (rev 0)
+++ trunk/opentrep/db/maintenance/tables/ref_db.sql 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,6 @@
+
+--
+-- Create the opentrep database
+--
+create database if not exists opentrep
+default character set utf8 collate utf8_unicode_ci;
Added: trunk/opentrep/db/maintenance/tables/sources.mk
===================================================================
--- trunk/opentrep/db/maintenance/tables/sources.mk (rev 0)
+++ trunk/opentrep/db/maintenance/tables/sources.mk 2009-07-18 00:50:00 UTC (rev 135)
@@ -0,0 +1,4 @@
+dbscript_mysql_sources = \
+ $(top_srcdir)/db/maintenance/tables/create_and_fill_mysql_db.sql \
+ $(top_srcdir)/db/maintenance/tables/ref_db.sql \
+ $(top_srcdir)/db/maintenance/tables/ref_city.sql
Deleted: trunk/opentrep/db/mysql/create_and_fill_mysql_db.sh
===================================================================
--- trunk/opentrep/refdata/mysql/create_and_fill_mysql_db.sh 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/db/mysql/create_and_fill_mysql_db.sh 2009-07-18 00:50:00 UTC (rev 135)
@@ -1,98 +0,0 @@
-#!/bin/sh
-#
-# One parameter is required for this script:
-# - the username
-#
-# Two parameters are optional:
-# - the host server of the database
-# - the port of the database
-#
-
-if [ "$1" = "" -o "$1" = "-h" -o "$1" = "--help" ];
-then
- echo "Usage: $0 <Database Username> [<Database Server Hostname> [<Database Server Port>]]"
- echo ""
- exit -1
-fi
-
-##
-# Database Server Hostname
-DB_HOST="localhost"
-if [ "$2" != "" ];
-then
- DB_HOST="$2"
-fi
-
-# Database Server Port
-DB_PORT="3306"
-if [ "$3" != "" ];
-then
- DB_PORT="$3"
-fi
-
-# Database User
-DB_USER="$1"
-
-# Database Password
-DB_PASSWD="${DB_USER}"
-
-# Database Name
-DB_NAME="opentrep"
-
-# Create the database
-function createDatabase() {
- echo "The '${DB_NAME}' database will be created:"
- mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} mysql < ${SQL_FILE}
-}
-
-# Scan a SQL script for the names of (database) tables
-function createTable() {
- echo "The ref_place_details and ref_place_names tables will be created:"
- mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} < ${SQL_FILE}
-}
-
-#
-function loadData() {
- echo "The ref_place_details and ref_place_names tables will be filled from ../data/*.csv files:"
- mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} < ${SQL_LOADER_FILE}
- echo "Done"
-}
-
-#
-function trimStateCode() {
- echo "Triming the spaces from the state_code field of the ${TABLE} table:"
- mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} -e "update ${TABLE} set city_code=NULL where city_code='';"
- mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} -e "update ${TABLE} set state_code=NULL where state_code like '%null%';"
- mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} -e "update ${TABLE} set state_code=NULL where length(state_code)=2;"
- mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} -e "update ${TABLE} set state_code=substring(state_code,2,2) where length(state_code)=4;"
-}
-
-#
-function countRows() {
- echo "Counting the rows from the ${TABLE} table:"
- mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} -e "select count(*) from ${TABLE};"
-}
-
-# Database
-SQL_FILE="../data_structure/ref_db.sql"
-createDatabase
-
-# Table: Airport and City
-SQL_FILE="../data_structure/ref_city.sql"
-createTable
-
-# Load data into the table
-SQL_LOADER_FILE="create_and_fill_mysql_db.sql"
-loadData
-
-# Trim the spaces from the state_code field of the ref_place_details table
-TABLE=ref_place_details
-trimStateCode
-
-# Count the rows from the ref_place_details table
-TABLE=ref_place_details
-countRows
-
-# Count the rows from the ref_place_names table
-TABLE=ref_place_names
-countRows
Deleted: trunk/opentrep/db/mysql/create_and_fill_mysql_db.sql
===================================================================
--- trunk/opentrep/refdata/mysql/create_and_fill_mysql_db.sql 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/db/mysql/create_and_fill_mysql_db.sql 2009-07-18 00:50:00 UTC (rev 135)
@@ -1,16 +0,0 @@
-
---
--- Load the Airport and City geographical details into the MySQL table
---
-load data local infile '../data/ref_place_details.csv' ignore
-into table ref_place_details
-fields terminated by ',' enclosed by '' escaped by '\\'
-ignore 1 lines;
-
---
--- Load the Airport and City names into the MySQL table
---
-load data local infile '../data/ref_place_names.csv' ignore
-into table ref_place_names
-fields terminated by ',' enclosed by '' escaped by '\\'
-ignore 1 lines;
Deleted: trunk/opentrep/db/mysql/create_opentrep_user.sh
===================================================================
--- trunk/opentrep/refdata/mysql/create_opentrep_user.sh 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/db/mysql/create_opentrep_user.sh 2009-07-18 00:50:00 UTC (rev 135)
@@ -1,52 +0,0 @@
-#!/bin/sh
-#
-# Two parameters are required for this script:
-# - the administrator username
-# - the administrator password
-#
-# Two parameters are optional:
-# - the host server of the database
-# - the port of the database
-#
-
-if [ "$1" = "" -o "$2" = "" -o "$1" = "-h" -o "$1" = "--help" ];
-then
- echo "Usage: $0 <Admin Username> <Admin password> [<Database Server Hostname> [<Database Server Port>]]"
- echo ""
- exit -1
-fi
-
-##
-# Database Server Hostname
-DB_HOST="localhost"
-if [ "$3" != "" ];
-then
- DB_HOST="$3"
-fi
-
-# Database Server Port
-DB_PORT="3306"
-if [ "$4" != "" ];
-then
- DB_PORT="$4"
-fi
-
-# Database User
-DB_USER="$1"
-
-# Database Password
-DB_PASSWD="$2"
-
-# Database Name
-DB_NAME="mysql"
-
-function createOpenTrepUser() {
- echo "Creating the opentrep user within the database:"
- mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} < ${SQL_FILE}
- mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} -e "flush privileges"
-}
-
-# Creating the opentrep user
-SQL_FILE="create_opentrep_user.sql"
-createOpenTrepUser
-
Deleted: trunk/opentrep/db/mysql/create_opentrep_user.sql
===================================================================
--- trunk/opentrep/refdata/mysql/create_opentrep_user.sql 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/db/mysql/create_opentrep_user.sql 2009-07-18 00:50:00 UTC (rev 135)
@@ -1,7 +0,0 @@
-
-insert into `user` (`Host`, `User`, `Password`, `Select_priv`, `Insert_priv`, `Update_priv`, `Delete_priv`, `Create_priv`, `Drop_priv`, `Reload_priv`, `Shutdown_priv`, `Process_priv`, `File_priv`, `Grant_priv`, `References_priv`, `Index_priv`, `Alter_priv`, `Show_db_priv`, `Super_priv`, `Create_tmp_table_priv`, `Lock_tables_priv`, `Execute_priv`, `Repl_slave_priv`, `Repl_client_priv`, `Create_view_priv`, `Show_view_priv`, `Create_routine_priv`, `Alter_routine_priv`, `Create_user_priv`, `ssl_type`, `ssl_cipher`, `x509_issuer`, `x509_subject`, `max_questions`, `max_updates`, `max_connections`, `max_user_connections`) values
-('%', 'opentrep', '*C21B5F0DB6BBABAA20B5496E75D652982A6AC65C', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'N', 'N', 'N', 'Y', 'N', 'N', 'Y', 'Y', 'N', 'N', 'Y', 'N', 'Y', 'N', 'N', 'Y', 'Y', 'Y', 'Y', 'N', '', '', '', '', 0, 0, 0, 0),
-('localhost', 'opentrep', '*C21B5F0DB6BBABAA20B5496E75D652982A6AC65C', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'N', 'N', 'N', 'Y', 'N', 'N', 'Y', 'Y', 'N', 'N', 'Y', 'N', 'Y', 'N', 'N', 'Y', 'Y', 'Y', 'Y', 'N', '', '', '', '', 0, 0, 0, 0);
-
-flush privileges;
-
Deleted: trunk/opentrep/db/mysql/drop_tables_from_mysql_db.sh
===================================================================
--- trunk/opentrep/refdata/mysql/drop_tables_from_mysql_db.sh 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/db/mysql/drop_tables_from_mysql_db.sh 2009-07-18 00:50:00 UTC (rev 135)
@@ -1,54 +0,0 @@
-#!/bin/sh
-#
-# One parameter is required for this script:
-# - the username
-#
-# Two parameters are optional:
-# - the host server of the database
-# - the port of the database
-#
-
-if [ "$1" = "" -o "$1" = "-h" -o "$1" = "--help" ];
-then
- echo "Usage: $0 <Database Username> [<Database Server Hostname> [<Database Server Port>]]"
- echo ""
- exit -1
-fi
-
-##
-# Database Server Hostname
-DB_HOST="localhost"
-if [ "$2" != "" ];
-then
- DB_HOST="$2"
-fi
-
-# Database Server Port
-DB_PORT="3306"
-if [ "$3" != "" ];
-then
- DB_PORT="$3"
-fi
-
-# Database User
-DB_USER="$1"
-
-# Database Password
-DB_PASSWD="${DB_USER}"
-
-# Database Name
-DB_NAME="opentrep"
-
-# Drop a table
-function dropTable() {
- echo "The ${TABLE} table will be dropped:"
- mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} -e "drop table ${DB_NAME}.${TABLE}"
-}
-
-# Table: drop the ref_place_details table
-TABLE=ref_place_details
-dropTable
-
-# Table: drop the ref_place_names table
-TABLE=ref_place_names
-dropTable
Modified: trunk/opentrep/opentrep/OPENTREP_Types.hpp
===================================================================
--- trunk/opentrep/opentrep/OPENTREP_Types.hpp 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/opentrep/OPENTREP_Types.hpp 2009-07-18 00:50:00 UTC (rev 135)
@@ -56,8 +56,11 @@
/** Xapian document ID. */
typedef int XapianDocID_T;
- /** Travel Search Query. */
+ /** Travel search query. */
typedef std::string TravelQuery_T;
+
+ /** Number of matching documents. */
+ typedef unsigned short NbOfMatches_T;
}
#endif // __OPENTREP_OPENTREP_TYPES_HPP
Property changes on: trunk/opentrep/opentrep/batches
___________________________________________________________________
Modified: svn:ignore
- .deps
.libs
Makefile
Makefile.in
opentrep_indexer
opentrep_searcher
+ .deps
.libs
Makefile
Makefile.in
opentrep_indexer*
opentrep_searcher*
Modified: trunk/opentrep/opentrep/batches/indexer.cpp
===================================================================
--- trunk/opentrep/opentrep/batches/indexer.cpp 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/opentrep/batches/indexer.cpp 2009-07-18 00:50:00 UTC (rev 135)
@@ -12,43 +12,56 @@
#include <boost/program_options.hpp>
// OPENTREP
#include <opentrep/OPENTREP_Service.hpp>
+#include <opentrep/config/opentrep-paths.hpp>
+
+// //////// Type definitions ///////
+typedef std::vector<std::string> WordList_T;
+
+
+// //////// Constants //////
+/** Default name and location for the log file. */
+const std::string K_OPENTREP_DEFAULT_LOG_FILENAME ("opentrep_indexer.log");
+
+/** Default name and location for the Xapian database. */
+const std::string K_OPENTREP_DEFAULT_DATABSE_FILEPATH("/tmp/opentrep/traveldb");
+
+
// ///////// Parsing of Options & Configuration /////////
-// A helper function to simplify the main part.
-template<class T> std::ostream& operator<< (std::ostream& os,
- const std::vector<T>& v) {
- std::copy (v.begin(), v.end(), std::ostream_iterator<T> (std::cout, " "));
- return os;
-}
+/** Early return status (so that it can be differentiated from an error). */
+const int K_OPENTREP_EARLY_RETURN_STATUS = 99;
-int readConfiguration (int argc, char* argv[]) {
- int opt;
-
- // Declare a group of options that will be
- // allowed only on command line
- boost::program_options::options_description generic("Generic options");
+/** Read and parse the command line options. */
+int readConfiguration (int argc, char* argv[],
+ std::string& ioDatabaseFilepath,
+ std::string& ioLogFilename) {
+
+ // Declare a group of options that will be allowed only on command line
+ boost::program_options::options_description generic ("Generic options");
generic.add_options()
+ ("prefix", "print installation prefix")
("version,v", "print version string")
("help,h", "produce help message");
- // Declare a group of options that will be allowed both on command line and in
- // config file
- boost::program_options::options_description config("Configuration");
+ // Declare a group of options that will be allowed both on command
+ // line and in config file
+ boost::program_options::options_description config ("Configuration");
config.add_options()
- ("optimization",
- boost::program_options::value<int>(&opt)->default_value(10),
- "optimization level")
- ("include-path,I",
- boost::program_options::value< std::vector<std::string> >()->composing(),
- "include path");
+ ("database,d",
+ boost::program_options::value< std::string >(&ioDatabaseFilepath)->default_value(K_OPENTREP_DEFAULT_DATABSE_FILEPATH),
+ "Xapian database filepath (e.g., /tmp/opentrep/traveldb)")
+ ("log,l",
+ boost::program_options::value< std::string >(&ioLogFilename)->default_value(K_OPENTREP_DEFAULT_LOG_FILENAME),
+ "Filepath for the logs")
+ ;
// Hidden options, will be allowed both on command line and
// in config file, but will not be shown to the user.
- boost::program_options::options_description hidden("Hidden options");
+ boost::program_options::options_description hidden ("Hidden options");
hidden.add_options()
- ("input-file",
+ ("copyright",
boost::program_options::value< std::vector<std::string> >(),
- "input file");
+ "Show the copyright (license)");
boost::program_options::options_description cmdline_options;
cmdline_options.add(generic).add(config).add(hidden);
@@ -56,46 +69,48 @@
boost::program_options::options_description config_file_options;
config_file_options.add(config).add(hidden);
- boost::program_options::options_description visible("Allowed options");
+ boost::program_options::options_description visible ("Allowed options");
visible.add(generic).add(config);
boost::program_options::positional_options_description p;
- p.add("input-file", -1);
+ p.add ("copyright", -1);
boost::program_options::variables_map vm;
boost::program_options::
- store (boost::program_options::command_line_parser(argc, argv).
- options (cmdline_options).positional(p).run(), vm);
+ store (boost::program_options::command_line_parser (argc, argv).
+ options (cmdline_options).positional(p).run(), vm);
- std::ifstream ifs ("request_parser.cfg");
+ std::ifstream ifs ("opentrep_indexer.cfg");
boost::program_options::store (parse_config_file (ifs, config_file_options),
vm);
boost::program_options::notify (vm);
if (vm.count ("help")) {
std::cout << visible << std::endl;
- return 0;
+ return K_OPENTREP_EARLY_RETURN_STATUS;
}
if (vm.count ("version")) {
- std::cout << "Open Travel Request Parser, version 1.0" << std::endl;
- return 0;
+ std::cout << PACKAGE_NAME << ", version " << PACKAGE_VERSION << std::endl;
+ return K_OPENTREP_EARLY_RETURN_STATUS;
}
- if (vm.count ("include-path")) {
- std::cout << "Include paths are: "
- << vm["include-path"].as< std::vector<std::string> >()
- << std::endl;
+ if (vm.count ("prefix")) {
+ std::cout << "Installation prefix: " << PREFIXDIR << std::endl;
+ return K_OPENTREP_EARLY_RETURN_STATUS;
}
- if (vm.count ("input-file")) {
- std::cout << "Input files are: "
- << vm["input-file"].as< std::vector<std::string> >()
+ if (vm.count ("database")) {
+ ioDatabaseFilepath = vm["database"].as< std::string >();
+ std::cout << "Xapian database filepath is: " << ioDatabaseFilepath
<< std::endl;
}
- std::cout << "Optimization level is " << opt << std::endl;
-
+ if (vm.count ("log")) {
+ ioLogFilename = vm["log"].as< std::string >();
+ std::cout << "Log filename is: " << ioLogFilename << std::endl;
+ }
+
return 0;
}
@@ -105,19 +120,17 @@
try {
// Output log File
- std::string lLogFilename ("indexer.log");
+ std::string lLogFilename;
// Xapian database name (directory of the index)
- OPENTREP::TravelDatabaseName_T lXapianDatabaseName ("traveldb");
+ OPENTREP::TravelDatabaseName_T lXapianDatabaseName;
- if (argc >= 1 && argv[1] != NULL) {
- std::istringstream istr (argv[1]);
- istr >> lLogFilename;
- }
+ // Call the command-line option parser
+ const int lOptionParserStatus =
+ readConfiguration (argc, argv, lXapianDatabaseName, lLogFilename);
- if (argc >= 2 && argv[2] != NULL) {
- std::istringstream istr (argv[2]);
- istr >> lXapianDatabaseName;
+ if (lOptionParserStatus == K_OPENTREP_EARLY_RETURN_STATUS) {
+ return 0;
}
// Set the log parameters
@@ -135,7 +148,6 @@
// Close the Log outputFile
logOutputFile.close();
-
} catch (const OPENTREP::RootException& otexp) {
std::cerr << "Standard exception: " << otexp.what() << std::endl;
Modified: trunk/opentrep/opentrep/batches/searcher.cpp
===================================================================
--- trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-18 00:50:00 UTC (rev 135)
@@ -15,15 +15,17 @@
#include <opentrep/OPENTREP_Service.hpp>
#include <opentrep/config/opentrep-paths.hpp>
+
// //////// Type definitions ///////
typedef std::vector<std::string> WordList_T;
+
// //////// Constants //////
/** Default name and location for the log file. */
const std::string K_OPENTREP_DEFAULT_LOG_FILENAME ("opentrep_searcher.log");
/** Default name and location for the Xapian database. */
-const std::string K_OPENTREP_DEFAULT_DATABSE_FILEPATH ("/tmp/opentrep/traveldb");
+const std::string K_OPENTREP_DEFAULT_DATABSE_FILEPATH("/tmp/opentrep/traveldb");
/** Default travel query string, to be seached against the Xapian database. */
const std::string K_OPENTREP_DEFAULT_QUERY_STRING ("sna francicso rio de janero lso anglese reykyavki");
@@ -31,6 +33,7 @@
/** Default error distance for spelling corrections. */
const unsigned short K_OPENTREP_DEFAULT_SPELLING_ERROR_DISTANCE = 3;
+
// //////////////////////////////////////////////////////////////////////
void tokeniseStringIntoWordList (const std::string& iPhrase,
WordList_T& ioWordList) {
Modified: trunk/opentrep/opentrep/bom/DocumentList.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/DocumentList.hpp 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/opentrep/bom/DocumentList.hpp 2009-07-18 00:50:00 UTC (rev 135)
@@ -5,14 +5,23 @@
// Import section
// //////////////////////////////////////////////////////////////////////
// STL
-#include <map>
+#include <list>
+// OpenTREP
+#include <opentrep/OPENTREP_Types.hpp>
// Xapian
#include <xapian.h>
namespace OPENTREP {
- /** List of Xapian documents. */
- typedef std::multimap<Xapian::percent, Xapian::Document> DocumentList_T;
+ /** Xapian document and its associated matching percentage. */
+ typedef std::pair<Xapian::percent, Xapian::Document> MatchingDocument_T;
+ /** A matching Xapian document, along with the query string which it
+ matches. */
+ typedef std::pair<TravelQuery_T, MatchingDocument_T> QueryAndDocument_T;
+
+ /** List of matching Xapian documents. */
+ typedef std::list<QueryAndDocument_T> DocumentList_T;
+
}
#endif // __OPENTREP_BOM_DOCUMENTLIST_HPP
Modified: trunk/opentrep/opentrep/bom/Result.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/Result.cpp 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/opentrep/bom/Result.cpp 2009-07-18 00:50:00 UTC (rev 135)
@@ -5,7 +5,7 @@
#include <cassert>
#include <string>
#include <sstream>
-// OPENTREP
+// OpenTREP
#include <opentrep/bom/StringMatcher.hpp>
#include <opentrep/bom/Result.hpp>
#include <opentrep/service/Logger.hpp>
@@ -24,7 +24,6 @@
// //////////////////////////////////////////////////////////////////////
void Result::init () {
- _documentList.clear();
}
// //////////////////////////////////////////////////////////////////////
@@ -44,14 +43,11 @@
std::ostringstream oStr;
oStr << describeShortKey() << std::endl;
- for (DocumentList_T::const_iterator itDoc = _documentList.begin();
- itDoc != _documentList.end(); ++itDoc) {
- const Xapian::percent& lPercent = itDoc->first;
- const Xapian::Document& lDocument = itDoc->second;
- const Xapian::docid& lDocID = lDocument.get_docid();
- oStr << "Document ID " << lDocID << "\t" << lPercent
- << "% [" << lDocument.get_data() << "]" << std::endl;
- }
+ const Xapian::percent& lPercentage = _matchingDocument.first;
+ const Xapian::Document& lDocument = _matchingDocument.second;
+ const Xapian::docid& lDocID = lDocument.get_docid();
+ oStr << "Document ID " << lDocID << "\t" << lPercentage
+ << "% [" << lDocument.get_data() << "]" << std::endl;
return oStr.str();
}
@@ -65,67 +61,4 @@
void Result::fromStream (std::istream& ioIn) {
}
- // //////////////////////////////////////////////////////////////////////
- const Xapian::Document& Result::getBestMatchingDocument() const {
- /**
- Retrieve the best matching document. As the document list (STL map)
- is sorted by ascending order of the matching percentage, the best
- matching one is located at the end (back) of the list (STL map).
- */
- DocumentList_T::const_reverse_iterator itDocument = _documentList.rbegin();
- return itDocument->second;
- }
-
- // //////////////////////////////////////////////////////////////////////
- const Xapian::percent& Result::getBestMatchingPercentage() const {
- /**
- Retrieve the best matching document. As the document list (STL map)
- is sorted by ascending order of the matching percentage, the best
- matching one is located at the end (back) of the list (STL map).
- */
- DocumentList_T::const_reverse_iterator itDocument = _documentList.rbegin();
- return itDocument->first;
- }
-
- // //////////////////////////////////////////////////////////////////////
- void Result::searchString () {
-
- // Catch any Xapian::Error exceptions thrown
- try {
-
- bool shouldStop = false;
- while (shouldStop == false) {
- // DEBUG
- /*
- OPENTREP_LOG_DEBUG (std::endl << "--------------------------------"
- << std::endl << "Current query string: `" << ioQueryString << "'");
- */
-
- // Retrieve the list of documents matching the query string
- Xapian::MSet lMatchingSet;
- StringMatcher::searchString (lMatchingSet, _queryString, _database);
-
- // Create the corresponding list of documents
- StringMatcher::createDocumentListFromMSet (lMatchingSet, _documentList);
-
- // Stop if a result is found.
- if (_documentList.empty() == false) {
- shouldStop = true;
- break;
- }
-
- // Remove a word from the query string
- StringMatcher::removeOneWord (_queryString);
-
- // Stop when the resulting string gets empty.
- if (_queryString.empty() == true) {
- shouldStop = true;
- }
- }
-
- } catch (const Xapian::Error& error) {
- OPENTREP_LOG_ERROR ("Exception: " << error.get_msg());
- }
- }
-
}
Modified: trunk/opentrep/opentrep/bom/Result.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/Result.hpp 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/opentrep/bom/Result.hpp 2009-07-18 00:50:00 UTC (rev 135)
@@ -25,31 +25,43 @@
return _queryString;
}
- /** Get the list of Xapian document objects. */
- const DocumentList_T& getDocumentList() const {
- return _documentList;
+ /** Get the Matching Xapian document object, along with its
+ corresponding matching percentage. */
+ const MatchingDocument_T& getMatchingDocument() const {
+ return _matchingDocument;
}
- /** Retrieve the best matching Xapian document object. */
- const Xapian::Document& getBestMatchingDocument() const;
+ /** Retrieve the percentage corresponding to the matching Xapian
+ document object. */
+ const Xapian::percent& getPercentage() const {
+ return _matchingDocument.first;
+ }
+
+ /** Retrieve the matching Xapian document object. */
+ const Xapian::Document& getDocument() const {
+ return _matchingDocument.second;
+ }
- /** Retrieve the percentage corresponding to the best matching
- Xapian document object. */
- const Xapian::percent& getBestMatchingPercentage() const;
-
// ////////////// Setters /////////////
/** Set the query string. */
void setQueryString (const TravelQuery_T& iQueryString) {
_queryString = iQueryString;
}
-
- public:
- // /////////// Business methods /////////
- /** Retrieve the list of documents matching the query string. */
- void searchString ();
+ /** Set the matching Xapian document object and its corresponding
+ matching percentage. */
+ void setMatchingDocument (const MatchingDocument_T& iMatchingDocument) {
+ _matchingDocument = iMatchingDocument;
+ }
+ /** Set the matching Xapian document object and its corresponding
+ matching percentage. */
+ void setQueryAndDocument (const QueryAndDocument_T& iQueryAndDocument) {
+ _queryString = iQueryAndDocument.first;
+ _matchingDocument = iQueryAndDocument.second;
+ }
+
public:
// /////////// Display support methods /////////
@@ -98,8 +110,9 @@
/** Xapian database. */
const Xapian::Database& _database;
- /** List of Xapian document objects. */
- DocumentList_T _documentList;
+ /** Matching Xapian document object, along with its corresponding
+ matching percentage. */
+ MatchingDocument_T _matchingDocument;
};
}
Modified: trunk/opentrep/opentrep/bom/ResultHolder.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/ResultHolder.cpp 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/opentrep/bom/ResultHolder.cpp 2009-07-18 00:50:00 UTC (rev 135)
@@ -10,9 +10,6 @@
#include <opentrep/bom/StringMatcher.hpp>
#include <opentrep/bom/Result.hpp>
#include <opentrep/bom/ResultHolder.hpp>
-// TODO: move that out of the BOM layer
-#include <opentrep/factory/FacResultHolder.hpp>
-#include <opentrep/factory/FacResult.hpp>
#include <opentrep/service/Logger.hpp>
namespace OPENTREP {
@@ -71,11 +68,59 @@
}
// //////////////////////////////////////////////////////////////////////
- void ResultHolder::searchString () {
+ bool ResultHolder::searchString (TravelQuery_T& ioPartialQueryString,
+ MatchingDocument_T& ioMatchingDocument) {
+ bool oFoundDocument = false;
// Catch any Xapian::Error exceptions thrown
try {
+ bool shouldStop = false;
+ while (shouldStop == false) {
+ // DEBUG
+ /*
+ OPENTREP_LOG_DEBUG (std::endl << "--------------------------------"
+ << std::endl << "Current query string: `" << ioPartialQueryString
+ << "'");
+ */
+
+ // Retrieve the list of documents matching the query string
+ Xapian::MSet lMatchingSet;
+ StringMatcher::searchString (lMatchingSet, ioPartialQueryString,
+ _database);
+
+ // Create the corresponding list of documents
+ oFoundDocument = StringMatcher::
+ extractBestMatchingDocumentFromMSet(lMatchingSet, ioMatchingDocument);
+
+ // Stop if a result is found.
+ if (oFoundDocument == true) {
+ shouldStop = true;
+ break;
+ }
+
+ // Remove a word from the query string
+ StringMatcher::removeOneWord (ioPartialQueryString);
+
+ // Stop when the resulting string gets empty.
+ if (ioPartialQueryString.empty() == true) {
+ shouldStop = true;
+ }
+ }
+
+ } catch (const Xapian::Error& error) {
+ OPENTREP_LOG_ERROR ("Exception: " << error.get_msg());
+ }
+
+ return oFoundDocument;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void ResultHolder::searchString (DocumentList_T& ioDocumentList) {
+
+ // Catch any Xapian::Error exceptions thrown
+ try {
+
std::string lRemainingQueryString (_queryString);
bool shouldStop = false;
while (shouldStop == false) {
@@ -91,21 +136,23 @@
again no result, until either a result is found or the
resulting string gets empty.
*/
- DocumentList_T lDocumentList;
- // TODO: move that out of the BOM layer
- Result& lResult = FacResult::instance().create (_database);
-
std::string lQueryString (lRemainingQueryString);
- //
- lResult.setQueryString (lQueryString);
- lResult.searchString ();
+ /**
+ Main algorithm, altering the query string (suppressing the
+ furthest right words, so that the remaining left part be matched
+ against the Xapian database).
+ */
+ MatchingDocument_T lMatchingDocument;
+ const bool hasFoundDocument = searchString (lQueryString,
+ lMatchingDocument);
+
+ if (hasFoundDocument == true) {
+ const QueryAndDocument_T lQueryAndDocument (lQueryString,
+ lMatchingDocument);
+ ioDocumentList.push_back (lQueryAndDocument);
+ }
- // Add the Result object (holding the list of matching
- // documents) to the dedicated list.
- // TODO: move that out of the BOM layer
- FacResultHolder::initLinkWithResult (*this, lResult);
-
/**
Remove, from the lRemainingQueryString string, the part which
has been already successfully parsed.
@@ -117,7 +164,6 @@
'rio de janeiro'. So, the already parsed part, namely
'sna francisco', must be subtracted from the initial query string.
*/
- lQueryString = lResult.getQueryString();
StringMatcher::subtractParsedToRemaining (lQueryString,
lRemainingQueryString);
Modified: trunk/opentrep/opentrep/bom/ResultHolder.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/ResultHolder.hpp 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/opentrep/bom/ResultHolder.hpp 2009-07-18 00:50:00 UTC (rev 135)
@@ -7,6 +7,7 @@
// OpenTREP
#include <opentrep/OPENTREP_Types.hpp>
#include <opentrep/bom/BomAbstract.hpp>
+#include <opentrep/bom/DocumentList.hpp>
#include <opentrep/bom/ResultList.hpp>
// Forward declarations
@@ -38,9 +39,17 @@
public:
// /////////// Business methods /////////
/** Retrieve the list of documents matching the query string. */
- void searchString ();
+ void searchString (DocumentList_T&);
+ private:
+ /** Retrieve the document best matching the query string.
+ @param TravelQuery_T& The partial query string.
+ @param MatchingDocument_T& The best matching Xapian document (if found).
+ @return bool Whether such a best matching document has been found. */
+ bool searchString(TravelQuery_T& ioPartialQueryString, MatchingDocument_T&);
+
+
public:
// /////////// Display support methods /////////
/** Dump a Business Object into an output stream.
Modified: trunk/opentrep/opentrep/bom/StringMatcher.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/StringMatcher.cpp 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/opentrep/bom/StringMatcher.cpp 2009-07-18 00:50:00 UTC (rev 135)
@@ -272,21 +272,30 @@
}
// //////////////////////////////////////////////////////////////////////
- void StringMatcher::
- createDocumentListFromMSet (const Xapian::MSet& iMatchingSet,
- DocumentList_T& ioDocumentList) {
- // Empty the list of documents
- ioDocumentList.clear();
+ bool StringMatcher::
+ extractBestMatchingDocumentFromMSet (const Xapian::MSet& iMatchingSet,
+ MatchingDocument_T& ioMatchingDocument) {
+ bool oFoundDocument = false;
- for (Xapian::MSetIterator itDoc = iMatchingSet.begin();
- itDoc != iMatchingSet.end(); ++itDoc) {
- const Xapian::Document& lDocument = itDoc.get_document();
+ if (iMatchingSet.empty() == true) {
+ return oFoundDocument;
+ }
+ oFoundDocument = true;
- ioDocumentList.insert (DocumentList_T::value_type (itDoc.get_percent(),
- lDocument));
- }
+ /**
+ Retrieve the best matching document. If there are several such
+ best matching documents (for instance, several at, say, 100%),
+ one is taken randomly (well, we take the first one of the STL
+ multimap, so it is not exactly randomly, but the result is the
+ same: it appears random).
+ */
+ Xapian::MSetIterator itDoc = iMatchingSet.begin();
+ ioMatchingDocument.first = itDoc.get_percent();
+ ioMatchingDocument.second = itDoc.get_document();
+
+ return oFoundDocument;
}
-
+
// //////////////////////////////////////////////////////////////////////
void StringMatcher::removeOneWord (std::string& ioQueryString) {
assert (ioQueryString.empty() == false);
Modified: trunk/opentrep/opentrep/bom/StringMatcher.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/StringMatcher.hpp 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/opentrep/bom/StringMatcher.hpp 2009-07-18 00:50:00 UTC (rev 135)
@@ -29,10 +29,17 @@
static void searchString (Xapian::MSet&, const std::string& iSearchString,
const Xapian::Database&);
- /** Copy the Xapian MSet (matching set) object into a document
- list object. */
- static void createDocumentListFromMSet (const Xapian::MSet&,
- DocumentList_T&);
+ /**
+ Extract the best matching Xapian document.
+ <br>If there are several such best matching documents (for
+ instance, several at, say, 100%), one is taken randomly. Well,
+ as we take the first one of the STL multimap, it is not exactly
+ randomly, but the result is the same: it appears to be random.
+ @return bool Whether or not there was a matching document.
+ */
+ static bool
+ extractBestMatchingDocumentFromMSet (const Xapian::MSet&,
+ MatchingDocument_T&);
/** Remove the word furthest at right. */
static void removeOneWord (std::string& ioQueryString);
Modified: trunk/opentrep/opentrep/command/RequestInterpreter.cpp
===================================================================
--- trunk/opentrep/opentrep/command/RequestInterpreter.cpp 2009-07-17 17:19:13 UTC (rev 134)
+++ trunk/opentrep/opentrep/command/RequestInterpreter.cpp 2009-07-18 00:50:00 UTC (rev 135)
@@ -15,6 +15,7 @@
#include <opentrep/factory/FacPlaceHolder.hpp>
#include <opentrep/factory/FacPlace.hpp>
#include <opentrep/factory/FacResultHolder.hpp>
+#include <opentrep/factory/FacResult.hpp>
#include <opentrep/command/DBManager.hpp>
#include <opentrep/command/RequestInterpreter.hpp>
#include <opentrep/service/Logger.hpp>
@@ -27,8 +28,7 @@
void RequestInterpreter::
interpretTravelRequest (soci::session& ioSociSession,
const TravelDatabaseName_T& iTravelDatabaseName,
- const TravelQuery_T& iTravelQuery,
- PlaceHolder& ioPlaceHolder) {
+ const TravelQuery_T& iTravelQuery) {
try {
@@ -39,9 +39,30 @@
ResultHolder& lResultHolder =
FacResultHolder::instance().create (iTravelQuery, lXapianDatabase);
- //
- lResultHolder.searchString();
+ // Main algorithm
+ DocumentList_T lDocumentList;
+ lResultHolder.searchString (lDocumentList);
+ // Back-up the (retrieved) matching Xapian documents into still
+ // to-be-created Result objects.
+ for (DocumentList_T::const_iterator itDoc = lDocumentList.begin();
+ itDoc != lDocumentList.end(); ++itDoc) {
+ // Retrieve both the Xapian document object and the corresponding
+ // matching percentage (most of the time, it is 100%)
+ const QueryAndDocument_T& lQueryAndDocument = *itDoc;
+
+ // Create a Result object
+ Result& lResult = FacResult::instance().create (lXapianDatabase);
+
+ // Fill the Result object with both the corresponding Document object
+ // and its associated query string
+ lResult.setQueryAndDocument (lQueryAndDocument);
+
+ // Add the Result object (holding the list of matching
+ // documents) to the dedicated list.
+ FacResultHolder::initLinkWithResult (lResultHolder, lResult);
+ }
+
// DEBUG
OPENTREP_LOG_DEBUG (std::endl
<< "========================================="
@@ -50,6 +71,9 @@
<< "========================================="
<< std::endl << std::endl);
+ // Create a PlaceHolder object, to collect the matching Place objects
+ PlaceHolder& lPlaceHolder = FacPlaceHolder::instance().create();
+
// Browse the list of result objects
const ResultList_T& lResultList = lResultHolder.getResultList();
for (ResultList_T::const_iterator itResult = lResultList.begin();
@@ -59,10 +83,8 @@
assert (lResult_ptr != NULL);
// Retrieve the parameters of the best matching document
- const Xapian::Document& lDocument =
- lResult_ptr->getBestMatchingDocumen...
[truncated message content] |
|
From: <den...@us...> - 2009-07-17 17:19:16
|
Revision: 134
http://opentrep.svn.sourceforge.net/opentrep/?rev=134&view=rev
Author: denis_arnaud
Date: 2009-07-17 17:19:13 +0000 (Fri, 17 Jul 2009)
Log Message:
-----------
[Dev] Added a few factory classes.
Modified Paths:
--------------
trunk/opentrep/opentrep/OPENTREP_Service.hpp
trunk/opentrep/opentrep/batches/indexer.cpp
trunk/opentrep/opentrep/batches/searcher.cpp
trunk/opentrep/opentrep/bom/Names.cpp
trunk/opentrep/opentrep/bom/Names.hpp
trunk/opentrep/opentrep/bom/Place.cpp
trunk/opentrep/opentrep/bom/Place.hpp
trunk/opentrep/opentrep/bom/PlaceList.hpp
trunk/opentrep/opentrep/bom/Result.cpp
trunk/opentrep/opentrep/bom/Result.hpp
trunk/opentrep/opentrep/bom/ResultHolder.cpp
trunk/opentrep/opentrep/bom/ResultHolder.hpp
trunk/opentrep/opentrep/bom/StringMatcher.cpp
trunk/opentrep/opentrep/bom/World.cpp
trunk/opentrep/opentrep/bom/World.hpp
trunk/opentrep/opentrep/bom/sources.mk
trunk/opentrep/opentrep/command/DBManager.cpp
trunk/opentrep/opentrep/command/RequestInterpreter.cpp
trunk/opentrep/opentrep/command/RequestInterpreter.hpp
trunk/opentrep/opentrep/factory/FacPlace.cpp
trunk/opentrep/opentrep/factory/FacPlace.hpp
trunk/opentrep/opentrep/factory/FacWorld.cpp
trunk/opentrep/opentrep/factory/sources.mk
trunk/opentrep/opentrep/service/OPENTREP_Service.cpp
trunk/opentrep/refdata/data/ref_place_names.csv
trunk/opentrep/test/parsers/search_string_parser.cpp
Added Paths:
-----------
trunk/opentrep/opentrep/bom/PlaceHolder.cpp
trunk/opentrep/opentrep/bom/PlaceHolder.hpp
trunk/opentrep/opentrep/factory/FacPlaceHolder.cpp
trunk/opentrep/opentrep/factory/FacPlaceHolder.hpp
trunk/opentrep/opentrep/factory/FacResult.cpp
trunk/opentrep/opentrep/factory/FacResult.hpp
trunk/opentrep/opentrep/factory/FacResultHolder.cpp
trunk/opentrep/opentrep/factory/FacResultHolder.hpp
Modified: trunk/opentrep/opentrep/OPENTREP_Service.hpp
===================================================================
--- trunk/opentrep/opentrep/OPENTREP_Service.hpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/OPENTREP_Service.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -6,6 +6,7 @@
// //////////////////////////////////////////////////////////////////////
// STL
#include <ostream>
+#include <string>
// OPENTREP
#include <opentrep/OPENTREP_Types.hpp>
@@ -18,14 +19,12 @@
class OPENTREP_Service {
public:
/** Constructor. */
- OPENTREP_Service ();
+ OPENTREP_Service (std::ostream& ioLogStream,
+ const std::string& iXapianDatabaseFilepath);
+
/** Destructor. */
~OPENTREP_Service();
- /** Initialise. */
- void init (std::ostream& ioLogStream,
- const std::string& iTravelDatabaseName);
-
/** Build the Xapian database (index) on the BOM held in memory. */
void buildSearchIndex ();
@@ -35,9 +34,15 @@
private:
// /////// Construction and Destruction helper methods ///////
- /** Default Constructor. */
+ /** Default constructor. */
+ OPENTREP_Service ();
+ /** Default copy constructor. */
OPENTREP_Service (const OPENTREP_Service&);
+ /** Initialise. */
+ void init (std::ostream& ioLogStream,
+ const std::string& iXapianDatabaseFilepath);
+
/** Initilise the log. */
void logInit (const LOG::EN_LogLevel iLogLevel, std::ostream& ioLogStream);
Modified: trunk/opentrep/opentrep/batches/indexer.cpp
===================================================================
--- trunk/opentrep/opentrep/batches/indexer.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/batches/indexer.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -127,8 +127,8 @@
logOutputFile.clear();
// Initialise the context
- OPENTREP::OPENTREP_Service opentrepService;
- opentrepService.init (logOutputFile, lXapianDatabaseName);
+ OPENTREP::OPENTREP_Service opentrepService (logOutputFile,
+ lXapianDatabaseName);
// Launch the indexation
opentrepService.buildSearchIndex();
Modified: trunk/opentrep/opentrep/batches/searcher.cpp
===================================================================
--- trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -4,15 +4,73 @@
#include <iostream>
#include <sstream>
#include <fstream>
-#include <map>
#include <vector>
+#include <string>
// Boost (Extended STL)
#include <boost/date_time/posix_time/posix_time.hpp>
#include <boost/date_time/gregorian/gregorian.hpp>
+#include <boost/tokenizer.hpp>
#include <boost/program_options.hpp>
// OPENTREP
#include <opentrep/OPENTREP_Service.hpp>
+#include <opentrep/config/opentrep-paths.hpp>
+// //////// Type definitions ///////
+typedef std::vector<std::string> WordList_T;
+
+// //////// Constants //////
+/** Default name and location for the log file. */
+const std::string K_OPENTREP_DEFAULT_LOG_FILENAME ("opentrep_searcher.log");
+
+/** Default name and location for the Xapian database. */
+const std::string K_OPENTREP_DEFAULT_DATABSE_FILEPATH ("/tmp/opentrep/traveldb");
+
+/** Default travel query string, to be seached against the Xapian database. */
+const std::string K_OPENTREP_DEFAULT_QUERY_STRING ("sna francicso rio de janero lso anglese reykyavki");
+
+/** Default error distance for spelling corrections. */
+const unsigned short K_OPENTREP_DEFAULT_SPELLING_ERROR_DISTANCE = 3;
+
+// //////////////////////////////////////////////////////////////////////
+void tokeniseStringIntoWordList (const std::string& iPhrase,
+ WordList_T& ioWordList) {
+ // Empty the word list
+ ioWordList.clear();
+
+ // Boost Tokeniser
+ typedef boost::tokenizer<boost::char_separator<char> > Tokeniser_T;
+
+ // Define the separators
+ const boost::char_separator<char> lSepatorList(" .,;:|+-*/_=!@#$%`~^&(){}[]?'<>\"");
+
+ // Initialise the phrase to be tokenised
+ Tokeniser_T lTokens (iPhrase, lSepatorList);
+ for (Tokeniser_T::const_iterator tok_iter = lTokens.begin();
+ tok_iter != lTokens.end(); ++tok_iter) {
+ const std::string& lTerm = *tok_iter;
+ ioWordList.push_back (lTerm);
+ }
+
+}
+
+// //////////////////////////////////////////////////////////////////////
+std::string createStringFromWordList (const WordList_T& iWordList) {
+ std::ostringstream oStr;
+
+ unsigned short idx = iWordList.size();
+ for (WordList_T::const_iterator itWord = iWordList.begin();
+ itWord != iWordList.end(); ++itWord, --idx) {
+ const std::string& lWord = *itWord;
+ oStr << lWord;
+ if (idx > 1) {
+ oStr << " ";
+ }
+ }
+
+ return oStr.str();
+}
+
+
// ///////// Parsing of Options & Configuration /////////
// A helper function to simplify the main part.
template<class T> std::ostream& operator<< (std::ostream& os,
@@ -21,34 +79,57 @@
return os;
}
-int readConfiguration (int argc, char* argv[]) {
- int opt;
-
- // Declare a group of options that will be
- // allowed only on command line
- boost::program_options::options_description generic("Generic options");
+/** Early return status (so that it can be differentiated from an error). */
+const int K_OPENTREP_EARLY_RETURN_STATUS = 99;
+
+/** Read and parse the command line options. */
+int readConfiguration (int argc, char* argv[],
+ unsigned short& ioSpellingErrorDistance,
+ std::string& ioQueryString,
+ std::string& ioDatabaseFilepath,
+ std::string& ioLogFilename) {
+
+ // Initialise the travel query string, if that one is empty
+ if (ioQueryString.empty() == true) {
+ ioQueryString = K_OPENTREP_DEFAULT_QUERY_STRING;
+ }
+
+ // Transform the query string into a list of words (STL strings)
+ WordList_T lWordList;
+ tokeniseStringIntoWordList (ioQueryString, lWordList);
+
+ // Declare a group of options that will be allowed only on command line
+ boost::program_options::options_description generic ("Generic options");
generic.add_options()
+ ("prefix", "print installation prefix")
("version,v", "print version string")
("help,h", "produce help message");
- // Declare a group of options that will be allowed both on command line and in
- // config file
- boost::program_options::options_description config("Configuration");
+ // Declare a group of options that will be allowed both on command
+ // line and in config file
+ boost::program_options::options_description config ("Configuration");
config.add_options()
- ("optimization",
- boost::program_options::value<int>(&opt)->default_value(10),
- "optimization level")
- ("include-path,I",
- boost::program_options::value< std::vector<std::string> >()->composing(),
- "include path");
+ ("error,e",
+ boost::program_options::value< unsigned short >(&ioSpellingErrorDistance)->default_value(K_OPENTREP_DEFAULT_SPELLING_ERROR_DISTANCE),
+ "Spelling error distance (e.g., 3)")
+ ("query,q",
+ boost::program_options::value< WordList_T >(&lWordList)->multitoken(),
+ "Traval query word list (e.g. sna francicso rio de janero lso anglese reykyavki")
+ ("database,d",
+ boost::program_options::value< std::string >(&ioDatabaseFilepath)->default_value(K_OPENTREP_DEFAULT_DATABSE_FILEPATH),
+ "Xapian database filepath (e.g., /tmp/opentrep/traveldb)")
+ ("log,l",
+ boost::program_options::value< std::string >(&ioLogFilename)->default_value(K_OPENTREP_DEFAULT_LOG_FILENAME),
+ "Filepath for the logs")
+ ;
// Hidden options, will be allowed both on command line and
// in config file, but will not be shown to the user.
- boost::program_options::options_description hidden("Hidden options");
+ boost::program_options::options_description hidden ("Hidden options");
hidden.add_options()
- ("input-file",
+ ("copyright",
boost::program_options::value< std::vector<std::string> >(),
- "input file");
+ "Show the copyright (license)");
boost::program_options::options_description cmdline_options;
cmdline_options.add(generic).add(config).add(hidden);
@@ -56,45 +137,53 @@
boost::program_options::options_description config_file_options;
config_file_options.add(config).add(hidden);
- boost::program_options::options_description visible("Allowed options");
+ boost::program_options::options_description visible ("Allowed options");
visible.add(generic).add(config);
boost::program_options::positional_options_description p;
- p.add("input-file", -1);
+ p.add ("copyright", -1);
boost::program_options::variables_map vm;
boost::program_options::
- store (boost::program_options::command_line_parser(argc, argv).
- options (cmdline_options).positional(p).run(), vm);
+ store (boost::program_options::command_line_parser (argc, argv).
+ options (cmdline_options).positional(p).run(), vm);
- std::ifstream ifs ("request_parser.cfg");
+ std::ifstream ifs ("opentrep_searcher.cfg");
boost::program_options::store (parse_config_file (ifs, config_file_options),
vm);
boost::program_options::notify (vm);
if (vm.count ("help")) {
std::cout << visible << std::endl;
- return 0;
+ return K_OPENTREP_EARLY_RETURN_STATUS;
}
if (vm.count ("version")) {
- std::cout << "Open Travel Request Parser, version 1.0" << std::endl;
- return 0;
+ std::cout << PACKAGE_NAME << ", version " << PACKAGE_VERSION << std::endl;
+ return K_OPENTREP_EARLY_RETURN_STATUS;
}
- if (vm.count ("include-path")) {
- std::cout << "Include paths are: "
- << vm["include-path"].as< std::vector<std::string> >()
- << std::endl;
+ if (vm.count ("prefix")) {
+ std::cout << "Installation prefix: " << PREFIXDIR << std::endl;
+ return K_OPENTREP_EARLY_RETURN_STATUS;
}
- if (vm.count ("input-file")) {
- std::cout << "Input files are: "
- << vm["input-file"].as< std::vector<std::string> >()
+ if (vm.count ("database")) {
+ ioDatabaseFilepath = vm["database"].as< std::string >();
+ std::cout << "Xapian database filepath is: " << ioDatabaseFilepath
<< std::endl;
}
- std::cout << "Optimization level is " << opt << std::endl;
+ if (vm.count ("log")) {
+ ioLogFilename = vm["log"].as< std::string >();
+ std::cout << "Log filename is: " << ioLogFilename << std::endl;
+ }
+
+ std::cout << "The spelling error distance is: " << ioSpellingErrorDistance
+ << std::endl;
+
+ ioQueryString = createStringFromWordList (lWordList);
+ std::cout << "The travel query string is: " << ioQueryString << std::endl;
return 0;
}
@@ -105,29 +194,26 @@
try {
// Travel query
- OPENTREP::TravelQuery_T lTravelQuery ("sna francisco rio de janero lso angeles");
+ OPENTREP::TravelQuery_T lTravelQuery;
// Output log File
- std::string lLogFilename ("searcher.log");
+ std::string lLogFilename;
// Xapian database name (directory of the index)
- OPENTREP::TravelDatabaseName_T lXapianDatabaseName ("traveldb");
+ OPENTREP::TravelDatabaseName_T lXapianDatabaseName;
- if (argc >= 1 && argv[1] != NULL) {
- std::istringstream istr (argv[1]);
- istr >> lTravelQuery;
- }
+ // Xapian spelling error distance
+ unsigned short lSpellingErrorDistance;
- if (argc >= 2 && argv[2] != NULL) {
- std::istringstream istr (argv[2]);
- istr >> lLogFilename;
+ // Call the command-line option parser
+ const int lOptionParserStatus =
+ readConfiguration (argc, argv, lSpellingErrorDistance, lTravelQuery,
+ lXapianDatabaseName, lLogFilename);
+
+ if (lOptionParserStatus == K_OPENTREP_EARLY_RETURN_STATUS) {
+ return 0;
}
- if (argc >= 3 && argv[3] != NULL) {
- std::istringstream istr (argv[3]);
- istr >> lXapianDatabaseName;
- }
-
// Set the log parameters
std::ofstream logOutputFile;
// open and clean the log outputfile
@@ -135,8 +221,8 @@
logOutputFile.clear();
// Initialise the context
- OPENTREP::OPENTREP_Service opentrepService;
- opentrepService.init (logOutputFile, lXapianDatabaseName);
+ OPENTREP::OPENTREP_Service opentrepService (logOutputFile,
+ lXapianDatabaseName);
// Query the Xapian database (index)
opentrepService.interpretTravelRequest (lTravelQuery);
Modified: trunk/opentrep/opentrep/bom/Names.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/Names.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/Names.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -28,6 +28,17 @@
// //////////////////////////////////////////////////////////////////////
Names::~Names() {
}
+
+ // //////////////////////////////////////////////////////////////////////
+ std::string Names::getFirstName() const {
+ if (_nameList.empty() == true) {
+ return "";
+ }
+ NameList_T::const_iterator itName = _nameList.begin();
+ assert (itName != _nameList.end());
+ const std::string& lName = *itName;
+ return lName;
+ }
// //////////////////////////////////////////////////////////////////////
const std::string Names::describeShortKey() const {
Modified: trunk/opentrep/opentrep/bom/Names.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/Names.hpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/Names.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -46,6 +46,11 @@
return _nameList;
}
+ /** Get the first name of the list.
+ <br>Note that it can be empty (when the list is itself empty). */
+ std::string getFirstName() const;
+
+
// /////////// Setters ///////////////
/** Set the language code. */
void setLanguageCode (const Language::EN_Language& iLanguageCode) {
Modified: trunk/opentrep/opentrep/bom/Place.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/Place.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/Place.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -2,7 +2,7 @@
// Import section
// //////////////////////////////////////////////////////////////////////
// C
-#include <assert.h>
+#include <cassert>
// OpenTrep BOM
#include <opentrep/bom/Place.hpp>
#include <opentrep/service/Logger.hpp>
@@ -10,12 +10,12 @@
namespace OPENTREP {
// //////////////////////////////////////////////////////////////////////
- Place::Place () : _world (NULL) {
+ Place::Place () : _world (NULL), _placeHolder (NULL) {
}
// //////////////////////////////////////////////////////////////////////
Place::Place (const Place& iPlace) :
- _world (iPlace._world),
+ _world (iPlace._world), _placeHolder (iPlace._placeHolder),
_placeCode (iPlace._placeCode), _cityCode (iPlace._cityCode),
_stateCode (iPlace._stateCode), _countryCode (iPlace._countryCode),
_regionCode (iPlace._regionCode), _continentCode (iPlace._continentCode),
@@ -67,6 +67,33 @@
}
// //////////////////////////////////////////////////////////////////////
+ std::string Place::toShortString() const {
+ /* When the city code is empty, it means that the place is a city and
+ not an airport. The city code is thus the same as the place code
+ itself. */
+ std::ostringstream oStr;
+ oStr << describeShortKey() << ", ";
+ if (_cityCode.empty()) {
+ oStr << _placeCode << ", ";
+ } else {
+ oStr << _cityCode << ", ";
+ }
+ oStr << _stateCode
+ << ", " << _countryCode << ", " << _regionCode
+ << ", " << _continentCode << ", " << _timeZoneGroup
+ << ", " << _longitude << ", " << _latitude << ", " << _docID;
+
+ NameMatrix_T::const_iterator itNameHolder = _nameMatrix.begin();
+ const Names& lNameHolder = itNameHolder->second;
+ const std::string& lFirstName = lNameHolder.getFirstName();
+ if (lFirstName.empty() == false) {
+ oStr << ", " << lFirstName << ".";
+ }
+
+ return oStr.str();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
void Place::toStream (std::ostream& ioOut) const {
ioOut << toString();
}
Modified: trunk/opentrep/opentrep/bom/Place.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/Place.hpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/Place.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -18,11 +18,13 @@
// Forward declarations
class World;
+ class PlaceHolder;
/** Structure modelling an place. */
class Place : public BomAbstract {
friend class FacWorld;
friend class FacPlace;
+ friend class FacPlaceHolder;
friend class DbaPlace;
public:
// ///////// Getters ////////
@@ -141,6 +143,7 @@
/** Reset the map of name lists. */
void resetMatrix();
+
public:
// ///////// Display methods ////////
@@ -155,6 +158,9 @@
/** Get the serialised version of the Place object. */
std::string toString() const;
+ /** Get a short display of the Business Object. */
+ std::string toShortString() const;
+
/** Get a string describing the whole key (differentiating two objects
at any level). */
const std::string describeKey() const;
@@ -182,6 +188,9 @@
/** Parent World. */
World* _world;
+ /** Parent PlaceHolder. */
+ PlaceHolder* _placeHolder;
+
private:
// /////// Attributes /////////
/** Place code. */
Added: trunk/opentrep/opentrep/bom/PlaceHolder.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/PlaceHolder.cpp (rev 0)
+++ trunk/opentrep/opentrep/bom/PlaceHolder.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -0,0 +1,81 @@
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// C
+#include <cassert>
+#include <sstream>
+// OpenTREP
+#include <opentrep/bom/Place.hpp>
+#include <opentrep/bom/PlaceHolder.hpp>
+#include <opentrep/service/Logger.hpp>
+
+namespace OPENTREP {
+
+ // //////////////////////////////////////////////////////////////////////
+ PlaceHolder::PlaceHolder () {
+ init();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ PlaceHolder::~PlaceHolder () {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void PlaceHolder::init () {
+ _placeList.clear();
+ _placeOrderedList.clear();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ const std::string PlaceHolder::describeShortKey() const {
+ std::ostringstream oStr;
+ return oStr.str();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ const std::string PlaceHolder::describeKey() const {
+ return describeShortKey();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ std::string PlaceHolder::toString() const {
+ std::ostringstream oStr;
+ oStr << describeShortKey() << std::endl;
+
+ for (PlaceOrderedList_T::const_iterator itPlace = _placeOrderedList.begin();
+ itPlace != _placeOrderedList.end(); ++itPlace) {
+ const Place* lPlace_ptr = *itPlace;
+ assert (lPlace_ptr != NULL);
+
+ oStr << lPlace_ptr->toString() << std::endl;
+ }
+
+ return oStr.str();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ std::string PlaceHolder::toShortString() const {
+ std::ostringstream oStr;
+ oStr << describeShortKey() << std::endl;
+
+ for (PlaceOrderedList_T::const_iterator itPlace = _placeOrderedList.begin();
+ itPlace != _placeOrderedList.end(); ++itPlace) {
+ const Place* lPlace_ptr = *itPlace;
+ assert (lPlace_ptr != NULL);
+
+ oStr << lPlace_ptr->toShortString() << std::endl;
+ }
+
+ return oStr.str();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void PlaceHolder::toStream (std::ostream& ioOut) const {
+ ioOut << toString();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void PlaceHolder::fromStream (std::istream& ioIn) {
+ }
+
+}
Added: trunk/opentrep/opentrep/bom/PlaceHolder.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/PlaceHolder.hpp (rev 0)
+++ trunk/opentrep/opentrep/bom/PlaceHolder.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -0,0 +1,80 @@
+#ifndef __OPENTREP_BOM_PLACEHOLDER_HPP
+#define __OPENTREP_BOM_PLACEHOLDER_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// OpenTREP
+#include <opentrep/OPENTREP_Types.hpp>
+#include <opentrep/bom/BomAbstract.hpp>
+#include <opentrep/bom/PlaceList.hpp>
+
+namespace OPENTREP {
+
+ /** Class wrapping functions on a list of Place objects. */
+ class PlaceHolder : public BomAbstract {
+ friend class FacPlaceHolder;
+ public:
+ // ////////////// Getters /////////////
+ /** Retrieve the list of place objects. */
+ const PlaceList_T& getPlaceList() const {
+ return _placeList;
+ }
+
+
+ // ////////////// Setters /////////////
+
+
+ public:
+ // /////////// Business methods /////////
+
+
+ public:
+ // /////////// Display support methods /////////
+ /** Dump a Business Object into an output stream.
+ @param ostream& the output stream. */
+ void toStream (std::ostream& ioOut) const;
+
+ /** Read a Business Object from an input stream.
+ @param istream& the input stream. */
+ void fromStream (std::istream& ioIn);
+
+ /** Get the serialised version of the Business Object. */
+ std::string toString() const;
+
+ /** Get a short display of the Business Object. */
+ std::string toShortString() const;
+
+ /** Get a string describing the whole key (differentiating two objects
+ at any level). */
+ const std::string describeKey() const;
+
+ /** Get a string describing the short key (differentiating two objects
+ at the same level). */
+ const std::string describeShortKey() const;
+
+
+ private:
+ // ////////////// Constructors and Destructors /////////////
+ /** Default constructor. */
+ PlaceHolder ();
+ /** Default copy constructor. */
+ PlaceHolder (const PlaceHolder&);
+ /** Destructor. */
+ ~PlaceHolder ();
+ /** Initialise (reset the list of documents). */
+ void init ();
+
+
+ private:
+ // /////////////// Attributes ////////////////
+ /** List of place objects, sorted by Place ID. */
+ PlaceList_T _placeList;
+
+ /** List of place objects, the sort order corresponding to their
+ insertion order. */
+ PlaceOrderedList_T _placeOrderedList;
+ };
+
+}
+#endif // __OPENTREP_BOM_PLACEHOLDER_HPP
Modified: trunk/opentrep/opentrep/bom/PlaceList.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/PlaceList.hpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/PlaceList.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -15,8 +15,9 @@
// ///////////// Type definitions ////////////////////
typedef std::size_t PlaceID_T;
- typedef std::map<PlaceID_T, Place*> PlaceList_T;
- typedef std::list<Place*> SimplePlaceList_T;
+ // typedef std::map<PlaceID_T, Place*> PlaceDirectList_T;
+ typedef std::map<std::string, Place*> PlaceList_T;
+ typedef std::list<Place*> PlaceOrderedList_T;
}
#endif // __OPENTREP_BOM_PLACELIST_HPP
Modified: trunk/opentrep/opentrep/bom/Result.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/Result.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/Result.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -14,7 +14,7 @@
// //////////////////////////////////////////////////////////////////////
Result::Result (const Xapian::Database& iDatabase)
- : _database (iDatabase) {
+ : _resultHolder (NULL), _database (iDatabase) {
init();
}
Modified: trunk/opentrep/opentrep/bom/Result.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/Result.hpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/Result.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -11,10 +11,13 @@
namespace OPENTREP {
+ // Forward declarations
+ class ResultHolder;
+
/** Class wrapping functions on a list of Xapian Document objects. */
class Result : public BomAbstract {
+ friend class FacResultHolder;
friend class FacResult;
- friend class ResultHolder;
public:
// ////////////// Getters /////////////
/** Get the query string. */
@@ -86,6 +89,9 @@
private:
// /////////////// Attributes ////////////////
+ /** Parent ResultHolder. */
+ ResultHolder* _resultHolder;
+
/** Query string having generated the list of document. */
TravelQuery_T _queryString;
Modified: trunk/opentrep/opentrep/bom/ResultHolder.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/ResultHolder.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/ResultHolder.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -10,6 +10,9 @@
#include <opentrep/bom/StringMatcher.hpp>
#include <opentrep/bom/Result.hpp>
#include <opentrep/bom/ResultHolder.hpp>
+// TODO: move that out of the BOM layer
+#include <opentrep/factory/FacResultHolder.hpp>
+#include <opentrep/factory/FacResult.hpp>
#include <opentrep/service/Logger.hpp>
namespace OPENTREP {
@@ -89,18 +92,19 @@
resulting string gets empty.
*/
DocumentList_T lDocumentList;
- Result* lResult_ptr = new Result (_database);
- assert (lResult_ptr != NULL);
+ // TODO: move that out of the BOM layer
+ Result& lResult = FacResult::instance().create (_database);
std::string lQueryString (lRemainingQueryString);
//
- lResult_ptr->setQueryString (lQueryString);
- lResult_ptr->searchString ();
+ lResult.setQueryString (lQueryString);
+ lResult.searchString ();
// Add the Result object (holding the list of matching
// documents) to the dedicated list.
- _resultList.push_back (lResult_ptr);
+ // TODO: move that out of the BOM layer
+ FacResultHolder::initLinkWithResult (*this, lResult);
/**
Remove, from the lRemainingQueryString string, the part which
@@ -113,7 +117,7 @@
'rio de janeiro'. So, the already parsed part, namely
'sna francisco', must be subtracted from the initial query string.
*/
- lQueryString = lResult_ptr->getQueryString();
+ lQueryString = lResult.getQueryString();
StringMatcher::subtractParsedToRemaining (lQueryString,
lRemainingQueryString);
Modified: trunk/opentrep/opentrep/bom/ResultHolder.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/ResultHolder.hpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/ResultHolder.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -19,7 +19,6 @@
/** Class wrapping functions on a list of Result objects. */
class ResultHolder : public BomAbstract {
friend class FacResultHolder;
- friend class RequestInterpreter;
public:
// ////////////// Getters /////////////
/** Get the query string. */
Modified: trunk/opentrep/opentrep/bom/StringMatcher.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/StringMatcher.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/StringMatcher.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -20,13 +20,31 @@
namespace OPENTREP {
+ // ///////////// Type definitions //////////
+ /** Edit distance (e.g., 2 or 3). */
+ typedef unsigned int EditDistance_T;
+
+ /** Given the size of the phrase, determine the allowed edit distance for
+ spelling purpose. For instance, an edit distance of 1 will be allowed
+ on a 4-letter word, while an edit distance of 3 will be allowed on an
+ 11-letter word. */
+ // //////////////////////////////////////////////////////////////////////
+ static unsigned int calculateEditDistance (const std::string& iPhrase) {
+ EditDistance_T oEditDistance = 2;
+
+ const EditDistance_T lQueryStringSize = iPhrase.size();
+
+ oEditDistance = lQueryStringSize / 3;
+ return oEditDistance;
+ }
+
/** For each of the word in the given list, perform spelling
corrections. If the word is correctly spelled, it is copied as
is. Otherwise, a corrected version is stored. */
// //////////////////////////////////////////////////////////////////////
static void createCorrectedWordList (const WordList_T& iOriginalWordList,
- WordList_T& ioCorrectedWordList,
- const Xapian::Database& iDatabase) {
+ WordList_T& ioCorrectedWordList,
+ const Xapian::Database& iDatabase) {
// Empty the target list
ioCorrectedWordList.clear();
@@ -36,8 +54,9 @@
for (WordList_T::const_iterator itWord = iOriginalWordList.begin();
itWord != iOriginalWordList.end(); ++itWord) {
const std::string& lOriginalWord = *itWord;
+ const EditDistance_T lEditDistance= calculateEditDistance(lOriginalWord);
const std::string& lSuggestedWord =
- iDatabase.get_spelling_suggestion (lOriginalWord, 3);
+ iDatabase.get_spelling_suggestion (lOriginalWord, lEditDistance);
if (lSuggestedWord.empty() == true) {
ioCorrectedWordList.push_back (lOriginalWord);
@@ -110,8 +129,10 @@
phrase/string. With the above example, 'sna francisco' yields the
suggestion 'san francisco'.
*/
+ const EditDistance_T lEditDistance =
+ calculateEditDistance (lOriginalQueryString);
const std::string lFullWordCorrectedString =
- ioDatabase.get_spelling_suggestion (lOriginalQueryString, 3);
+ ioDatabase.get_spelling_suggestion (lOriginalQueryString, lEditDistance);
// DEBUG
/*
Modified: trunk/opentrep/opentrep/bom/World.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/World.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/World.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -62,8 +62,8 @@
oStr << shortDisplay();
- for (SimplePlaceList_T::const_iterator itPlace = _simplePlaceList.begin();
- itPlace != _simplePlaceList.end(); ++itPlace) {
+ for (PlaceOrderedList_T::const_iterator itPlace = _placeOrderedList.begin();
+ itPlace != _placeOrderedList.end(); ++itPlace) {
const Place* lPlace_ptr = *itPlace;
assert (lPlace_ptr != NULL);
Modified: trunk/opentrep/opentrep/bom/World.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/World.hpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/World.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -23,14 +23,14 @@
public:
// //////////// Getters /////////////
/** Get the list of Place objects.
- const PlaceList_T& getPlaceList () const {
+ const PlaceDirectList_T& getPlaceList () const {
return _placeList;
}
*/
/** Get the list of Place objects. */
- const SimplePlaceList_T& getSimplePlaceList () const {
- return _simplePlaceList;
+ const PlaceOrderedList_T& getSimplePlaceList () const {
+ return _placeOrderedList;
}
// //////////// Setters /////////////
@@ -80,12 +80,12 @@
/** List of Place objects.
<br>That list is actually a STL map, indexed on the
Xapian document ID. */
- // PlaceList_T _placeList;
+ // PlaceDirectList_T _placeList;
/** List of Place objects.
<br>That list is actually a STL list, to store temporarily Place
objects when indexing the Xapian database. */
- SimplePlaceList_T _simplePlaceList;
+ PlaceOrderedList_T _placeOrderedList;
};
// ///////////// Type definitions ////////////////////
Modified: trunk/opentrep/opentrep/bom/sources.mk
===================================================================
--- trunk/opentrep/opentrep/bom/sources.mk 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/bom/sources.mk 2009-07-17 17:19:13 UTC (rev 134)
@@ -3,11 +3,12 @@
$(top_srcdir)/opentrep/bom/Language.hpp \
$(top_srcdir)/opentrep/bom/GenericBom.hpp \
$(top_srcdir)/opentrep/bom/World.hpp \
+ $(top_srcdir)/opentrep/bom/WordList.hpp \
+ $(top_srcdir)/opentrep/bom/WordHolder.hpp \
$(top_srcdir)/opentrep/bom/Names.hpp \
$(top_srcdir)/opentrep/bom/Place.hpp \
$(top_srcdir)/opentrep/bom/PlaceList.hpp \
- $(top_srcdir)/opentrep/bom/WordList.hpp \
- $(top_srcdir)/opentrep/bom/WordHolder.hpp \
+ $(top_srcdir)/opentrep/bom/PlaceHolder.hpp \
$(top_srcdir)/opentrep/bom/DocumentList.hpp \
$(top_srcdir)/opentrep/bom/Result.hpp \
$(top_srcdir)/opentrep/bom/ResultList.hpp \
@@ -17,9 +18,10 @@
$(top_srcdir)/opentrep/bom/BomType.cpp \
$(top_srcdir)/opentrep/bom/Language.cpp \
$(top_srcdir)/opentrep/bom/World.cpp \
+ $(top_srcdir)/opentrep/bom/WordHolder.cpp \
$(top_srcdir)/opentrep/bom/Names.cpp \
$(top_srcdir)/opentrep/bom/Place.cpp \
- $(top_srcdir)/opentrep/bom/WordHolder.cpp \
+ $(top_srcdir)/opentrep/bom/PlaceHolder.cpp \
$(top_srcdir)/opentrep/bom/Result.cpp \
$(top_srcdir)/opentrep/bom/ResultHolder.cpp \
$(top_srcdir)/opentrep/bom/StringMatcher.cpp
Modified: trunk/opentrep/opentrep/command/DBManager.cpp
===================================================================
--- trunk/opentrep/opentrep/command/DBManager.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/command/DBManager.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -185,7 +185,7 @@
hasStillData = iterateOnStatement (lSelectStatement, ioPlace,
shouldNotDoReset);
if (hasStillData == true) {
- throw new MultipleRowsForASingleDocIDException();
+ throw MultipleRowsForASingleDocIDException();
}
// Debug
Modified: trunk/opentrep/opentrep/command/RequestInterpreter.cpp
===================================================================
--- trunk/opentrep/opentrep/command/RequestInterpreter.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/command/RequestInterpreter.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -11,7 +11,10 @@
#include <opentrep/bom/Place.hpp>
#include <opentrep/bom/ResultHolder.hpp>
#include <opentrep/bom/Result.hpp>
+#include <opentrep/bom/PlaceHolder.hpp>
+#include <opentrep/factory/FacPlaceHolder.hpp>
#include <opentrep/factory/FacPlace.hpp>
+#include <opentrep/factory/FacResultHolder.hpp>
#include <opentrep/command/DBManager.hpp>
#include <opentrep/command/RequestInterpreter.hpp>
#include <opentrep/service/Logger.hpp>
@@ -24,40 +27,31 @@
void RequestInterpreter::
interpretTravelRequest (soci::session& ioSociSession,
const TravelDatabaseName_T& iTravelDatabaseName,
- const TravelQuery_T& iTravelQuery) {
+ const TravelQuery_T& iTravelQuery,
+ PlaceHolder& ioPlaceHolder) {
try {
// Make the database
Xapian::Database lXapianDatabase (iTravelDatabaseName);
- // TODO: Use FacResultHolder for the following
- ResultHolder* lResultHolder_ptr = new ResultHolder (iTravelQuery,
- lXapianDatabase);
- assert (lResultHolder_ptr != NULL);
+ // Create a ResultHolder object
+ ResultHolder& lResultHolder =
+ FacResultHolder::instance().create (iTravelQuery, lXapianDatabase);
//
- lResultHolder_ptr->searchString();
+ lResultHolder.searchString();
// DEBUG
OPENTREP_LOG_DEBUG (std::endl
- << "_________________________________________"
- << std::endl
<< "========================================="
- << std::endl
- << "-----------------------------------------"
- << std::endl
- << "Matching list: " << std::endl
- << lResultHolder_ptr->toString()
- << "_________________________________________"
- << std::endl
+ << std::endl << "Matching list: " << std::endl
+ << lResultHolder.toString()
<< "========================================="
- << std::endl
- << "-----------------------------------------"
<< std::endl << std::endl);
// Browse the list of result objects
- const ResultList_T& lResultList = lResultHolder_ptr->getResultList();
+ const ResultList_T& lResultList = lResultHolder.getResultList();
for (ResultList_T::const_iterator itResult = lResultList.begin();
itResult != lResultList.end(); ++itResult) {
// Retrieve the result object
@@ -86,8 +80,14 @@
DBManager::retrievePlace (ioSociSession, lDocID, lPlace);
if (hasRetrievedPlace == true) {
+ // Insert the Place object within the PlaceHolder object
+ FacPlaceHolder::initLinkWithPlace (ioPlaceHolder, lPlace);
+
+ // DEBUG
OPENTREP_LOG_DEBUG ("Retrieved Document: " << lPlace.toString());
+
} else {
+ // DEBUG
OPENTREP_LOG_DEBUG ("No retrieved Document for ID = " << lDocID);
}
}
Modified: trunk/opentrep/opentrep/command/RequestInterpreter.hpp
===================================================================
--- trunk/opentrep/opentrep/command/RequestInterpreter.hpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/command/RequestInterpreter.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -16,6 +16,9 @@
namespace OPENTREP {
+ // Forward declarations
+ class PlaceHolder;
+
/** Command wrapping the travel request process. */
class RequestInterpreter {
friend class OPENTREP_Service;
@@ -23,7 +26,7 @@
/** Interpret a search query. */
static void interpretTravelRequest (soci::session&,
const TravelDatabaseName_T&,
- const TravelQuery_T&);
+ const TravelQuery_T&, PlaceHolder&);
private:
/** Constructors. */
Modified: trunk/opentrep/opentrep/factory/FacPlace.cpp
===================================================================
--- trunk/opentrep/opentrep/factory/FacPlace.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/factory/FacPlace.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -2,7 +2,7 @@
// Import section
// //////////////////////////////////////////////////////////////////////
// C
-#include <assert.h>
+#include <cassert>
// OPENTREP
#include <opentrep/bom/Place.hpp>
#include <opentrep/factory/FacSupervisor.hpp>
@@ -13,6 +13,14 @@
FacPlace* FacPlace::_instance = NULL;
// //////////////////////////////////////////////////////////////////////
+ FacPlace::FacPlace () {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ FacPlace::FacPlace (const FacPlace&) {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
FacPlace::~FacPlace () {
_instance = NULL;
}
Modified: trunk/opentrep/opentrep/factory/FacPlace.hpp
===================================================================
--- trunk/opentrep/opentrep/factory/FacPlace.hpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/factory/FacPlace.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -40,8 +40,8 @@
/** Default Constructor.
<br>This constructor is private in order to ensure the singleton
pattern.*/
- FacPlace () {}
- FacPlace (const FacPlace&) {}
+ FacPlace ();
+ FacPlace (const FacPlace&);
private:
/** The unique instance.*/
Added: trunk/opentrep/opentrep/factory/FacPlaceHolder.cpp
===================================================================
--- trunk/opentrep/opentrep/factory/FacPlaceHolder.cpp (rev 0)
+++ trunk/opentrep/opentrep/factory/FacPlaceHolder.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -0,0 +1,76 @@
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// C
+#include <cassert>
+// OPENTREP
+#include <opentrep/bom/PlaceHolder.hpp>
+#include <opentrep/bom/Place.hpp>
+#include <opentrep/factory/FacSupervisor.hpp>
+#include <opentrep/factory/FacPlaceHolder.hpp>
+#include <opentrep/service/Logger.hpp>
+
+namespace OPENTREP {
+
+ FacPlaceHolder* FacPlaceHolder::_instance = NULL;
+
+ // //////////////////////////////////////////////////////////////////////
+ FacPlaceHolder::FacPlaceHolder () {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ FacPlaceHolder::FacPlaceHolder (const FacPlaceHolder&) {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ FacPlaceHolder::~FacPlaceHolder () {
+ _instance = NULL;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ FacPlaceHolder& FacPlaceHolder::instance () {
+
+ if (_instance == NULL) {
+ _instance = new FacPlaceHolder();
+ assert (_instance != NULL);
+
+ FacSupervisor::instance().registerBomFactory (_instance);
+ }
+ return *_instance;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ PlaceHolder& FacPlaceHolder::create () {
+ PlaceHolder* oPlaceHolder_ptr = NULL;
+
+ oPlaceHolder_ptr = new PlaceHolder ();
+ assert (oPlaceHolder_ptr != NULL);
+
+ // The new object is added to the Bom pool
+ _pool.push_back (oPlaceHolder_ptr);
+
+ return *oPlaceHolder_ptr;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void FacPlaceHolder::initLinkWithPlace (PlaceHolder& ioPlaceHolder,
+ Place& ioPlace) {
+ // Link the PlaceHolder to the Place, and vice versa
+ ioPlace._placeHolder = &ioPlaceHolder;
+
+ // Add the Place to the PlaceHolder internal map (of Place objects)
+ const bool insertSucceeded = ioPlaceHolder._placeList.
+ insert (PlaceList_T::value_type (ioPlace.describeShortKey(),
+ &ioPlace)).second;
+ if (insertSucceeded == false) {
+ OPENTREP_LOG_ERROR ("Insertion failed for "
+ << ioPlaceHolder.describeKey()
+ << " and " << ioPlace.describeShortKey());
+ assert (insertSucceeded == true);
+ }
+
+ // Add the Place to the PlaceHolder internal list (of Place objects)
+ ioPlaceHolder._placeOrderedList.push_back (&ioPlace);
+ }
+
+}
Added: trunk/opentrep/opentrep/factory/FacPlaceHolder.hpp
===================================================================
--- trunk/opentrep/opentrep/factory/FacPlaceHolder.hpp (rev 0)
+++ trunk/opentrep/opentrep/factory/FacPlaceHolder.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -0,0 +1,56 @@
+#ifndef __OPENTREP_FAC_FACPLACEHOLDER_HPP
+#define __OPENTREP_FAC_FACPLACEHOLDER_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// OPENTREP
+#include <opentrep/factory/FacBomAbstract.hpp>
+
+namespace OPENTREP {
+
+ /** Forward declarations. */
+ class PlaceHolder;
+ class Place;
+
+ /** Factory for Place. */
+ class FacPlaceHolder : public FacBomAbstract {
+ public:
+
+ /** Provide the unique instance.
+ <br> The singleton is instantiated when first used
+ @return FacPlaceHolder& */
+ static FacPlaceHolder& instance();
+
+ /** Destructor.
+ <br> The Destruction put the _instance to NULL
+ in order to be clean for the next FacPlaceHolder::instance() */
+ virtual ~FacPlaceHolder();
+
+ /** Create a new PlaceHolder object.
+ <br>This new object is added to the list of instantiated objects.
+ @return PlaceHolder& The newly created object. */
+ PlaceHolder& create ();
+
+ /** Initialise the link between a PlaceHolder and a Place.
+ @param PlaceHolder&
+ @param Place&
+ @exception FacExceptionNullPointer
+ @exception FacException.*/
+ static void initLinkWithPlace (PlaceHolder&, Place&);
+
+
+ private:
+ /** Default Constructor.
+ <br>This constructor is private in order to ensure the singleton
+ pattern.*/
+ FacPlaceHolder ();
+ FacPlaceHolder (const FacPlaceHolder&);
+
+ private:
+ /** The unique instance.*/
+ static FacPlaceHolder* _instance;
+
+ };
+}
+#endif // __OPENTREP_FAC_FACPLACEHOLDER_HPP
Added: trunk/opentrep/opentrep/factory/FacResult.cpp
===================================================================
--- trunk/opentrep/opentrep/factory/FacResult.cpp (rev 0)
+++ trunk/opentrep/opentrep/factory/FacResult.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -0,0 +1,66 @@
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// C
+#include <cassert>
+// OPENTREP
+#include <opentrep/bom/Result.hpp>
+#include <opentrep/factory/FacSupervisor.hpp>
+#include <opentrep/factory/FacResult.hpp>
+
+namespace OPENTREP {
+
+ FacResult* FacResult::_instance = NULL;
+
+ // //////////////////////////////////////////////////////////////////////
+ FacResult::FacResult () {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ FacResult::FacResult (const FacResult&) {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ FacResult::~FacResult () {
+ _instance = NULL;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ FacResult& FacResult::instance () {
+
+ if (_instance == NULL) {
+ _instance = new FacResult();
+ assert (_instance != NULL);
+
+ FacSupervisor::instance().registerBomFactory (_instance);
+ }
+ return *_instance;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ Result& FacResult::create (const Xapian::Database& iXapianDatabase) {
+ Result* oResult_ptr = NULL;
+
+ oResult_ptr = new Result (iXapianDatabase);
+ assert (oResult_ptr != NULL);
+
+ // The new object is added to the Bom pool
+ _pool.push_back (oResult_ptr);
+
+ return *oResult_ptr;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ Result& FacResult::clone (const Result& iResult) {
+ Result* oResult_ptr = NULL;
+
+ oResult_ptr = new Result (iResult);
+ assert (oResult_ptr != NULL);
+
+ // The new object is added to the Bom pool
+ _pool.push_back (oResult_ptr);
+
+ return *oResult_ptr;
+ }
+
+}
Added: trunk/opentrep/opentrep/factory/FacResult.hpp
===================================================================
--- trunk/opentrep/opentrep/factory/FacResult.hpp (rev 0)
+++ trunk/opentrep/opentrep/factory/FacResult.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -0,0 +1,52 @@
+#ifndef __OPENTREP_FAC_FACRESULT_HPP
+#define __OPENTREP_FAC_FACRESULT_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// OPENTREP
+#include <opentrep/factory/FacBomAbstract.hpp>
+
+namespace OPENTREP {
+
+ /** Forward declarations. */
+ class Result;
+
+ /** Factory for Result. */
+ class FacResult : public FacBomAbstract {
+ public:
+
+ /** Provide the unique instance.
+ <br> The singleton is instantiated when first used
+ @return FacResult& */
+ static FacResult& instance();
+
+ /** Destructor.
+ <br> The Destruction put the _instance to NULL
+ in order to be clean for the next FacResult::instance() */
+ virtual ~FacResult();
+
+ /** Create a new Result object.
+ <br>This new object is added to the list of instantiated objects.
+ @return Result& The newly created object. */
+ Result& create (const Xapian::Database&);
+
+ /** Create a copy of a Result object.
+ <br>This new object is added to the list of instantiated objects.
+ @return Result& The newly created object. */
+ Result& clone (const Result&);
+
+ private:
+ /** Default Constructor.
+ <br>This constructor is private in order to ensure the singleton
+ pattern.*/
+ FacResult ();
+ FacResult (const FacResult&);
+
+ private:
+ /** The unique instance.*/
+ static FacResult* _instance;
+
+ };
+}
+#endif // __OPENTREP_FAC_FACRESULT_HPP
Added: trunk/opentrep/opentrep/factory/FacResultHolder.cpp
===================================================================
--- trunk/opentrep/opentrep/factory/FacResultHolder.cpp (rev 0)
+++ trunk/opentrep/opentrep/factory/FacResultHolder.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -0,0 +1,66 @@
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// C
+#include <cassert>
+// OPENTREP
+#include <opentrep/bom/ResultHolder.hpp>
+#include <opentrep/bom/Result.hpp>
+#include <opentrep/factory/FacSupervisor.hpp>
+#include <opentrep/factory/FacResultHolder.hpp>
+#include <opentrep/service/Logger.hpp>
+
+namespace OPENTREP {
+
+ FacResultHolder* FacResultHolder::_instance = NULL;
+
+ // //////////////////////////////////////////////////////////////////////
+ FacResultHolder::FacResultHolder () {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ FacResultHolder::FacResultHolder (const FacResultHolder&) {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ FacResultHolder::~FacResultHolder () {
+ _instance = NULL;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ FacResultHolder& FacResultHolder::instance () {
+
+ if (_instance == NULL) {
+ _instance = new FacResultHolder();
+ assert (_instance != NULL);
+
+ FacSupervisor::instance().registerBomFactory (_instance);
+ }
+ return *_instance;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ ResultHolder& FacResultHolder::create (const TravelQuery_T& iQueryString,
+ const Xapian::Database& iDatabase) {
+ ResultHolder* oResultHolder_ptr = NULL;
+
+ oResultHolder_ptr = new ResultHolder (iQueryString, iDatabase);
+ assert (oResultHolder_ptr != NULL);
+
+ // The new object is added to the Bom pool
+ _pool.push_back (oResultHolder_ptr);
+
+ return *oResultHolder_ptr;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void FacResultHolder::initLinkWithResult (ResultHolder& ioResultHolder,
+ Result& ioResult) {
+ // Link the ResultHolder to the Result, and vice versa
+ ioResult._resultHolder = &ioResultHolder;
+
+ // Add the Result to the ResultHolder internal list (of Result objects)
+ ioResultHolder._resultList.push_back (&ioResult);
+ }
+
+}
Added: trunk/opentrep/opentrep/factory/FacResultHolder.hpp
===================================================================
--- trunk/opentrep/opentrep/factory/FacResultHolder.hpp (rev 0)
+++ trunk/opentrep/opentrep/factory/FacResultHolder.hpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -0,0 +1,63 @@
+#ifndef __OPENTREP_FAC_FACRESULTHOLDER_HPP
+#define __OPENTREP_FAC_FACRESULTHOLDER_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// OpenTREP
+#include <opentrep/factory/FacBomAbstract.hpp>
+#include <opentrep/OPENTREP_Types.hpp>
+
+// Forward declarations
+namespace Xapian {
+ class Database;
+}
+
+namespace OPENTREP {
+
+ /** Forward declarations. */
+ class ResultHolder;
+ class Result;
+
+ /** Factory for Result. */
+ class FacResultHolder : public FacBomAbstract {
+ public:
+
+ /** Provide the unique instance.
+ <br> The singleton is instantiated when first used
+ @return FacResultHolder& */
+ static FacResultHolder& instance();
+
+ /** Destructor.
+ <br> The Destruction put the _instance to NULL
+ in order to be clean for the next FacResultHolder::instance() */
+ virtual ~FacResultHolder();
+
+ /** Create a new ResultHolder object.
+ <br>This new object is added to the list of instantiated objects.
+ @return ResultHolder& The newly created object. */
+ ResultHolder& create (const TravelQuery_T& iQueryString,
+ const Xapian::Database& iDatabase);
+
+ /** Initialise the link between a ResultHolder and a Result.
+ @param ResultHolder&
+ @param Result&
+ @exception FacExceptionNullPointer
+ @exception FacException.*/
+ static void initLinkWithResult (ResultHolder&, Result&);
+
+
+ private:
+ /** Default Constructor.
+ <br>This constructor is private in order to ensure the singleton
+ pattern.*/
+ FacResultHolder ();
+ FacResultHolder (const FacResultHolder&);
+
+ private:
+ /** The unique instance.*/
+ static FacResultHolder* _instance;
+
+ };
+}
+#endif // __OPENTREP_FAC_FACRESULTHOLDER_HPP
Modified: trunk/opentrep/opentrep/factory/FacWorld.cpp
===================================================================
--- trunk/opentrep/opentrep/factory/FacWorld.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/factory/FacWorld.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -71,8 +71,8 @@
// Add the Place pointer to the dedicated list within the World object
/*
- const bool insertSucceeded2 = ioWorld._placeList.
- insert (PlaceList_T::value_type (lPlaceID, &ioPlace)).second;
+ const bool insertSucceeded2 = ioWorld._placeDirectList.
+ insert (PlaceDirectList_T::value_type (lPlaceID, &ioPlace)).second;
if (insertSucceeded2 == false) {
OPENTREP_LOG_ERROR ("Insertion failed for " << ioWorld.describeKey()
<< " and " << ioPlace.describeShortKey());
@@ -83,7 +83,7 @@
}
// Add the Place pointer to the dedicated list within the World object
- ioWorld._simplePlaceList.push_back (&ioPlace);
+ ioWorld._placeOrderedList.push_back (&ioPlace);
}
// //////////////////////////////////////////////////////////////////////
Modified: trunk/opentrep/opentrep/factory/sources.mk
===================================================================
--- trunk/opentrep/opentrep/factory/sources.mk 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/factory/sources.mk 2009-07-17 17:19:13 UTC (rev 134)
@@ -3,10 +3,16 @@
$(top_srcdir)/opentrep/factory/FacSupervisor.hpp \
$(top_srcdir)/opentrep/factory/FacOpenTrepServiceContext.hpp \
$(top_srcdir)/opentrep/factory/FacWorld.hpp \
- $(top_srcdir)/opentrep/factory/FacPlace.hpp
+ $(top_srcdir)/opentrep/factory/FacPlaceHolder.hpp \
+ $(top_srcdir)/opentrep/factory/FacPlace.hpp \
+ $(top_srcdir)/opentrep/factory/FacResultHolder.hpp \
+ $(top_srcdir)/opentrep/factory/FacResult.hpp
fac_cc_sources = $(top_srcdir)/opentrep/factory/FacBomAbstract.cpp \
$(top_srcdir)/opentrep/factory/FacServiceAbstract.cpp \
$(top_srcdir)/opentrep/factory/FacSupervisor.cpp \
$(top_srcdir)/opentrep/factory/FacOpenTrepServiceContext.cpp \
$(top_srcdir)/opentrep/factory/FacWorld.cpp \
- $(top_srcdir)/opentrep/factory/FacPlace.cpp
+ $(top_srcdir)/opentrep/factory/FacPlaceHolder.cpp \
+ $(top_srcdir)/opentrep/factory/FacPlace.cpp \
+ $(top_srcdir)/opentrep/factory/FacResultHolder.cpp \
+ $(top_srcdir)/opentrep/factory/FacResult.cpp
Modified: trunk/opentrep/opentrep/service/OPENTREP_Service.cpp
===================================================================
--- trunk/opentrep/opentrep/service/OPENTREP_Service.cpp 2009-07-17 00:10:38 UTC (rev 133)
+++ trunk/opentrep/opentrep/service/OPENTREP_Service.cpp 2009-07-17 17:19:13 UTC (rev 134)
@@ -2,11 +2,13 @@
// Import section
// //////////////////////////////////////////////////////////////////////
// C
-#include <assert.h>
+#include <cassert>
// OpenTrep
#include <opentrep/basic/BasConst_OPENTREP_Service.hpp>
#include <opentrep/basic/BasChronometer.hpp>
+#include <opentrep/bom/PlaceHolder.hpp>
#include <opentrep/factory/FacWorld.hpp>
+#include <opentrep/factory/FacPlaceHolder.hpp>
#include <opentrep/command/SociSessionManager.hpp>
#include <opentrep/command/DBManager.hpp>
#include <opentrep/command/IndexBuilder.hpp>
@@ -19,12 +21,21 @@
namespace OPENTREP {
// //////////////////////////////////////////////////////////////////////
+ OPENTREP_Service::OPENTREP_Service (std::ostream& ioLogStream,
+ const std::string& iXapianDatabaseFilepath)
+ : _opentrepSer...
[truncated message content] |
|
From: <den...@us...> - 2009-07-17 00:10:41
|
Revision: 133
http://opentrep.svn.sourceforge.net/opentrep/?rev=133&view=rev
Author: denis_arnaud
Date: 2009-07-17 00:10:38 +0000 (Fri, 17 Jul 2009)
Log Message:
-----------
[Dev] Retrofitted the code of test/xapian/string_search.cpp into the OpenTREP library, and created the corresponding objects and support classes.
Modified Paths:
--------------
trunk/opentrep/opentrep/batches/indexer.cpp
trunk/opentrep/opentrep/batches/searcher.cpp
trunk/opentrep/opentrep/bom/BomAbstract.hpp
trunk/opentrep/opentrep/bom/sources.mk
trunk/opentrep/opentrep/command/RequestInterpreter.cpp
Added Paths:
-----------
trunk/opentrep/opentrep/bom/DocumentList.hpp
trunk/opentrep/opentrep/bom/Result.cpp
trunk/opentrep/opentrep/bom/Result.hpp
trunk/opentrep/opentrep/bom/ResultHolder.cpp
trunk/opentrep/opentrep/bom/ResultHolder.hpp
trunk/opentrep/opentrep/bom/ResultList.hpp
trunk/opentrep/opentrep/bom/StringMatcher.cpp
trunk/opentrep/opentrep/bom/StringMatcher.hpp
trunk/opentrep/opentrep/bom/WordHolder.cpp
trunk/opentrep/opentrep/bom/WordHolder.hpp
trunk/opentrep/opentrep/bom/WordList.hpp
Modified: trunk/opentrep/opentrep/batches/indexer.cpp
===================================================================
--- trunk/opentrep/opentrep/batches/indexer.cpp 2009-07-16 16:45:48 UTC (rev 132)
+++ trunk/opentrep/opentrep/batches/indexer.cpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -1,153 +1,153 @@
-// C
-#include <assert.h>
-// STL
-#include <iostream>
-#include <sstream>
-#include <fstream>
-#include <map>
-#include <vector>
-// Boost (Extended STL)
-#include <boost/date_time/posix_time/posix_time.hpp>
-#include <boost/date_time/gregorian/gregorian.hpp>
-#include <boost/program_options.hpp>
-// OPENTREP
-#include <opentrep/OPENTREP_Service.hpp>
-
-// ///////// Parsing of Options & Configuration /////////
-// A helper function to simplify the main part.
-template<class T> std::ostream& operator<< (std::ostream& os,
- const std::vector<T>& v) {
- std::copy (v.begin(), v.end(), std::ostream_iterator<T> (std::cout, " "));
- return os;
-}
-
-int readConfiguration (int argc, char* argv[]) {
- int opt;
-
- // Declare a group of options that will be
- // allowed only on command line
- boost::program_options::options_description generic("Generic options");
- generic.add_options()
- ("version,v", "print version string")
- ("help,h", "produce help message");
-
- // Declare a group of options that will be allowed both on command line and in
- // config file
- boost::program_options::options_description config("Configuration");
- config.add_options()
- ("optimization",
- boost::program_options::value<int>(&opt)->default_value(10),
- "optimization level")
- ("include-path,I",
- boost::program_options::value< std::vector<std::string> >()->composing(),
- "include path");
-
- // Hidden options, will be allowed both on command line and
- // in config file, but will not be shown to the user.
- boost::program_options::options_description hidden("Hidden options");
- hidden.add_options()
- ("input-file",
- boost::program_options::value< std::vector<std::string> >(),
- "input file");
-
- boost::program_options::options_description cmdline_options;
- cmdline_options.add(generic).add(config).add(hidden);
-
- boost::program_options::options_description config_file_options;
- config_file_options.add(config).add(hidden);
-
- boost::program_options::options_description visible("Allowed options");
- visible.add(generic).add(config);
-
- boost::program_options::positional_options_description p;
- p.add("input-file", -1);
-
- boost::program_options::variables_map vm;
- boost::program_options::
- store (boost::program_options::command_line_parser(argc, argv).
- options (cmdline_options).positional(p).run(), vm);
-
- std::ifstream ifs ("request_parser.cfg");
- boost::program_options::store (parse_config_file (ifs, config_file_options),
- vm);
- boost::program_options::notify (vm);
-
- if (vm.count ("help")) {
- std::cout << visible << std::endl;
- return 0;
- }
-
- if (vm.count ("version")) {
- std::cout << "Open Travel Request Parser, version 1.0" << std::endl;
- return 0;
- }
-
- if (vm.count ("include-path")) {
- std::cout << "Include paths are: "
- << vm["include-path"].as< std::vector<std::string> >()
- << std::endl;
- }
-
- if (vm.count ("input-file")) {
- std::cout << "Input files are: "
- << vm["input-file"].as< std::vector<std::string> >()
- << std::endl;
- }
-
- std::cout << "Optimization level is " << opt << std::endl;
-
- return 0;
-}
-
-
-// /////////////// M A I N /////////////////
-int main (int argc, char* argv[]) {
- try {
-
- // Output log File
- std::string lLogFilename ("indexer.log");
-
- // Xapian database name (directory of the index)
- OPENTREP::TravelDatabaseName_T lXapianDatabaseName ("traveldb");
-
- if (argc >= 1 && argv[1] != NULL) {
- std::istringstream istr (argv[1]);
- istr >> lLogFilename;
- }
-
- if (argc >= 2 && argv[2] != NULL) {
- std::istringstream istr (argv[2]);
- istr >> lXapianDatabaseName;
- }
-
- // Set the log parameters
- std::ofstream logOutputFile;
- // open and clean the log outputfile
- logOutputFile.open (lLogFilename.c_str());
- logOutputFile.clear();
-
- // Initialise the context
- OPENTREP::OPENTREP_Service opentrepService;
- opentrepService.init (logOutputFile, lXapianDatabaseName);
-
- // Launch the indexation
- opentrepService.buildSearchIndex();
-
- // Close the Log outputFile
- logOutputFile.close();
-
-
- } catch (const OPENTREP::RootException& otexp) {
- std::cerr << "Standard exception: " << otexp.what() << std::endl;
- return -1;
-
- } catch (const std::exception& stde) {
- std::cerr << "Standard exception: " << stde.what() << std::endl;
- return -1;
-
- } catch (...) {
- return -1;
- }
-
- return 0;
-}
+// C
+#include <assert.h>
+// STL
+#include <iostream>
+#include <sstream>
+#include <fstream>
+#include <map>
+#include <vector>
+// Boost (Extended STL)
+#include <boost/date_time/posix_time/posix_time.hpp>
+#include <boost/date_time/gregorian/gregorian.hpp>
+#include <boost/program_options.hpp>
+// OPENTREP
+#include <opentrep/OPENTREP_Service.hpp>
+
+// ///////// Parsing of Options & Configuration /////////
+// A helper function to simplify the main part.
+template<class T> std::ostream& operator<< (std::ostream& os,
+ const std::vector<T>& v) {
+ std::copy (v.begin(), v.end(), std::ostream_iterator<T> (std::cout, " "));
+ return os;
+}
+
+int readConfiguration (int argc, char* argv[]) {
+ int opt;
+
+ // Declare a group of options that will be
+ // allowed only on command line
+ boost::program_options::options_description generic("Generic options");
+ generic.add_options()
+ ("version,v", "print version string")
+ ("help,h", "produce help message");
+
+ // Declare a group of options that will be allowed both on command line and in
+ // config file
+ boost::program_options::options_description config("Configuration");
+ config.add_options()
+ ("optimization",
+ boost::program_options::value<int>(&opt)->default_value(10),
+ "optimization level")
+ ("include-path,I",
+ boost::program_options::value< std::vector<std::string> >()->composing(),
+ "include path");
+
+ // Hidden options, will be allowed both on command line and
+ // in config file, but will not be shown to the user.
+ boost::program_options::options_description hidden("Hidden options");
+ hidden.add_options()
+ ("input-file",
+ boost::program_options::value< std::vector<std::string> >(),
+ "input file");
+
+ boost::program_options::options_description cmdline_options;
+ cmdline_options.add(generic).add(config).add(hidden);
+
+ boost::program_options::options_description config_file_options;
+ config_file_options.add(config).add(hidden);
+
+ boost::program_options::options_description visible("Allowed options");
+ visible.add(generic).add(config);
+
+ boost::program_options::positional_options_description p;
+ p.add("input-file", -1);
+
+ boost::program_options::variables_map vm;
+ boost::program_options::
+ store (boost::program_options::command_line_parser(argc, argv).
+ options (cmdline_options).positional(p).run(), vm);
+
+ std::ifstream ifs ("request_parser.cfg");
+ boost::program_options::store (parse_config_file (ifs, config_file_options),
+ vm);
+ boost::program_options::notify (vm);
+
+ if (vm.count ("help")) {
+ std::cout << visible << std::endl;
+ return 0;
+ }
+
+ if (vm.count ("version")) {
+ std::cout << "Open Travel Request Parser, version 1.0" << std::endl;
+ return 0;
+ }
+
+ if (vm.count ("include-path")) {
+ std::cout << "Include paths are: "
+ << vm["include-path"].as< std::vector<std::string> >()
+ << std::endl;
+ }
+
+ if (vm.count ("input-file")) {
+ std::cout << "Input files are: "
+ << vm["input-file"].as< std::vector<std::string> >()
+ << std::endl;
+ }
+
+ std::cout << "Optimization level is " << opt << std::endl;
+
+ return 0;
+}
+
+
+// /////////////// M A I N /////////////////
+int main (int argc, char* argv[]) {
+ try {
+
+ // Output log File
+ std::string lLogFilename ("indexer.log");
+
+ // Xapian database name (directory of the index)
+ OPENTREP::TravelDatabaseName_T lXapianDatabaseName ("traveldb");
+
+ if (argc >= 1 && argv[1] != NULL) {
+ std::istringstream istr (argv[1]);
+ istr >> lLogFilename;
+ }
+
+ if (argc >= 2 && argv[2] != NULL) {
+ std::istringstream istr (argv[2]);
+ istr >> lXapianDatabaseName;
+ }
+
+ // Set the log parameters
+ std::ofstream logOutputFile;
+ // open and clean the log outputfile
+ logOutputFile.open (lLogFilename.c_str());
+ logOutputFile.clear();
+
+ // Initialise the context
+ OPENTREP::OPENTREP_Service opentrepService;
+ opentrepService.init (logOutputFile, lXapianDatabaseName);
+
+ // Launch the indexation
+ opentrepService.buildSearchIndex();
+
+ // Close the Log outputFile
+ logOutputFile.close();
+
+
+ } catch (const OPENTREP::RootException& otexp) {
+ std::cerr << "Standard exception: " << otexp.what() << std::endl;
+ return -1;
+
+ } catch (const std::exception& stde) {
+ std::cerr << "Standard exception: " << stde.what() << std::endl;
+ return -1;
+
+ } catch (...) {
+ return -1;
+ }
+
+ return 0;
+}
Modified: trunk/opentrep/opentrep/batches/searcher.cpp
===================================================================
--- trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-16 16:45:48 UTC (rev 132)
+++ trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -1,160 +1,160 @@
-// C
-#include <assert.h>
-// STL
-#include <iostream>
-#include <sstream>
-#include <fstream>
-#include <map>
-#include <vector>
-// Boost (Extended STL)
-#include <boost/date_time/posix_time/posix_time.hpp>
-#include <boost/date_time/gregorian/gregorian.hpp>
-#include <boost/program_options.hpp>
-// OPENTREP
-#include <opentrep/OPENTREP_Service.hpp>
-
-// ///////// Parsing of Options & Configuration /////////
-// A helper function to simplify the main part.
-template<class T> std::ostream& operator<< (std::ostream& os,
- const std::vector<T>& v) {
- std::copy (v.begin(), v.end(), std::ostream_iterator<T> (std::cout, " "));
- return os;
-}
-
-int readConfiguration (int argc, char* argv[]) {
- int opt;
-
- // Declare a group of options that will be
- // allowed only on command line
- boost::program_options::options_description generic("Generic options");
- generic.add_options()
- ("version,v", "print version string")
- ("help,h", "produce help message");
-
- // Declare a group of options that will be allowed both on command line and in
- // config file
- boost::program_options::options_description config("Configuration");
- config.add_options()
- ("optimization",
- boost::program_options::value<int>(&opt)->default_value(10),
- "optimization level")
- ("include-path,I",
- boost::program_options::value< std::vector<std::string> >()->composing(),
- "include path");
-
- // Hidden options, will be allowed both on command line and
- // in config file, but will not be shown to the user.
- boost::program_options::options_description hidden("Hidden options");
- hidden.add_options()
- ("input-file",
- boost::program_options::value< std::vector<std::string> >(),
- "input file");
-
- boost::program_options::options_description cmdline_options;
- cmdline_options.add(generic).add(config).add(hidden);
-
- boost::program_options::options_description config_file_options;
- config_file_options.add(config).add(hidden);
-
- boost::program_options::options_description visible("Allowed options");
- visible.add(generic).add(config);
-
- boost::program_options::positional_options_description p;
- p.add("input-file", -1);
-
- boost::program_options::variables_map vm;
- boost::program_options::
- store (boost::program_options::command_line_parser(argc, argv).
- options (cmdline_options).positional(p).run(), vm);
-
- std::ifstream ifs ("request_parser.cfg");
- boost::program_options::store (parse_config_file (ifs, config_file_options),
- vm);
- boost::program_options::notify (vm);
-
- if (vm.count ("help")) {
- std::cout << visible << std::endl;
- return 0;
- }
-
- if (vm.count ("version")) {
- std::cout << "Open Travel Request Parser, version 1.0" << std::endl;
- return 0;
- }
-
- if (vm.count ("include-path")) {
- std::cout << "Include paths are: "
- << vm["include-path"].as< std::vector<std::string> >()
- << std::endl;
- }
-
- if (vm.count ("input-file")) {
- std::cout << "Input files are: "
- << vm["input-file"].as< std::vector<std::string> >()
- << std::endl;
- }
-
- std::cout << "Optimization level is " << opt << std::endl;
-
- return 0;
-}
-
-
-// /////////////// M A I N /////////////////
-int main (int argc, char* argv[]) {
- try {
-
- // Travel query
- OPENTREP::TravelQuery_T lTravelQuery ("cdg");
-
- // Output log File
- std::string lLogFilename ("searcher.log");
-
- // Xapian database name (directory of the index)
- OPENTREP::TravelDatabaseName_T lXapianDatabaseName ("traveldb");
-
- if (argc >= 1 && argv[1] != NULL) {
- std::istringstream istr (argv[1]);
- istr >> lTravelQuery;
- }
-
- if (argc >= 2 && argv[2] != NULL) {
- std::istringstream istr (argv[2]);
- istr >> lLogFilename;
- }
-
- if (argc >= 3 && argv[3] != NULL) {
- std::istringstream istr (argv[3]);
- istr >> lXapianDatabaseName;
- }
-
- // Set the log parameters
- std::ofstream logOutputFile;
- // open and clean the log outputfile
- logOutputFile.open (lLogFilename.c_str());
- logOutputFile.clear();
-
- // Initialise the context
- OPENTREP::OPENTREP_Service opentrepService;
- opentrepService.init (logOutputFile, lXapianDatabaseName);
-
- // Query the Xapian database (index)
- opentrepService.interpretTravelRequest (lTravelQuery);
-
- // Close the Log outputFile
- logOutputFile.close();
-
- } catch (const OPENTREP::RootException& otexp) {
- std::cerr << "Standard exception: " << otexp.what() << std::endl;
- return -1;
-
- } catch (const std::exception& stde) {
- std::cerr << "Standard exception: " << stde.what() << std::endl;
- return -1;
-
- } catch (...) {
- return -1;
- }
-
- return 0;
-}
+// C
+#include <assert.h>
+// STL
+#include <iostream>
+#include <sstream>
+#include <fstream>
+#include <map>
+#include <vector>
+// Boost (Extended STL)
+#include <boost/date_time/posix_time/posix_time.hpp>
+#include <boost/date_time/gregorian/gregorian.hpp>
+#include <boost/program_options.hpp>
+// OPENTREP
+#include <opentrep/OPENTREP_Service.hpp>
+
+// ///////// Parsing of Options & Configuration /////////
+// A helper function to simplify the main part.
+template<class T> std::ostream& operator<< (std::ostream& os,
+ const std::vector<T>& v) {
+ std::copy (v.begin(), v.end(), std::ostream_iterator<T> (std::cout, " "));
+ return os;
+}
+
+int readConfiguration (int argc, char* argv[]) {
+ int opt;
+
+ // Declare a group of options that will be
+ // allowed only on command line
+ boost::program_options::options_description generic("Generic options");
+ generic.add_options()
+ ("version,v", "print version string")
+ ("help,h", "produce help message");
+
+ // Declare a group of options that will be allowed both on command line and in
+ // config file
+ boost::program_options::options_description config("Configuration");
+ config.add_options()
+ ("optimization",
+ boost::program_options::value<int>(&opt)->default_value(10),
+ "optimization level")
+ ("include-path,I",
+ boost::program_options::value< std::vector<std::string> >()->composing(),
+ "include path");
+
+ // Hidden options, will be allowed both on command line and
+ // in config file, but will not be shown to the user.
+ boost::program_options::options_description hidden("Hidden options");
+ hidden.add_options()
+ ("input-file",
+ boost::program_options::value< std::vector<std::string> >(),
+ "input file");
+
+ boost::program_options::options_description cmdline_options;
+ cmdline_options.add(generic).add(config).add(hidden);
+
+ boost::program_options::options_description config_file_options;
+ config_file_options.add(config).add(hidden);
+
+ boost::program_options::options_description visible("Allowed options");
+ visible.add(generic).add(config);
+
+ boost::program_options::positional_options_description p;
+ p.add("input-file", -1);
+
+ boost::program_options::variables_map vm;
+ boost::program_options::
+ store (boost::program_options::command_line_parser(argc, argv).
+ options (cmdline_options).positional(p).run(), vm);
+
+ std::ifstream ifs ("request_parser.cfg");
+ boost::program_options::store (parse_config_file (ifs, config_file_options),
+ vm);
+ boost::program_options::notify (vm);
+
+ if (vm.count ("help")) {
+ std::cout << visible << std::endl;
+ return 0;
+ }
+
+ if (vm.count ("version")) {
+ std::cout << "Open Travel Request Parser, version 1.0" << std::endl;
+ return 0;
+ }
+
+ if (vm.count ("include-path")) {
+ std::cout << "Include paths are: "
+ << vm["include-path"].as< std::vector<std::string> >()
+ << std::endl;
+ }
+
+ if (vm.count ("input-file")) {
+ std::cout << "Input files are: "
+ << vm["input-file"].as< std::vector<std::string> >()
+ << std::endl;
+ }
+
+ std::cout << "Optimization level is " << opt << std::endl;
+
+ return 0;
+}
+
+
+// /////////////// M A I N /////////////////
+int main (int argc, char* argv[]) {
+ try {
+
+ // Travel query
+ OPENTREP::TravelQuery_T lTravelQuery ("sna francisco rio de janero lso angeles");
+
+ // Output log File
+ std::string lLogFilename ("searcher.log");
+
+ // Xapian database name (directory of the index)
+ OPENTREP::TravelDatabaseName_T lXapianDatabaseName ("traveldb");
+
+ if (argc >= 1 && argv[1] != NULL) {
+ std::istringstream istr (argv[1]);
+ istr >> lTravelQuery;
+ }
+
+ if (argc >= 2 && argv[2] != NULL) {
+ std::istringstream istr (argv[2]);
+ istr >> lLogFilename;
+ }
+
+ if (argc >= 3 && argv[3] != NULL) {
+ std::istringstream istr (argv[3]);
+ istr >> lXapianDatabaseName;
+ }
+
+ // Set the log parameters
+ std::ofstream logOutputFile;
+ // open and clean the log outputfile
+ logOutputFile.open (lLogFilename.c_str());
+ logOutputFile.clear();
+
+ // Initialise the context
+ OPENTREP::OPENTREP_Service opentrepService;
+ opentrepService.init (logOutputFile, lXapianDatabaseName);
+
+ // Query the Xapian database (index)
+ opentrepService.interpretTravelRequest (lTravelQuery);
+
+ // Close the Log outputFile
+ logOutputFile.close();
+
+ } catch (const OPENTREP::RootException& otexp) {
+ std::cerr << "Standard exception: " << otexp.what() << std::endl;
+ return -1;
+
+ } catch (const std::exception& stde) {
+ std::cerr << "Standard exception: " << stde.what() << std::endl;
+ return -1;
+
+ } catch (...) {
+ return -1;
+ }
+
+ return 0;
+}
Modified: trunk/opentrep/opentrep/bom/BomAbstract.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/BomAbstract.hpp 2009-07-16 16:45:48 UTC (rev 132)
+++ trunk/opentrep/opentrep/bom/BomAbstract.hpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -1,5 +1,5 @@
-#ifndef __OPENTREP_BOMABSTRACT_HPP
-#define __OPENTREP_BOMABSTRACT_HPP
+#ifndef __OPENTREP_BOM_BOMABSTRACT_HPP
+#define __OPENTREP_BOM_BOMABSTRACT_HPP
// //////////////////////////////////////////////////////////////////////
// Import section
@@ -92,4 +92,4 @@
return ioIn;
}
-#endif // __OPENTREP_BOMABSTRACT_HPP
+#endif // __OPENTREP_BOM_BOMABSTRACT_HPP
Added: trunk/opentrep/opentrep/bom/DocumentList.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/DocumentList.hpp (rev 0)
+++ trunk/opentrep/opentrep/bom/DocumentList.hpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -0,0 +1,18 @@
+#ifndef __OPENTREP_BOM_DOCUMENTLIST_HPP
+#define __OPENTREP_BOM_DOCUMENTLIST_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// STL
+#include <map>
+// Xapian
+#include <xapian.h>
+
+namespace OPENTREP {
+
+ /** List of Xapian documents. */
+ typedef std::multimap<Xapian::percent, Xapian::Document> DocumentList_T;
+
+}
+#endif // __OPENTREP_BOM_DOCUMENTLIST_HPP
Added: trunk/opentrep/opentrep/bom/Result.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/Result.cpp (rev 0)
+++ trunk/opentrep/opentrep/bom/Result.cpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -0,0 +1,131 @@
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// C
+#include <cassert>
+#include <string>
+#include <sstream>
+// OPENTREP
+#include <opentrep/bom/StringMatcher.hpp>
+#include <opentrep/bom/Result.hpp>
+#include <opentrep/service/Logger.hpp>
+
+namespace OPENTREP {
+
+ // //////////////////////////////////////////////////////////////////////
+ Result::Result (const Xapian::Database& iDatabase)
+ : _database (iDatabase) {
+ init();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ Result::~Result () {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void Result::init () {
+ _documentList.clear();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ const std::string Result::describeShortKey() const {
+ std::ostringstream oStr;
+ oStr << _queryString;
+ return oStr.str();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ const std::string Result::describeKey() const {
+ return describeShortKey();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ std::string Result::toString() const {
+ std::ostringstream oStr;
+ oStr << describeShortKey() << std::endl;
+
+ for (DocumentList_T::const_iterator itDoc = _documentList.begin();
+ itDoc != _documentList.end(); ++itDoc) {
+ const Xapian::percent& lPercent = itDoc->first;
+ const Xapian::Document& lDocument = itDoc->second;
+ const Xapian::docid& lDocID = lDocument.get_docid();
+ oStr << "Document ID " << lDocID << "\t" << lPercent
+ << "% [" << lDocument.get_data() << "]" << std::endl;
+ }
+
+ return oStr.str();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void Result::toStream (std::ostream& ioOut) const {
+ ioOut << toString();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void Result::fromStream (std::istream& ioIn) {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ const Xapian::Document& Result::getBestMatchingDocument() const {
+ /**
+ Retrieve the best matching document. As the document list (STL map)
+ is sorted by ascending order of the matching percentage, the best
+ matching one is located at the end (back) of the list (STL map).
+ */
+ DocumentList_T::const_reverse_iterator itDocument = _documentList.rbegin();
+ return itDocument->second;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ const Xapian::percent& Result::getBestMatchingPercentage() const {
+ /**
+ Retrieve the best matching document. As the document list (STL map)
+ is sorted by ascending order of the matching percentage, the best
+ matching one is located at the end (back) of the list (STL map).
+ */
+ DocumentList_T::const_reverse_iterator itDocument = _documentList.rbegin();
+ return itDocument->first;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void Result::searchString () {
+
+ // Catch any Xapian::Error exceptions thrown
+ try {
+
+ bool shouldStop = false;
+ while (shouldStop == false) {
+ // DEBUG
+ /*
+ OPENTREP_LOG_DEBUG (std::endl << "--------------------------------"
+ << std::endl << "Current query string: `" << ioQueryString << "'");
+ */
+
+ // Retrieve the list of documents matching the query string
+ Xapian::MSet lMatchingSet;
+ StringMatcher::searchString (lMatchingSet, _queryString, _database);
+
+ // Create the corresponding list of documents
+ StringMatcher::createDocumentListFromMSet (lMatchingSet, _documentList);
+
+ // Stop if a result is found.
+ if (_documentList.empty() == false) {
+ shouldStop = true;
+ break;
+ }
+
+ // Remove a word from the query string
+ StringMatcher::removeOneWord (_queryString);
+
+ // Stop when the resulting string gets empty.
+ if (_queryString.empty() == true) {
+ shouldStop = true;
+ }
+ }
+
+ } catch (const Xapian::Error& error) {
+ OPENTREP_LOG_ERROR ("Exception: " << error.get_msg());
+ }
+ }
+
+}
Added: trunk/opentrep/opentrep/bom/Result.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/Result.hpp (rev 0)
+++ trunk/opentrep/opentrep/bom/Result.hpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -0,0 +1,100 @@
+#ifndef __OPENTREP_BOM_RESULT_HPP
+#define __OPENTREP_BOM_RESULT_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// OpenTREP
+#include <opentrep/OPENTREP_Types.hpp>
+#include <opentrep/bom/BomAbstract.hpp>
+#include <opentrep/bom/DocumentList.hpp>
+
+namespace OPENTREP {
+
+ /** Class wrapping functions on a list of Xapian Document objects. */
+ class Result : public BomAbstract {
+ friend class FacResult;
+ friend class ResultHolder;
+ public:
+ // ////////////// Getters /////////////
+ /** Get the query string. */
+ const TravelQuery_T& getQueryString () const {
+ return _queryString;
+ }
+
+ /** Get the list of Xapian document objects. */
+ const DocumentList_T& getDocumentList() const {
+ return _documentList;
+ }
+
+ /** Retrieve the best matching Xapian document object. */
+ const Xapian::Document& getBestMatchingDocument() const;
+
+ /** Retrieve the percentage corresponding to the best matching
+ Xapian document object. */
+ const Xapian::percent& getBestMatchingPercentage() const;
+
+
+ // ////////////// Setters /////////////
+ /** Set the query string. */
+ void setQueryString (const TravelQuery_T& iQueryString) {
+ _queryString = iQueryString;
+ }
+
+
+ public:
+ // /////////// Business methods /////////
+ /** Retrieve the list of documents matching the query string. */
+ void searchString ();
+
+
+ public:
+ // /////////// Display support methods /////////
+ /** Dump a Business Object into an output stream.
+ @param ostream& the output stream. */
+ void toStream (std::ostream& ioOut) const;
+
+ /** Read a Business Object from an input stream.
+ @param istream& the input stream. */
+ void fromStream (std::istream& ioIn);
+
+ /** Get the serialised version of the Business Object. */
+ std::string toString() const;
+
+ /** Get a string describing the whole key (differentiating two objects
+ at any level). */
+ const std::string describeKey() const;
+
+ /** Get a string describing the short key (differentiating two objects
+ at the same level). */
+ const std::string describeShortKey() const;
+
+
+ private:
+ // ////////////// Constructors and Destructors /////////////
+ /** Main constructor. */
+ Result (const Xapian::Database&);
+ /** Default constructor. */
+ Result ();
+ /** Default copy constructor. */
+ Result (const Result&);
+ /** Destructor. */
+ ~Result ();
+ /** Initialise (reset the list of documents). */
+ void init ();
+
+
+ private:
+ // /////////////// Attributes ////////////////
+ /** Query string having generated the list of document. */
+ TravelQuery_T _queryString;
+
+ /** Xapian database. */
+ const Xapian::Database& _database;
+
+ /** List of Xapian document objects. */
+ DocumentList_T _documentList;
+ };
+
+}
+#endif // __OPENTREP_BOM_RESULT_HPP
Added: trunk/opentrep/opentrep/bom/ResultHolder.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/ResultHolder.cpp (rev 0)
+++ trunk/opentrep/opentrep/bom/ResultHolder.cpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -0,0 +1,133 @@
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// C
+#include <cassert>
+#include <sstream>
+// Xapian
+#include <xapian.h>
+// OpenTREP
+#include <opentrep/bom/StringMatcher.hpp>
+#include <opentrep/bom/Result.hpp>
+#include <opentrep/bom/ResultHolder.hpp>
+#include <opentrep/service/Logger.hpp>
+
+namespace OPENTREP {
+
+ // //////////////////////////////////////////////////////////////////////
+ ResultHolder::ResultHolder (const TravelQuery_T& iQueryString,
+ const Xapian::Database& iDatabase)
+ : _queryString (iQueryString), _database (iDatabase) {
+ init();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ ResultHolder::~ResultHolder () {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void ResultHolder::init () {
+ _resultList.clear();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ const std::string ResultHolder::describeShortKey() const {
+ std::ostringstream oStr;
+ oStr << _queryString;
+ return oStr.str();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ const std::string ResultHolder::describeKey() const {
+ return describeShortKey();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ std::string ResultHolder::toString() const {
+ std::ostringstream oStr;
+ oStr << describeShortKey() << std::endl;
+
+ for (ResultList_T::const_iterator itResult = _resultList.begin();
+ itResult != _resultList.end(); ++itResult) {
+ const Result* lResult_ptr = *itResult;
+ assert (lResult_ptr != NULL);
+
+ oStr << " ==> " << std::endl << lResult_ptr->toString();
+ }
+
+ return oStr.str();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void ResultHolder::toStream (std::ostream& ioOut) const {
+ ioOut << toString();
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void ResultHolder::fromStream (std::istream& ioIn) {
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void ResultHolder::searchString () {
+
+ // Catch any Xapian::Error exceptions thrown
+ try {
+
+ std::string lRemainingQueryString (_queryString);
+ bool shouldStop = false;
+ while (shouldStop == false) {
+ // DEBUG
+ /*
+ OPENTREP_LOG_DEBUG (std::endl
+ << "================================" << std::endl
+ << "Current query string: `" << lRemainingQueryString << "'");
+ */
+ /**
+ Search with the initial full string, then by removing a word if
+ there was no result, then by removing another word if there was
+ again no result, until either a result is found or the
+ resulting string gets empty.
+ */
+ DocumentList_T lDocumentList;
+ Result* lResult_ptr = new Result (_database);
+ assert (lResult_ptr != NULL);
+
+ std::string lQueryString (lRemainingQueryString);
+
+ //
+ lResult_ptr->setQueryString (lQueryString);
+ lResult_ptr->searchString ();
+
+ // Add the Result object (holding the list of matching
+ // documents) to the dedicated list.
+ _resultList.push_back (lResult_ptr);
+
+ /**
+ Remove, from the lRemainingQueryString string, the part which
+ has been already successfully parsed.
+ <br>For instance, when 'sna francisco rio de janeiro' is the
+ initial full clean query string, the searchString() method
+ first reduce the query string to 'sna francisco', which
+ successfully matches against SFO (San Francisco airport).
+ <br>Then, the remaining part of the query string to be parsed is
+ 'rio de janeiro'. So, the already parsed part, namely
+ 'sna francisco', must be subtracted from the initial query string.
+ */
+ lQueryString = lResult_ptr->getQueryString();
+ StringMatcher::subtractParsedToRemaining (lQueryString,
+ lRemainingQueryString);
+
+ // If there is nothing left to be parsed, we have then finished
+ // to parse the initial string.
+ if (lRemainingQueryString.empty() == true) {
+ shouldStop = true;
+ break;
+ }
+ }
+
+ } catch (const Xapian::Error& error) {
+ OPENTREP_LOG_ERROR ("Exception: " << error.get_msg());
+ }
+ }
+
+}
Added: trunk/opentrep/opentrep/bom/ResultHolder.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/ResultHolder.hpp (rev 0)
+++ trunk/opentrep/opentrep/bom/ResultHolder.hpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -0,0 +1,94 @@
+#ifndef __OPENTREP_BOM_RESULTHOLDER_HPP
+#define __OPENTREP_BOM_RESULTHOLDER_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// OpenTREP
+#include <opentrep/OPENTREP_Types.hpp>
+#include <opentrep/bom/BomAbstract.hpp>
+#include <opentrep/bom/ResultList.hpp>
+
+// Forward declarations
+namespace Xapian {
+ class Database;
+}
+
+namespace OPENTREP {
+
+ /** Class wrapping functions on a list of Result objects. */
+ class ResultHolder : public BomAbstract {
+ friend class FacResultHolder;
+ friend class RequestInterpreter;
+ public:
+ // ////////////// Getters /////////////
+ /** Get the query string. */
+ const TravelQuery_T& getQueryString () const {
+ return _queryString;
+ }
+
+ /** Retrieve the list of result objects. */
+ const ResultList_T& getResultList() const {
+ return _resultList;
+ }
+
+
+ // ////////////// Setters /////////////
+
+
+ public:
+ // /////////// Business methods /////////
+ /** Retrieve the list of documents matching the query string. */
+ void searchString ();
+
+
+ public:
+ // /////////// Display support methods /////////
+ /** Dump a Business Object into an output stream.
+ @param ostream& the output stream. */
+ void toStream (std::ostream& ioOut) const;
+
+ /** Read a Business Object from an input stream.
+ @param istream& the input stream. */
+ void fromStream (std::istream& ioIn);
+
+ /** Get the serialised version of the Business Object. */
+ std::string toString() const;
+
+ /** Get a string describing the whole key (differentiating two objects
+ at any level). */
+ const std::string describeKey() const;
+
+ /** Get a string describing the short key (differentiating two objects
+ at the same level). */
+ const std::string describeShortKey() const;
+
+
+ private:
+ // ////////////// Constructors and Destructors /////////////
+ /** Main constructor. */
+ ResultHolder (const TravelQuery_T&, const Xapian::Database&);
+ /** Default constructor. */
+ ResultHolder ();
+ /** Default copy constructor. */
+ ResultHolder (const ResultHolder&);
+ /** Destructor. */
+ ~ResultHolder ();
+ /** Initialise (reset the list of documents). */
+ void init ();
+
+
+ private:
+ // /////////////// Attributes ////////////////
+ /** Query string having generated the list of document. */
+ const TravelQuery_T _queryString;
+
+ /** Xapian database. */
+ const Xapian::Database& _database;
+
+ /** List of result objects. */
+ ResultList_T _resultList;
+ };
+
+}
+#endif // __OPENTREP_BOM_RESULTHOLDER_HPP
Added: trunk/opentrep/opentrep/bom/ResultList.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/ResultList.hpp (rev 0)
+++ trunk/opentrep/opentrep/bom/ResultList.hpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -0,0 +1,19 @@
+#ifndef __OPENTREP_BOM_RESULTLIST_HPP
+#define __OPENTREP_BOM_RESULTLIST_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// STL
+#include <list>
+
+namespace OPENTREP {
+
+ // Forward declarations
+ class Result;
+
+ /** List of result objects. */
+ typedef std::list<Result*> ResultList_T;
+
+}
+#endif // __OPENTREP_BOM_RESULTLIST_HPP
Added: trunk/opentrep/opentrep/bom/StringMatcher.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/StringMatcher.cpp (rev 0)
+++ trunk/opentrep/opentrep/bom/StringMatcher.cpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -0,0 +1,319 @@
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// C
+#include <cassert>
+// STL
+#include <iostream>
+#include <sstream>
+#include <string>
+#include <list>
+#include <map>
+// Boost
+#include <boost/tokenizer.hpp>
+// Xapian
+#include <xapian.h>
+// OpenTREP
+#include <opentrep/bom/WordHolder.hpp>
+#include <opentrep/bom/StringMatcher.hpp>
+#include <opentrep/service/Logger.hpp>
+
+namespace OPENTREP {
+
+ /** For each of the word in the given list, perform spelling
+ corrections. If the word is correctly spelled, it is copied as
+ is. Otherwise, a corrected version is stored. */
+ // //////////////////////////////////////////////////////////////////////
+ static void createCorrectedWordList (const WordList_T& iOriginalWordList,
+ WordList_T& ioCorrectedWordList,
+ const Xapian::Database& iDatabase) {
+ // Empty the target list
+ ioCorrectedWordList.clear();
+
+ // Catch any Xapian::Error exceptions thrown
+ try {
+
+ for (WordList_T::const_iterator itWord = iOriginalWordList.begin();
+ itWord != iOriginalWordList.end(); ++itWord) {
+ const std::string& lOriginalWord = *itWord;
+ const std::string& lSuggestedWord =
+ iDatabase.get_spelling_suggestion (lOriginalWord, 3);
+
+ if (lSuggestedWord.empty() == true) {
+ ioCorrectedWordList.push_back (lOriginalWord);
+
+ } else {
+ ioCorrectedWordList.push_back (lSuggestedWord);
+ }
+
+ // DEBUG
+ /*
+ OPENTREP_LOG_DEBUG ("Original word: `" << lOriginalWord
+ << "' ==> corrected word: `" << lSuggestedWord << "'");
+ */
+ }
+
+ } catch (const Xapian::Error& error) {
+ std::cerr << "Exception: " << error.get_msg() << std::endl;
+ }
+ }
+
+ // ///////////////////////////////////////////////////////////////////
+ void StringMatcher::searchString (Xapian::MSet& ioMatchingSet,
+ const std::string& iSearchString,
+ const Xapian::Database& ioDatabase) {
+
+ // Catch any Xapian::Error exceptions thrown
+ try {
+
+ /**
+ Build another string, in addition to the original one. Overall,
+ there are thus two strings:
+ <br><ul>
+ <li>One with the original words given by the user</li>
+ <li>One with the orthographic-corrected words, wherever
+ relevant (otherwise, the original word is taken)</li>
+ </ul>
+ <br>For instance, 'sna francisco' would give the following
+ two strings:
+ <br><ul>
+ <li>'sna francicso' (original)</li>
+ <li>'sna francisco' (corrected, where relevant, word by word)</li>
+ </ul>
+ <br>Note that, as 'sna' exists in the dictionary (Santa Ana, CA, USA),
+ it is not replaced. We shall take care of the whole string in a
+ further step below.
+ */
+ WordList_T lOriginalWordList;
+ WordHolder::tokeniseStringIntoWordList (iSearchString, lOriginalWordList);
+
+ /**
+ We rebuild a clean query string from the word list. Indeed, the original
+ string may have contained a few separators (e.g., '/', ';', etc.), which
+ have been removed by the tokeniseStringIntoWordList() method. All those
+ separators are thus replaced by spaces.
+ For instance, the 'san francisco, ca, us' initial string would be
+ replaced by 'san francisco ca us'.
+ */
+ const std::string lOriginalQueryString =
+ WordHolder::createStringFromWordList (lOriginalWordList);
+
+ WordList_T lCorrectedWordList;
+ createCorrectedWordList (lOriginalWordList, lCorrectedWordList,
+ ioDatabase);
+
+ const std::string lCorrectedQueryString =
+ WordHolder::createStringFromWordList (lCorrectedWordList);
+
+ /**
+ Try to find, if relevant, an orthographic suggestion for the whole
+ phrase/string. With the above example, 'sna francisco' yields the
+ suggestion 'san francisco'.
+ */
+ const std::string lFullWordCorrectedString =
+ ioDatabase.get_spelling_suggestion (lOriginalQueryString, 3);
+
+ // DEBUG
+ /*
+ OPENTREP_LOG_DEBUG ("Query string `" << lOriginalQueryString
+ << "' ==> corrected query string: `" << lCorrectedQueryString
+ << "' and correction for the full query string: `"
+ << lFullWordCorrectedString << "'");
+ */
+
+ // Build the query object
+ Xapian::QueryParser lQueryParser;
+ lQueryParser.set_database (ioDatabase);
+ /**
+ As explained in http://www.xapian.org/docs/queryparser.html,
+ Xapian::Query::OP_ADJ is better than Xapian::Query::OP_PHRASE,
+ but only available from version 1.0.13 of Xapian.
+ */
+ // lQueryParser.set_default_op (Xapian::Query::OP_ADJ);
+ lQueryParser.set_default_op (Xapian::Query::OP_PHRASE);
+
+ // DEBUG
+ /*
+ OPENTREP_LOG_DEBUG ("Query parser `" << lQueryParser.get_description()
+ << "'");
+ */
+
+ /**
+ The Xapian::QueryParser::parse_query() method aggregates all the words
+ with operators inbetween them (here, the "PHRASE" operator).
+ With the above example ('sna francicso'), it yields
+ "sna PHRASE 2 francicso".
+ */
+ Xapian::Query lQuery =
+ lQueryParser.parse_query (lOriginalQueryString,
+ Xapian::QueryParser::FLAG_BOOLEAN
+ | Xapian::QueryParser::FLAG_PHRASE
+ | Xapian::QueryParser::FLAG_LOVEHATE
+ | Xapian::QueryParser::FLAG_SPELLING_CORRECTION);
+ /**
+ Strangely enough (is it?), the corrected query given by the Xapian
+ QueryParser corresponds to the full original string, where words
+ have been corrected one by one, but considered as a single block.
+ With the above example, 'sna francicso' yields (wrongly)
+ 'sna francisco', instead of "sna PHRASE 2 francisco", as generated
+ by the following code.
+ */
+ // Xapian::Query lCorrectedQuery =
+ // lQueryParser.get_corrected_query_string();
+ Xapian::Query lCorrectedQuery =
+ lQueryParser.parse_query (lCorrectedQueryString,
+ Xapian::QueryParser::FLAG_BOOLEAN
+ | Xapian::QueryParser::FLAG_PHRASE
+ | Xapian::QueryParser::FLAG_LOVEHATE);
+
+ /**
+ As, with the above example, the full corrected string is
+ 'san francisco', it yields the query "san PHRASE 2 francisco",
+ which is eventually right.
+ */
+ Xapian::Query lFullQueryCorrected =
+ lQueryParser.parse_query (lFullWordCorrectedString,
+ Xapian::QueryParser::FLAG_BOOLEAN
+ | Xapian::QueryParser::FLAG_PHRASE
+ | Xapian::QueryParser::FLAG_LOVEHATE);
+
+ // DEBUG
+ /*
+ OPENTREP_LOG_DEBUG ("Query `" << lQuery.get_description()
+ << "', corrected query `" << lCorrectedQuery.get_description()
+ << "' and corrected for full query `"
+ << lFullQueryCorrected.get_description() << "' ");
+ */
+
+ // Start an enquire session
+ Xapian::Enquire enquire (ioDatabase);
+
+ // Give the query object to the enquire session
+ enquire.set_query (lQuery);
+
+ // Get the top 10 results of the query
+ ioMatchingSet = enquire.get_mset (0, 10);
+
+ // Display the results
+ int nbMatches = ioMatchingSet.size();
+
+ // DEBUG
+ /*
+ OPENTREP_LOG_DEBUG (nbMatches << " results found");
+ */
+
+ /**
+ When no match is found, we search on the corrected phrase/string
+ (where the words have been corrected one by one).
+ */
+ if (nbMatches == 0) {
+ enquire.set_query (lCorrectedQuery);
+ ioMatchingSet = enquire.get_mset (0, 10);
+
+ // Display the results
+ nbMatches = ioMatchingSet.size();
+
+ // DEBUG
+ /*
+ OPENTREP_LOG_DEBUG(nbMatches << " results found on corrected string");
+ */
+ }
+
+ /**
+ If there is still no match, we search on the string corrected
+ as a whole.
+ */
+ if (nbMatches == 0) {
+ enquire.set_query (lFullQueryCorrected);
+ ioMatchingSet = enquire.get_mset (0, 10);
+
+ // Display the results
+ nbMatches = ioMatchingSet.size();
+
+ // DEBUG
+ /*
+ OPENTREP_LOG_DEBUG (nbMatches
+ << " results found on corrected full string");
+ */
+ }
+
+
+ // DEBUG
+ /*
+ const Xapian::Query& lActualQuery = enquire.get_query();
+ OPENTREP_LOG_DEBUG ("Actual query `" << lActualQuery.get_description()
+ << "'");
+ */
+
+ } catch (const Xapian::Error& error) {
+ std::cerr << "Exception: " << error.get_msg() << std::endl;
+ }
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void StringMatcher::
+ createDocumentListFromMSet (const Xapian::MSet& iMatchingSet,
+ DocumentList_T& ioDocumentList) {
+ // Empty the list of documents
+ ioDocumentList.clear();
+
+ for (Xapian::MSetIterator itDoc = iMatchingSet.begin();
+ itDoc != iMatchingSet.end(); ++itDoc) {
+ const Xapian::Document& lDocument = itDoc.get_document();
+
+ ioDocumentList.insert (DocumentList_T::value_type (itDoc.get_percent(),
+ lDocument));
+ }
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void StringMatcher::removeOneWord (std::string& ioQueryString) {
+ assert (ioQueryString.empty() == false);
+
+ WordList_T lWordList;
+ WordHolder::tokeniseStringIntoWordList (ioQueryString, lWordList);
+ assert (lWordList.empty() == false);
+
+ // Remove the furthest right word
+ lWordList.pop_back();
+
+ const std::string& lReducedString =
+ WordHolder::createStringFromWordList (lWordList);
+ ioQueryString = lReducedString;
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ void StringMatcher::
+ subtractParsedToRemaining (const std::string& iAlreadyParsedQueryString,
+ std::string& ioRemainingQueryString) {
+ /**
+ Remove, from the lRemainingQueryString string, the part which
+ has been already successfully parsed.
+ <br>For instance, when 'sna francisco rio de janeiro' is the
+ initial full clean query string, the searchString() method
+ first reduce the query string to 'sna francisco', which
+ successfully matches against SFO (San Francisco airport).
+ <br>Then, the remaining part of the query string to be parsed is
+ 'rio de janeiro'. So, the already parsed part must be subtracted
+ from the initial query string.
+ */
+ WordList_T lRemainingWordList;
+ WordHolder::tokeniseStringIntoWordList (ioRemainingQueryString,
+ lRemainingWordList);
+
+ WordList_T lParsedWordList;
+ WordHolder::tokeniseStringIntoWordList (iAlreadyParsedQueryString,
+ lParsedWordList);
+
+ unsigned int idx = lParsedWordList.size();
+ for ( ; idx != 0 && lRemainingWordList.empty() == false; --idx) {
+ lRemainingWordList.pop_front();
+ }
+
+ // Build the remaining part of the string still to be parsed.
+ // <br>Note that that part may be empty.
+ ioRemainingQueryString =
+ WordHolder::createStringFromWordList (lRemainingWordList);
+ }
+
+}
Added: trunk/opentrep/opentrep/bom/StringMatcher.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/StringMatcher.hpp (rev 0)
+++ trunk/opentrep/opentrep/bom/StringMatcher.hpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -0,0 +1,52 @@
+#ifndef __OPENTREP_BOM_STRINGMATCHER_HPP
+#define __OPENTREP_BOM_STRINGMATCHER_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// STL
+#include <string>
+// OpenTREP
+#include <opentrep/bom/BomAbstract.hpp>
+#include <opentrep/bom/WordList.hpp>
+#include <opentrep/bom/DocumentList.hpp>
+
+// Forward declarations
+namespace Xapian {
+ class MSet;
+ class Database;
+}
+
+namespace OPENTREP {
+
+ /** Class grouping a few utility methods based on the Xapian library.
+ <br>See <a href="http://www.xapian.org">Xapian's Web site</a>
+ for more information. */
+ class StringMatcher : public BomAbstract {
+ public:
+ /** Search, within the Xapian database, for occurrences of the
+ words of the search string. */
+ static void searchString (Xapian::MSet&, const std::string& iSearchString,
+ const Xapian::Database&);
+
+ /** Copy the Xapian MSet (matching set) object into a document
+ list object. */
+ static void createDocumentListFromMSet (const Xapian::MSet&,
+ DocumentList_T&);
+
+ /** Remove the word furthest at right. */
+ static void removeOneWord (std::string& ioQueryString);
+
+ /** Remove, from a string, the part corresponding to the one given
+ as parameter. */
+ static void
+ subtractParsedToRemaining (const std::string& iAlreadyParsedQueryString,
+ std::string& ioRemainingQueryString);
+
+
+ private:
+
+ };
+
+}
+#endif // __OPENTREP_BOM_STRINGMATCHER_HPP
Added: trunk/opentrep/opentrep/bom/WordHolder.cpp
===================================================================
--- trunk/opentrep/opentrep/bom/WordHolder.cpp (rev 0)
+++ trunk/opentrep/opentrep/bom/WordHolder.cpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -0,0 +1,53 @@
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// C
+#include <cassert>
+// Boost
+#include <boost/tokenizer.hpp>
+// OpenTREP
+#include <opentrep/bom/WordHolder.hpp>
+
+namespace OPENTREP {
+
+ // Define the separators
+ static const boost::char_separator<char> SepatorList (" .,;:|+-*/_=!@#$%`~^&(){}[]?'<>\"");
+
+ // //////////////////////////////////////////////////////////////////////
+ void WordHolder::tokeniseStringIntoWordList (const std::string& iPhrase,
+ WordList_T& ioWordList) {
+ // Empty the word list
+ ioWordList.clear();
+
+ // Boost Tokeniser
+ typedef boost::tokenizer<boost::char_separator<char> > Tokeniser_T;
+
+ // Initialise the phrase to be tokenised
+ Tokeniser_T lTokens (iPhrase, SepatorList);
+ for (Tokeniser_T::const_iterator tok_iter = lTokens.begin();
+ tok_iter != lTokens.end(); ++tok_iter) {
+ const std::string& lTerm = *tok_iter;
+ ioWordList.push_back (lTerm);
+ }
+
+ }
+
+ // //////////////////////////////////////////////////////////////////////
+ std::string WordHolder::
+ createStringFromWordList (const WordList_T& iWordList) {
+ std::ostringstream oStr;
+
+ unsigned short idx = iWordList.size();
+ for (WordList_T::const_iterator itWord = iWordList.begin();
+ itWord != iWordList.end(); ++itWord, --idx) {
+ const std::string& lWord = *itWord;
+ oStr << lWord;
+ if (idx > 1) {
+ oStr << " ";
+ }
+ }
+
+ return oStr.str();
+ }
+
+}
Added: trunk/opentrep/opentrep/bom/WordHolder.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/WordHolder.hpp (rev 0)
+++ trunk/opentrep/opentrep/bom/WordHolder.hpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -0,0 +1,43 @@
+#ifndef __OPENTREP_BOM_WORDHOLDER_HPP
+#define __OPENTREP_BOM_WORDHOLDER_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// OpenTREP
+#include <opentrep/bom/BomAbstract.hpp>
+#include <opentrep/bom/WordList.hpp>
+
+namespace OPENTREP {
+
+ /** Class wrapping utility functions to transform back and forth
+ strings from and into a list of words. */
+ class WordHolder : public BomAbstract {
+ friend class FacWordHolder;
+ public:
+
+ // /////////////// Business Methods ////////////////
+ /** Tokenise a string into a list of words (STL strings).
+ <br>The Boost.Tokenizer library is used. */
+ static void tokeniseStringIntoWordList (const std::string& iPhrase,
+ WordList_T& ioWordList);
+
+ /** Serialise a list of words (STL strings) into a single (STL) string. */
+ static std::string createStringFromWordList (const WordList_T& iWordList);
+
+ private:
+ // ////////////// Constructors and Destructors /////////////
+ /** Default constructor. */
+ WordHolder ();
+ /** Default copy constructor. */
+ WordHolder (const WordHolder&);
+ /** Destructor. */
+ ~WordHolder ();
+
+
+ private:
+ // /////////////// Attributes ////////////////
+ };
+
+}
+#endif // __OPENTREP_BOM_WORDHOLDER_HPP
Added: trunk/opentrep/opentrep/bom/WordList.hpp
===================================================================
--- trunk/opentrep/opentrep/bom/WordList.hpp (rev 0)
+++ trunk/opentrep/opentrep/bom/WordList.hpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -0,0 +1,18 @@
+#ifndef __OPENTREP_BOM_WORDLIST_HPP
+#define __OPENTREP_BOM_WORDLIST_HPP
+
+// //////////////////////////////////////////////////////////////////////
+// Import section
+// //////////////////////////////////////////////////////////////////////
+// STL
+#include <string>
+#include <list>
+
+namespace OPENTREP {
+
+ /** List of simple words (STL strings). */
+ typedef std::list<std::string> WordList_T;
+
+}
+#endif // __OPENTREP_BOM_WORDLIST_HPP
+
Modified: trunk/opentrep/opentrep/bom/sources.mk
===================================================================
--- trunk/opentrep/opentrep/bom/sources.mk 2009-07-16 16:45:48 UTC (rev 132)
+++ trunk/opentrep/opentrep/bom/sources.mk 2009-07-17 00:10:38 UTC (rev 133)
@@ -5,10 +5,21 @@
$(top_srcdir)/opentrep/bom/World.hpp \
$(top_srcdir)/opentrep/bom/Names.hpp \
$(top_srcdir)/opentrep/bom/Place.hpp \
- $(top_srcdir)/opentrep/bom/PlaceList.hpp
+ $(top_srcdir)/opentrep/bom/PlaceList.hpp \
+ $(top_srcdir)/opentrep/bom/WordList.hpp \
+ $(top_srcdir)/opentrep/bom/WordHolder.hpp \
+ $(top_srcdir)/opentrep/bom/DocumentList.hpp \
+ $(top_srcdir)/opentrep/bom/Result.hpp \
+ $(top_srcdir)/opentrep/bom/ResultList.hpp \
+ $(top_srcdir)/opentrep/bom/ResultHolder.hpp \
+ $(top_srcdir)/opentrep/bom/StringMatcher.hpp
bom_cc_sources = $(top_srcdir)/opentrep/bom/BomAbstract.cpp \
$(top_srcdir)/opentrep/bom/BomType.cpp \
$(top_srcdir)/opentrep/bom/Language.cpp \
$(top_srcdir)/opentrep/bom/World.cpp \
$(top_srcdir)/opentrep/bom/Names.cpp \
- $(top_srcdir)/opentrep/bom/Place.cpp
+ $(top_srcdir)/opentrep/bom/Place.cpp \
+ $(top_srcdir)/opentrep/bom/WordHolder.cpp \
+ $(top_srcdir)/opentrep/bom/Result.cpp \
+ $(top_srcdir)/opentrep/bom/ResultHolder.cpp \
+ $(top_srcdir)/opentrep/bom/StringMatcher.cpp
Modified: trunk/opentrep/opentrep/command/RequestInterpreter.cpp
===================================================================
--- trunk/opentrep/opentrep/command/RequestInterpreter.cpp 2009-07-16 16:45:48 UTC (rev 132)
+++ trunk/opentrep/opentrep/command/RequestInterpreter.cpp 2009-07-17 00:10:38 UTC (rev 133)
@@ -9,6 +9,8 @@
#include <exception>
// OpenTrep
#include <opentrep/bom/Place.hpp>
+#include <opentrep/bom/ResultHolder.hpp>
+#include <opentrep/bom/Result.hpp>
#include <opentrep/factory/FacPlace.hpp>
#include <opentrep/command/DBManager.hpp>
#include <opentrep/command/RequestInterpreter.hpp>
@@ -23,44 +25,51 @@
interpretTravelRequest (soci::session& ioSociSession,
const TravelDatabaseName_T& iTravelDatabaseName,
...
[truncated message content] |
|
From: <den...@us...> - 2009-07-16 16:45:50
|
Revision: 132
http://opentrep.svn.sourceforge.net/opentrep/?rev=132&view=rev
Author: denis_arnaud
Date: 2009-07-16 16:45:48 +0000 (Thu, 16 Jul 2009)
Log Message:
-----------
[Dev] The (testing) string search fully works now.
Modified Paths:
--------------
trunk/opentrep/test/xapian/Makefile
trunk/opentrep/test/xapian/string_search.cpp
Modified: trunk/opentrep/test/xapian/Makefile
===================================================================
--- trunk/opentrep/test/xapian/Makefile 2009-07-15 23:42:07 UTC (rev 131)
+++ trunk/opentrep/test/xapian/Makefile 2009-07-16 16:45:48 UTC (rev 132)
@@ -1,7 +1,7 @@
# Xapian
-XAPIAN_DIR=/opt/xapian-1.0.10
-#XAPIAN_DIR=/usr
+#XAPIAN_DIR=/opt/xapian-1.0.11
+XAPIAN_DIR=/usr
XAPIAN_CFLAGS=-I${XAPIAN_DIR}/include
XAPIAN_LIBS=-L${XAPIAN_DIR}/lib -lxapian
Modified: trunk/opentrep/test/xapian/string_search.cpp
===================================================================
--- trunk/opentrep/test/xapian/string_search.cpp 2009-07-15 23:42:07 UTC (rev 131)
+++ trunk/opentrep/test/xapian/string_search.cpp 2009-07-16 16:45:48 UTC (rev 132)
@@ -5,6 +5,7 @@
#include <sstream>
#include <string>
#include <list>
+#include <map>
// Boost
#include <boost/tokenizer.hpp>
// Xapian
@@ -15,12 +16,25 @@
typedef std::list<std::string> WordList_T;
/** List of Xapian documents. */
-typedef std::list<Xapian::Document> DocumentList_T;
+typedef std::multimap<Xapian::percent, Xapian::Document> DocumentList_T;
+/** List of results, organised by sets, for a given full string. */
+typedef std::map<std::string, DocumentList_T> ResultList_T;
+/** Memory allocation error. */
+class MemoryError : public std::exception {
+};
+
+
+// ///////////////////////// Forward declarations ///////////////////////
+std::string display (const Xapian::MSet& iMatchingSet);
+std::string display (const DocumentList_T& iDocumentList);
+std::string display (const ResultList_T& iResultList);
+
+
// //////////////////////////////////////////////////////////////////////
-void tokeniseAndAddToDocument (const std::string& iPhrase,
- WordList_T& ioWordList) {
+void tokeniseStringIntoWordList (const std::string& iPhrase,
+ WordList_T& ioWordList) {
// Empty the word list
ioWordList.clear();
@@ -36,9 +50,8 @@
tok_iter != lTokens.end(); ++tok_iter) {
const std::string& lTerm = *tok_iter;
ioWordList.push_back (lTerm);
-
- // OPENTREP_LOG_DEBUG ("Added term: " << lTerm);
}
+
}
// //////////////////////////////////////////////////////////////////////
@@ -82,9 +95,11 @@
}
// DEBUG
+ /*
std::cout << "Original word: `" << lOriginalWord
<< "' ==> corrected word: `" << lSuggestedWord << "'"
<< std::endl;
+ */
}
} catch (const Xapian::Error& error) {
@@ -93,9 +108,8 @@
}
// ///////////////////////////////////////////////////////////////////
-void searchString (Xapian::MSet& ioMatchingSet,
- const std::string& iSearchString,
- Xapian::Database& ioDatabase) {
+void searchString (Xapian::MSet& ioMatchingSet, const std::string& iSearchString,
+ Xapian::Database& ioDatabase, std::ostream& ioStream) {
// Catch any Xapian::Error exceptions thrown
try {
@@ -119,8 +133,16 @@
further step below.
*/
WordList_T lOriginalWordList;
- tokeniseAndAddToDocument (iSearchString, lOriginalWordList);
-
+ tokeniseStringIntoWordList (iSearchString, lOriginalWordList);
+
+ /**
+ We rebuild a clean query string from the word list. Indeed, the original
+ string may have contained a few separators (e.g., '/', ';', etc.), which
+ have been removed by the tokeniseStringIntoWordList() method. All those
+ separators are thus replaced by spaces.
+ For instance, the 'san francisco, ca, us' initial string would be
+ replaced by 'san francisco ca us'.
+ */
const std::string lOriginalQueryString =
createStringFromWordList (lOriginalWordList);
@@ -137,11 +159,14 @@
*/
const std::string lFullWordCorrectedString =
ioDatabase.get_spelling_suggestion (lOriginalQueryString, 3);
-
- std::cout << "Query string `" << lOriginalQueryString
+
+ // DEBUG
+ /*
+ ioStream << "Query string `" << lOriginalQueryString
<< "' ==> corrected query string: `" << lCorrectedQueryString
<< "' and correction for the full query string: `"
<< lFullWordCorrectedString << "'" << std::endl;
+ */
// Build the query object
Xapian::QueryParser lQueryParser;
@@ -153,9 +178,12 @@
*/
// lQueryParser.set_default_op (Xapian::Query::OP_ADJ);
lQueryParser.set_default_op (Xapian::Query::OP_PHRASE);
-
- std::cout << "Query parser `" << lQueryParser.get_description() << "'"
+
+ // DEBUG
+ /*
+ ioStream << "Query parser `" << lQueryParser.get_description() << "'"
<< std::endl;
+ */
/**
The Xapian::QueryParser::parse_query() method aggregates all the words
@@ -196,10 +224,13 @@
| Xapian::QueryParser::FLAG_PHRASE
| Xapian::QueryParser::FLAG_LOVEHATE);
- std::cout << "Query `" << lQuery.get_description()
+ // DEBUG
+ /*
+ ioStream << "Query `" << lQuery.get_description()
<< "', corrected query `" << lCorrectedQuery.get_description()
<< "' and corrected for full query `"
<< lFullQueryCorrected.get_description() << "' " << std::endl;
+ */
// Start an enquire session
Xapian::Enquire enquire (ioDatabase);
@@ -212,8 +243,12 @@
// Display the results
int nbMatches = ioMatchingSet.size();
- std::cout << nbMatches << " results found" << std::endl;
+ // DEBUG
+ /*
+ ioStream << nbMatches << " results found" << std::endl;
+ */
+
/**
When no match is found, we search on the corrected phrase/string
(where the words have been corrected one by one).
@@ -224,8 +259,12 @@
// Display the results
nbMatches = ioMatchingSet.size();
- std::cout << nbMatches << " results found on corrected string"
+
+ // DEBUG
+ /*
+ ioStream << nbMatches << " results found on corrected string"
<< std::endl;
+ */
}
/**
@@ -238,13 +277,21 @@
// Display the results
nbMatches = ioMatchingSet.size();
- std::cout << nbMatches << " results found on corrected full string"
+
+ // DEBUG
+ /*
+ ioStream << nbMatches << " results found on corrected full string"
<< std::endl;
+ */
}
const Xapian::Query& lActualQuery = enquire.get_query();
- std::cout << "Actual query `" << lActualQuery.get_description()
+
+ // DEBUG
+ /*
+ ioStream << "Actual query `" << lActualQuery.get_description()
<< "'" << std::endl;
+ */
} catch (const Xapian::Error& error) {
std::cerr << "Exception: " << error.get_msg() << std::endl;
@@ -254,11 +301,15 @@
// //////////////////////////////////////////////////////////////////////
void createDocumentListFromMSet (const Xapian::MSet& iMatchingSet,
DocumentList_T& ioDocumentList) {
+ // Empty the list of documents
+ ioDocumentList.clear();
for (Xapian::MSetIterator itDoc = iMatchingSet.begin();
itDoc != iMatchingSet.end(); ++itDoc) {
const Xapian::Document& lDocument = itDoc.get_document();
- ioDocumentList.push_back (lDocument);
+
+ ioDocumentList.insert (DocumentList_T::value_type (itDoc.get_percent(),
+ lDocument));
}
}
@@ -267,7 +318,7 @@
assert (ioQueryString.empty() == false);
WordList_T lWordList;
- tokeniseAndAddToDocument (ioQueryString, lWordList);
+ tokeniseStringIntoWordList (ioQueryString, lWordList);
assert (lWordList.empty() == false);
// Remove the furthest right word
@@ -278,6 +329,169 @@
}
// //////////////////////////////////////////////////////////////////////
+void subtractParsedToRemaining (std::string& ioAlreadyParsedQueryString,
+ std::string& ioRemainingQueryString,
+ std::ostream& ioStream) {
+ /**
+ Remove, from the lRemainingQueryString string, the part which
+ has been already successfully parsed.
+ <br>For instance, when 'sna francisco rio de janeiro' is the
+ initial full clean query string, the searchString() method
+ first reduce the query string to 'sna francisco', which
+ successfully matches against SFO (San Francisco airport).
+ <br>Then, the remaining part of the query string to be parsed is
+ 'rio de janeiro'. So, the already parsed part must be subtracted
+ from the initial query string.
+ */
+ WordList_T lRemainingWordList;
+ tokeniseStringIntoWordList (ioRemainingQueryString, lRemainingWordList);
+
+ WordList_T lParsedWordList;
+ tokeniseStringIntoWordList (ioAlreadyParsedQueryString, lParsedWordList);
+
+ unsigned int idx = lParsedWordList.size();
+ for ( ; idx != 0 && lRemainingWordList.empty() == false; --idx) {
+ lRemainingWordList.pop_front();
+ }
+
+ // Build the remaining part of the string still to be parsed.
+ // <br>Note that that part may be empty.
+ ioRemainingQueryString = createStringFromWordList (lRemainingWordList);
+}
+
+// //////////////////////////////////////////////////////////////////////
+void searchString (DocumentList_T& ioDocumentList, std::string& ioQueryString,
+ Xapian::Database& ioDatabase, std::ostream& ioStream) {
+
+ // Catch any Xapian::Error exceptions thrown
+ try {
+
+ bool shouldStop = false;
+ while (shouldStop == false) {
+ // DEBUG
+ /*
+ ioStream << std::endl << "--------------------------------" << std::endl
+ << "Current query string: `" << ioQueryString << "'" << std::endl;
+ */
+
+ // Retrieve the list of documents matching the query string
+ Xapian::MSet lMatchingSet;
+ searchString (lMatchingSet, ioQueryString, ioDatabase, ioStream);
+
+ // DEBUG
+ /*
+ ioStream << "Matching set (" << lMatchingSet.size() << " elements): "
+ << display (lMatchingSet);
+ */
+
+ // Create the corresponding list of documents
+ createDocumentListFromMSet (lMatchingSet, ioDocumentList);
+
+ // Stop if a result is found.
+ if (ioDocumentList.empty() == false) {
+ shouldStop = true;
+ break;
+ }
+
+ // Remove a word from the query string
+ removeOneWord (ioQueryString);
+
+ // Stop when the resulting string gets empty.
+ if (ioQueryString.empty() == true) {
+ shouldStop = true;
+ }
+ }
+
+ } catch (const Xapian::Error& error) {
+ std::cerr << "Exception: " << error.get_msg() << std::endl;
+ }
+}
+
+// //////////////////////////////////////////////////////////////////////
+void searchString (const std::string& iFullCleanQueryString,
+ ResultList_T& ioResultList, Xapian::Database& ioDatabase,
+ std::ostream& ioStream) {
+
+ // Catch any Xapian::Error exceptions thrown
+ try {
+
+ std::string lRemainingQueryString (iFullCleanQueryString);
+ bool shouldStop = false;
+ while (shouldStop == false) {
+ // DEBUG
+ /*
+ ioStream << std::endl << "================================" << std::endl
+ << "Current query string: `" << lRemainingQueryString << "'"
+ << std::endl;
+ */
+ /**
+ Search with the initial full string, then by removing a word if
+ there was no result, then by removing another word if there was
+ again no result, until either a result is found or the
+ resulting string gets empty.
+ */
+ DocumentList_T lDocumentList;
+ std::string lQueryString (lRemainingQueryString);
+ searchString (lDocumentList, lQueryString, ioDatabase, ioStream);
+
+ // Add the list of matching documents to the result list
+ const bool hasInsertionBeenSuccessfull =
+ ioResultList.insert (ResultList_T::value_type (lQueryString,
+ lDocumentList)).second;
+ if (hasInsertionBeenSuccessfull == false) {
+ std::cerr << "Insertion of document list failed for: `"
+ << display (lDocumentList) << "\xB4" << std::endl;
+ throw MemoryError();
+ }
+
+ /**
+ Remove, from the lRemainingQueryString string, the part which
+ has been already successfully parsed.
+ <br>For instance, when 'sna francisco rio de janeiro' is the
+ initial full clean query string, the searchString() method
+ first reduce the query string to 'sna francisco', which
+ successfully matches against SFO (San Francisco airport).
+ <br>Then, the remaining part of the query string to be parsed is
+ 'rio de janeiro'. So, the already parsed part, namely 'sna francisco',
+ must be subtracted from the initial query string.
+ */
+ subtractParsedToRemaining (lQueryString, lRemainingQueryString, ioStream);
+
+ // If there is nothing left to be parsed, we have then finished
+ // to parse the initial string.
+ if (lRemainingQueryString.empty() == true) {
+ shouldStop = true;
+ break;
+ }
+ }
+
+ } catch (const Xapian::Error& error) {
+ std::cerr << "Exception: " << error.get_msg() << std::endl;
+ }
+}
+
+// //////////////////////////////////////////////////////////////////////
+void extractResults (const ResultList_T& iResultList,
+ DocumentList_T& ioDocumentList) {
+
+ for (ResultList_T::const_reverse_iterator itResult = iResultList.rbegin();
+ itResult != iResultList.rend(); ++itResult) {
+ const std::string& lQueryString = itResult->first;
+ const DocumentList_T& lDocumentList = itResult->second;
+
+ // Retrieve the best matching document. As the document list (STL map)
+ // is sorted by ascending order of the matching percentage, the best
+ // matching one is located at the end (back) of the list (STL map).
+ DocumentList_T::const_reverse_iterator itDocument = lDocumentList.rbegin();
+ const Xapian::percent& lBestMatchingPercentage = itDocument->first;
+ const Xapian::Document& lBestMatchingDocument = itDocument->second;
+
+ ioDocumentList.insert (DocumentList_T::value_type (lBestMatchingPercentage,
+ lBestMatchingDocument));
+ }
+}
+
+// //////////////////////////////////////////////////////////////////////
std::string display (const Xapian::MSet& iMatchingSet) {
std::ostringstream oStr;
@@ -299,16 +513,33 @@
for (DocumentList_T::const_iterator itDoc = iDocumentList.begin();
itDoc != iDocumentList.end(); ++itDoc) {
- const Xapian::Document& lDocument = *itDoc;
+ const Xapian::percent& lPercent = itDoc->first;
+ const Xapian::Document& lDocument = itDoc->second;
const Xapian::docid& lDocID = lDocument.get_docid();
- oStr << "Document ID " << lDocID << "\t["
- << lDocument.get_data() << "]" << std::endl;
+ oStr << "Document ID " << lDocID << "\t" << lPercent
+ << "% [" << lDocument.get_data() << "]" << std::endl;
}
return oStr.str();
}
+// //////////////////////////////////////////////////////////////////////
+std::string display (const ResultList_T& iResultList) {
+ std::ostringstream oStr;
+
+ for (ResultList_T::const_iterator itResult = iResultList.begin();
+ itResult != iResultList.end(); ++itResult) {
+ const std::string& lQueryString = itResult->first;
+ const DocumentList_T& lDocumentList = itResult->second;
+
+ oStr << "Result for query (sub-)string: `" << lQueryString << "\xB4:"
+ << std::endl << display (lDocumentList);
+ }
+
+ return oStr.str();
+}
+
// //////////////////////////// M A I N //////////////////////////////
int main (int argc, char* argv[]) {
@@ -324,7 +555,8 @@
try {
// Make the database
- Xapian::Database lDatabase (argv[1]);
+ const std::string lXapianDatabaseFilepath (argv[1]);
+ Xapian::Database lDatabase (lXapianDatabaseFilepath);
/**
Build a query string from the command line parameters.
@@ -339,45 +571,34 @@
const std::string lWord (argv[idx]);
lQueryStringStr << lWord;
}
- const std::string& lCommandLineQueryString = lQueryStringStr.str();
+ const std::string& lFullCleanQueryString = lQueryStringStr.str();
- /**
- Search with the initial full string, then by removing a word if
- no there was result, then by removing another word if there was
- again no result, until either a result is found or the
- resulting string gets empty.
- */
- std::string lQueryString (lCommandLineQueryString);
- bool shouldStop = false;
- while (shouldStop == false) {
- // DEBUG
- std::cout << std::endl << "================================" << std::endl
- << "New query string: `" << lQueryString << "'" << std::endl;
-
- // Retrieve the list of documents matching the query string
- Xapian::MSet lMatchingSet;
- searchString (lMatchingSet, lQueryString, lDatabase);
- std::cout << display (lMatchingSet);
-
- // Create the corresponding list of documents
- DocumentList_T lDocumentList;
- createDocumentListFromMSet (lMatchingSet, lDocumentList);
+ //
+ ResultList_T lResultList;
+ searchString (lFullCleanQueryString, lResultList, lDatabase, std::cout);
- // Stop if a result is found.
- if (lDocumentList.empty() == false) {
- shouldStop = true;
- break;
- }
+ // DEBUG
+ /*
+ std::cout << std::endl << "Result list: " << std::endl
+ << display (lResultList);
+ */
- // Remove a word from the query string
- removeOneWord (lQueryString);
-
- // Stop when the resulting string gets empty.
- if (lQueryString.empty() == true) {
- shouldStop = true;
- }
- }
+ // Extract the best matching results from each matching set (document list)
+ DocumentList_T lDocumentList;
+ extractResults (lResultList, lDocumentList);
+ // DEBUG
+ std::cout << std::endl
+ << "_________________________________________" << std::endl
+ << "=========================================" << std::endl
+ << "-----------------------------------------" << std::endl
+ << "Matching list: " << std::endl
+ << display (lDocumentList)
+ << "_________________________________________" << std::endl
+ << "=========================================" << std::endl
+ << "-----------------------------------------" << std::endl
+ << std::endl;
+
} catch (const Xapian::Error& error) {
std::cerr << "Exception: " << error.get_msg() << std::endl;
}
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-15 23:42:10
|
Revision: 131
http://opentrep.svn.sourceforge.net/opentrep/?rev=131&view=rev
Author: denis_arnaud
Date: 2009-07-15 23:42:07 +0000 (Wed, 15 Jul 2009)
Log Message:
-----------
[Dev] Improved the string search, by allowing several cities at once. There is still some work to do.
Modified Paths:
--------------
trunk/opentrep/test/xapian/string_search.cpp
Modified: trunk/opentrep/test/xapian/string_search.cpp
===================================================================
--- trunk/opentrep/test/xapian/string_search.cpp 2009-07-15 16:59:19 UTC (rev 130)
+++ trunk/opentrep/test/xapian/string_search.cpp 2009-07-15 23:42:07 UTC (rev 131)
@@ -1,140 +1,386 @@
+// C
+#include <cassert>
// STL
#include <iostream>
#include <sstream>
+#include <string>
+#include <list>
+// Boost
+#include <boost/tokenizer.hpp>
// Xapian
#include <xapian.h>
-// ////////////// M A I N //////////////
-int main (int argc, char* argv[]) {
+// /////////////////////// Type definitions /////////////////////////////
+/** List of simple words (STL strings). */
+typedef std::list<std::string> WordList_T;
- // Simplest possible options parsing: we just require two or more
- // parameters.
- if (argc < 3) {
- std::cout << "Usage: " << argv[0]
- << " <path to database> <search terms>" << std::endl;
- return -1;
+/** List of Xapian documents. */
+typedef std::list<Xapian::Document> DocumentList_T;
+
+
+// //////////////////////////////////////////////////////////////////////
+void tokeniseAndAddToDocument (const std::string& iPhrase,
+ WordList_T& ioWordList) {
+ // Empty the word list
+ ioWordList.clear();
+
+ // Boost Tokeniser
+ typedef boost::tokenizer<boost::char_separator<char> > Tokeniser_T;
+
+ // Define the separators
+ const boost::char_separator<char> lSepatorList(" .,;:|+-*/_=!@#$%`~^&(){}[]?'<>\"");
+
+ // Initialise the phrase to be tokenised
+ Tokeniser_T lTokens (iPhrase, lSepatorList);
+ for (Tokeniser_T::const_iterator tok_iter = lTokens.begin();
+ tok_iter != lTokens.end(); ++tok_iter) {
+ const std::string& lTerm = *tok_iter;
+ ioWordList.push_back (lTerm);
+
+ // OPENTREP_LOG_DEBUG ("Added term: " << lTerm);
+ }
+}
+
+// //////////////////////////////////////////////////////////////////////
+std::string createStringFromWordList (const WordList_T& iWordList) {
+ std::ostringstream oStr;
+
+ unsigned short idx = iWordList.size();
+ for (WordList_T::const_iterator itWord = iWordList.begin();
+ itWord != iWordList.end(); ++itWord, --idx) {
+ const std::string& lWord = *itWord;
+ oStr << lWord;
+ if (idx > 1) {
+ oStr << " ";
}
+ }
+
+ return oStr.str();
+}
- // Catch any Xapian::Error exceptions thrown
- try {
+// //////////////////////////////////////////////////////////////////////
+void createCorrectedWordList (const WordList_T& iOriginalWordList,
+ WordList_T& ioCorrectedWordList,
+ const Xapian::Database& iDatabase) {
+ // Empty the target list
+ ioCorrectedWordList.clear();
+
+ // Catch any Xapian::Error exceptions thrown
+ try {
- // Make the database
- Xapian::Database db (argv[1]);
+ for (WordList_T::const_iterator itWord = iOriginalWordList.begin();
+ itWord != iOriginalWordList.end(); ++itWord) {
+ const std::string& lOriginalWord = *itWord;
+ const std::string& lSuggestedWord =
+ iDatabase.get_spelling_suggestion (lOriginalWord, 3);
- // Start an enquire session
- Xapian::Enquire enquire (db);
+ if (lSuggestedWord.empty() == true) {
+ ioCorrectedWordList.push_back (lOriginalWord);
- std::ostringstream oOriginalStr;
- std::ostringstream oCorrectedStr;
- for (int idx=2; idx != argc; ++idx) {
- if (idx != 2) {
- oOriginalStr << " ";
- oCorrectedStr << " ";
- }
- const std::string lWord (argv[idx]);
- const std::string lSuggestedWord = db.get_spelling_suggestion(lWord, 3);
- std::cout << "Word `" << lWord << "' ==> Suggested word `"
- << lSuggestedWord << "'" << std::endl;
- oOriginalStr << lWord;
+ } else {
+ ioCorrectedWordList.push_back (lSuggestedWord);
+ }
- if (lSuggestedWord.empty() == true) {
- oCorrectedStr << lWord;
-
- } else {
- oCorrectedStr << lSuggestedWord;
- }
- }
+ // DEBUG
+ std::cout << "Original word: `" << lOriginalWord
+ << "' ==> corrected word: `" << lSuggestedWord << "'"
+ << std::endl;
+ }
+
+ } catch (const Xapian::Error& error) {
+ std::cerr << "Exception: " << error.get_msg() << std::endl;
+ }
+}
+
+// ///////////////////////////////////////////////////////////////////
+void searchString (Xapian::MSet& ioMatchingSet,
+ const std::string& iSearchString,
+ Xapian::Database& ioDatabase) {
+
+ // Catch any Xapian::Error exceptions thrown
+ try {
- const std::string lOriginalQueryString = oOriginalStr.str();
- const std::string lCorrectedQueryString = oCorrectedStr.str();
- const std::string lFullWordCorrectedString =
- db.get_spelling_suggestion (lOriginalQueryString, 4);
+ /**
+ Build another string, in addition to the original one. Overall,
+ there are thus two strings:
+ <br><ul>
+ <li>One with the original words given by the user</li>
+ <li>One with the orthographic-corrected words, wherever
+ relevant (otherwise, the original word is taken)</li>
+ </ul>
+ <br>For instance, 'sna francisco' would give the following
+ two strings:
+ <br><ul>
+ <li>'sna francicso' (original)</li>
+ <li>'sna francisco' (corrected, where relevant, word by word)</li>
+ </ul>
+ <br>Note that, as 'sna' exists in the dictionary (Santa Ana, CA, USA),
+ it is not replaced. We shall take care of the whole string in a
+ further step below.
+ */
+ WordList_T lOriginalWordList;
+ tokeniseAndAddToDocument (iSearchString, lOriginalWordList);
- std::cout << "Query string `" << lOriginalQueryString
- << "' ==> corrected query string: `" << lCorrectedQueryString
- << "' and correction for the full query string: `"
- << lFullWordCorrectedString << "'" << std::endl;
+ const std::string lOriginalQueryString =
+ createStringFromWordList (lOriginalWordList);
+
+ WordList_T lCorrectedWordList;
+ createCorrectedWordList (lOriginalWordList, lCorrectedWordList, ioDatabase);
+
+ const std::string lCorrectedQueryString =
+ createStringFromWordList (lCorrectedWordList);
- // Build the query object
- Xapian::QueryParser lQueryParser;
- lQueryParser.set_database (db);
- // As explained in http://www.xapian.org/docs/queryparser.html,
- // Xapian::Query::OP_ADJ is better than Xapian::Query::OP_PHRASE,
- // but only available from version 1.0.13 of Xapian.
- // lQueryParser.set_default_op (Xapian::Query::OP_ADJ);
- lQueryParser.set_default_op (Xapian::Query::OP_PHRASE);
+ /**
+ Try to find, if relevant, an orthographic suggestion for the whole
+ phrase/string. With the above example, 'sna francisco' yields the
+ suggestion 'san francisco'.
+ */
+ const std::string lFullWordCorrectedString =
+ ioDatabase.get_spelling_suggestion (lOriginalQueryString, 3);
- std::cout << "Query parser `" << lQueryParser.get_description() << "'"
- << std::endl;
+ std::cout << "Query string `" << lOriginalQueryString
+ << "' ==> corrected query string: `" << lCorrectedQueryString
+ << "' and correction for the full query string: `"
+ << lFullWordCorrectedString << "'" << std::endl;
- Xapian::Query lQuery =
- lQueryParser.parse_query (lOriginalQueryString,
- Xapian::QueryParser::FLAG_BOOLEAN
- | Xapian::QueryParser::FLAG_PHRASE
- | Xapian::QueryParser::FLAG_LOVEHATE
- | Xapian::QueryParser::FLAG_SPELLING_CORRECTION);
- //Xapian::Query lCorrectedQuery= lQueryParser.get_corrected_query_string();
- Xapian::Query lCorrectedQuery =
- lQueryParser.parse_query (lCorrectedQueryString,
- Xapian::QueryParser::FLAG_BOOLEAN
- | Xapian::QueryParser::FLAG_PHRASE
- | Xapian::QueryParser::FLAG_LOVEHATE);
+ // Build the query object
+ Xapian::QueryParser lQueryParser;
+ lQueryParser.set_database (ioDatabase);
+ /**
+ As explained in http://www.xapian.org/docs/queryparser.html,
+ Xapian::Query::OP_ADJ is better than Xapian::Query::OP_PHRASE,
+ but only available from version 1.0.13 of Xapian.
+ */
+ // lQueryParser.set_default_op (Xapian::Query::OP_ADJ);
+ lQueryParser.set_default_op (Xapian::Query::OP_PHRASE);
+
+ std::cout << "Query parser `" << lQueryParser.get_description() << "'"
+ << std::endl;
- Xapian::Query lFullQueryCorrected =
- lQueryParser.parse_query (lFullWordCorrectedString,
- Xapian::QueryParser::FLAG_BOOLEAN
- | Xapian::QueryParser::FLAG_PHRASE
- | Xapian::QueryParser::FLAG_LOVEHATE);
+ /**
+ The Xapian::QueryParser::parse_query() method aggregates all the words
+ with operators inbetween them (here, the "PHRASE" operator).
+ With the above example ('sna francicso'), it yields
+ "sna PHRASE 2 francicso".
+ */
+ Xapian::Query lQuery =
+ lQueryParser.parse_query (lOriginalQueryString,
+ Xapian::QueryParser::FLAG_BOOLEAN
+ | Xapian::QueryParser::FLAG_PHRASE
+ | Xapian::QueryParser::FLAG_LOVEHATE
+ | Xapian::QueryParser::FLAG_SPELLING_CORRECTION);
+ /**
+ Strangely enough (is it?), the corrected query given by the Xapian
+ QueryParser corresponds to the full original string, where words
+ have been corrected one by one, but considered as a single block.
+ With the above example, 'sna francicso' yields (wrongly)
+ 'sna francisco', instead of "sna PHRASE 2 francisco", as generated
+ by the following code.
+ */
+ // Xapian::Query lCorrectedQuery =
+ // lQueryParser.get_corrected_query_string();
+ Xapian::Query lCorrectedQuery =
+ lQueryParser.parse_query (lCorrectedQueryString,
+ Xapian::QueryParser::FLAG_BOOLEAN
+ | Xapian::QueryParser::FLAG_PHRASE
+ | Xapian::QueryParser::FLAG_LOVEHATE);
- std::cout << "Query `" << lQuery.get_description()
- << "', corrected query `" << lCorrectedQuery.get_description()
- << "' and corrected for full query `"
- << lFullQueryCorrected.get_description() << "' " << std::endl;
+ /**
+ As, with the above example, the full corrected string is
+ 'san francisco', it yields the query "san PHRASE 2 francisco",
+ which is eventually right.
+ */
+ Xapian::Query lFullQueryCorrected =
+ lQueryParser.parse_query (lFullWordCorrectedString,
+ Xapian::QueryParser::FLAG_BOOLEAN
+ | Xapian::QueryParser::FLAG_PHRASE
+ | Xapian::QueryParser::FLAG_LOVEHATE);
+
+ std::cout << "Query `" << lQuery.get_description()
+ << "', corrected query `" << lCorrectedQuery.get_description()
+ << "' and corrected for full query `"
+ << lFullQueryCorrected.get_description() << "' " << std::endl;
- // Give the query object to the enquire session
- enquire.set_query (lQuery);
+ // Start an enquire session
+ Xapian::Enquire enquire (ioDatabase);
- // Get the top 10 results of the query
- Xapian::MSet matches = enquire.get_mset (0, 10);
+ // Give the query object to the enquire session
+ enquire.set_query (lQuery);
+ // Get the top 10 results of the query
+ ioMatchingSet = enquire.get_mset (0, 10);
+
+ // Display the results
+ int nbMatches = ioMatchingSet.size();
+ std::cout << nbMatches << " results found" << std::endl;
+
+ /**
+ When no match is found, we search on the corrected phrase/string
+ (where the words have been corrected one by one).
+ */
+ if (nbMatches == 0) {
+ enquire.set_query (lCorrectedQuery);
+ ioMatchingSet = enquire.get_mset (0, 10);
+
// Display the results
- int nbMatches = matches.size();
- std::cout << nbMatches << " results found" << std::endl;
+ nbMatches = ioMatchingSet.size();
+ std::cout << nbMatches << " results found on corrected string"
+ << std::endl;
+ }
- if (nbMatches == 0) {
- enquire.set_query (lCorrectedQuery);
- matches = enquire.get_mset (0, 10);
+ /**
+ If there is still no match, we search on the string corrected
+ as a whole.
+ */
+ if (nbMatches == 0) {
+ enquire.set_query (lFullQueryCorrected);
+ ioMatchingSet = enquire.get_mset (0, 10);
+
+ // Display the results
+ nbMatches = ioMatchingSet.size();
+ std::cout << nbMatches << " results found on corrected full string"
+ << std::endl;
+ }
- // Display the results
- nbMatches = matches.size();
- std::cout << nbMatches << " results found on corrected string"
- << std::endl;
+ const Xapian::Query& lActualQuery = enquire.get_query();
+ std::cout << "Actual query `" << lActualQuery.get_description()
+ << "'" << std::endl;
+
+ } catch (const Xapian::Error& error) {
+ std::cerr << "Exception: " << error.get_msg() << std::endl;
+ }
+}
- if (nbMatches == 0) {
- enquire.set_query (lFullQueryCorrected);
- matches = enquire.get_mset (0, 10);
+// //////////////////////////////////////////////////////////////////////
+void createDocumentListFromMSet (const Xapian::MSet& iMatchingSet,
+ DocumentList_T& ioDocumentList) {
- // Display the results
- nbMatches = matches.size();
- std::cout << nbMatches << " results found on corrected full string"
- << std::endl;
- }
+ for (Xapian::MSetIterator itDoc = iMatchingSet.begin();
+ itDoc != iMatchingSet.end(); ++itDoc) {
+ const Xapian::Document& lDocument = itDoc.get_document();
+ ioDocumentList.push_back (lDocument);
+ }
+}
+
+// //////////////////////////////////////////////////////////////////////
+void removeOneWord (std::string& ioQueryString) {
+ assert (ioQueryString.empty() == false);
+
+ WordList_T lWordList;
+ tokeniseAndAddToDocument (ioQueryString, lWordList);
+ assert (lWordList.empty() == false);
+
+ // Remove the furthest right word
+ lWordList.pop_back();
+
+ const std::string& lReducedString = createStringFromWordList (lWordList);
+ ioQueryString = lReducedString;
+}
+
+// //////////////////////////////////////////////////////////////////////
+std::string display (const Xapian::MSet& iMatchingSet) {
+ std::ostringstream oStr;
+
+ for (Xapian::MSetIterator itDoc = iMatchingSet.begin();
+ itDoc != iMatchingSet.end(); ++itDoc) {
+ const Xapian::Document& lDocument = itDoc.get_document();
+ const Xapian::docid& lDocID = lDocument.get_docid();
+
+ oStr << "Document ID " << lDocID << "\t" << itDoc.get_percent()
+ << "% [" << lDocument.get_data() << "]" << std::endl;
+ }
+
+ return oStr.str();
+}
+
+// //////////////////////////////////////////////////////////////////////
+std::string display (const DocumentList_T& iDocumentList) {
+ std::ostringstream oStr;
+
+ for (DocumentList_T::const_iterator itDoc = iDocumentList.begin();
+ itDoc != iDocumentList.end(); ++itDoc) {
+ const Xapian::Document& lDocument = *itDoc;
+ const Xapian::docid& lDocID = lDocument.get_docid();
+
+ oStr << "Document ID " << lDocID << "\t["
+ << lDocument.get_data() << "]" << std::endl;
+ }
+
+ return oStr.str();
+}
+
+// //////////////////////////// M A I N //////////////////////////////
+int main (int argc, char* argv[]) {
+
+ // Simplest possible options parsing: we just require two or more
+ // parameters.
+ if (argc < 3) {
+ std::cout << "Usage: " << argv[0]
+ << " <path to database> <search terms>" << std::endl;
+ return -1;
+ }
+
+ // Catch any Xapian::Error exceptions thrown
+ try {
+
+ // Make the database
+ Xapian::Database lDatabase (argv[1]);
+
+ /**
+ Build a query string from the command line parameters.
+ That way, any other front end producing a query string will
+ be fine.
+ */
+ std::ostringstream lQueryStringStr;
+ for (unsigned int idx = 2; idx != argc; ++idx) {
+ if (idx != 2) {
+ lQueryStringStr << " ";
}
+ const std::string lWord (argv[idx]);
+ lQueryStringStr << lWord;
+ }
+ const std::string& lCommandLineQueryString = lQueryStringStr.str();
- const Xapian::Query& lActualQuery = enquire.get_query();
- std::cout << "Actual query `" << lActualQuery.get_description()
- << "'" << std::endl;
+ /**
+ Search with the initial full string, then by removing a word if
+ no there was result, then by removing another word if there was
+ again no result, until either a result is found or the
+ resulting string gets empty.
+ */
+ std::string lQueryString (lCommandLineQueryString);
+ bool shouldStop = false;
+ while (shouldStop == false) {
+ // DEBUG
+ std::cout << std::endl << "================================" << std::endl
+ << "New query string: `" << lQueryString << "'" << std::endl;
- for (Xapian::MSetIterator i = matches.begin(); i != matches.end(); ++i) {
- Xapian::Document doc = i.get_document();
- std::cout << "Document ID " << *i << "\t" <<
- i.get_percent() << "% [" <<
- doc.get_data() << "]" << std::endl;
+ // Retrieve the list of documents matching the query string
+ Xapian::MSet lMatchingSet;
+ searchString (lMatchingSet, lQueryString, lDatabase);
+ std::cout << display (lMatchingSet);
+
+ // Create the corresponding list of documents
+ DocumentList_T lDocumentList;
+ createDocumentListFromMSet (lMatchingSet, lDocumentList);
+
+ // Stop if a result is found.
+ if (lDocumentList.empty() == false) {
+ shouldStop = true;
+ break;
}
- } catch (const Xapian::Error& error) {
- std::cerr << "Exception: " << error.get_msg() << std::endl;
+ // Remove a word from the query string
+ removeOneWord (lQueryString);
+
+ // Stop when the resulting string gets empty.
+ if (lQueryString.empty() == true) {
+ shouldStop = true;
+ }
}
+
+ } catch (const Xapian::Error& error) {
+ std::cerr << "Exception: " << error.get_msg() << std::endl;
+ }
- return 0;
+ return 0;
}
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-15 16:59:26
|
Revision: 130
http://opentrep.svn.sourceforge.net/opentrep/?rev=130&view=rev
Author: denis_arnaud
Date: 2009-07-15 16:59:19 +0000 (Wed, 15 Jul 2009)
Log Message:
-----------
[TREP] Improved the string search (command line utility).
Modified Paths:
--------------
trunk/opentrep/opentrep/command/IndexBuilder.cpp
trunk/opentrep/test/xapian/string_search.cpp
Modified: trunk/opentrep/opentrep/command/IndexBuilder.cpp
===================================================================
--- trunk/opentrep/opentrep/command/IndexBuilder.cpp 2009-07-15 14:45:43 UTC (rev 129)
+++ trunk/opentrep/opentrep/command/IndexBuilder.cpp 2009-07-15 16:59:19 UTC (rev 130)
@@ -33,7 +33,7 @@
typedef boost::tokenizer<boost::char_separator<char> > Tokeniser_T;
// Define the separators
- boost::char_separator<char> lSepatorList(" .,;:|+-*/_=!@#$%`~^&(){}[]?'<>\"");
+ const boost::char_separator<char> lSepatorList(" .,;:|+-*/_=!@#$%`~^&(){}[]?'<>\"");
// Initialise the phrase to be tokenised
Tokeniser_T lTokens (iPhrase, lSepatorList);
@@ -44,7 +44,7 @@
ioDatabase.add_spelling (lTerm);
ioDocument.add_term (lTerm);
- OPENTREP_LOG_DEBUG ("Added term: " << lTerm);
+ // OPENTREP_LOG_DEBUG ("Added term: " << lTerm);
}
}
@@ -115,10 +115,14 @@
// Add the place name (it can be the classical one, or
// extended, alternate, etc.)
if (lName.empty() == false) {
+ // Add the full name (potentially containing spaces, e.g.,
+ // 'san francisco'), as well as each word
+ // within it (with the example above, 'san' and 'francisco').
+ lDocument.add_term (lName); ++idx;
+ ioDatabase.add_spelling (lName);
+ tokeniseAndAddToDocument (lName, lDocument, ioDatabase);
+
// OPENTREP_LOG_DEBUG ("Added name: " << lName);
- // lDocument.add_term (lName); ++idx;
- // ioDatabase.add_spelling (lName);
- tokeniseAndAddToDocument (lName, lDocument, ioDatabase);
}
}
}
Modified: trunk/opentrep/test/xapian/string_search.cpp
===================================================================
--- trunk/opentrep/test/xapian/string_search.cpp 2009-07-15 14:45:43 UTC (rev 129)
+++ trunk/opentrep/test/xapian/string_search.cpp 2009-07-15 16:59:19 UTC (rev 130)
@@ -24,26 +24,43 @@
// Start an enquire session
Xapian::Enquire enquire (db);
- std::ostringstream oStr;
+ std::ostringstream oOriginalStr;
+ std::ostringstream oCorrectedStr;
for (int idx=2; idx != argc; ++idx) {
if (idx != 2) {
- oStr << " ";
+ oOriginalStr << " ";
+ oCorrectedStr << " ";
}
const std::string lWord (argv[idx]);
const std::string lSuggestedWord = db.get_spelling_suggestion(lWord, 3);
std::cout << "Word `" << lWord << "' ==> Suggested word `"
<< lSuggestedWord << "'" << std::endl;
- oStr << lWord;
+ oOriginalStr << lWord;
+
+ if (lSuggestedWord.empty() == true) {
+ oCorrectedStr << lWord;
+
+ } else {
+ oCorrectedStr << lSuggestedWord;
+ }
}
- const std::string lQueryString = oStr.str();
- std::cout << "QueryString `" << lQueryString << "'" << std::endl;
+
+ const std::string lOriginalQueryString = oOriginalStr.str();
+ const std::string lCorrectedQueryString = oCorrectedStr.str();
+ const std::string lFullWordCorrectedString =
+ db.get_spelling_suggestion (lOriginalQueryString, 4);
+
+ std::cout << "Query string `" << lOriginalQueryString
+ << "' ==> corrected query string: `" << lCorrectedQueryString
+ << "' and correction for the full query string: `"
+ << lFullWordCorrectedString << "'" << std::endl;
// Build the query object
Xapian::QueryParser lQueryParser;
lQueryParser.set_database (db);
// As explained in http://www.xapian.org/docs/queryparser.html,
// Xapian::Query::OP_ADJ is better than Xapian::Query::OP_PHRASE,
- // but only available from version 1.0.13 of Xapian
+ // but only available from version 1.0.13 of Xapian.
// lQueryParser.set_default_op (Xapian::Query::OP_ADJ);
lQueryParser.set_default_op (Xapian::Query::OP_PHRASE);
@@ -51,17 +68,28 @@
<< std::endl;
Xapian::Query lQuery =
- lQueryParser.parse_query (lQueryString,
+ lQueryParser.parse_query (lOriginalQueryString,
Xapian::QueryParser::FLAG_BOOLEAN
| Xapian::QueryParser::FLAG_PHRASE
| Xapian::QueryParser::FLAG_LOVEHATE
| Xapian::QueryParser::FLAG_SPELLING_CORRECTION);
- Xapian::Query lCorrectedQuery = lQueryParser.get_corrected_query_string();
+ //Xapian::Query lCorrectedQuery= lQueryParser.get_corrected_query_string();
+ Xapian::Query lCorrectedQuery =
+ lQueryParser.parse_query (lCorrectedQueryString,
+ Xapian::QueryParser::FLAG_BOOLEAN
+ | Xapian::QueryParser::FLAG_PHRASE
+ | Xapian::QueryParser::FLAG_LOVEHATE);
+ Xapian::Query lFullQueryCorrected =
+ lQueryParser.parse_query (lFullWordCorrectedString,
+ Xapian::QueryParser::FLAG_BOOLEAN
+ | Xapian::QueryParser::FLAG_PHRASE
+ | Xapian::QueryParser::FLAG_LOVEHATE);
+
std::cout << "Query `" << lQuery.get_description()
- << "', Corrected query `" << lCorrectedQuery.get_description()
- << "' "
- << std::endl;
+ << "', corrected query `" << lCorrectedQuery.get_description()
+ << "' and corrected for full query `"
+ << lFullQueryCorrected.get_description() << "' " << std::endl;
// Give the query object to the enquire session
enquire.set_query (lQuery);
@@ -70,17 +98,25 @@
Xapian::MSet matches = enquire.get_mset (0, 10);
// Display the results
- const int nbMatches = matches.size();
+ int nbMatches = matches.size();
std::cout << nbMatches << " results found" << std::endl;
-
- // if (true) {
if (nbMatches == 0) {
enquire.set_query (lCorrectedQuery);
matches = enquire.get_mset (0, 10);
- //const Xapian::MSet matchesAll = enquire.get_mset ();
- if (matches.size() == matches.max_size()) {
- std::cout << "Corrected string matches all the documents"
+
+ // Display the results
+ nbMatches = matches.size();
+ std::cout << nbMatches << " results found on corrected string"
+ << std::endl;
+
+ if (nbMatches == 0) {
+ enquire.set_query (lFullQueryCorrected);
+ matches = enquire.get_mset (0, 10);
+
+ // Display the results
+ nbMatches = matches.size();
+ std::cout << nbMatches << " results found on corrected full string"
<< std::endl;
}
}
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-15 14:45:49
|
Revision: 129
http://opentrep.svn.sourceforge.net/opentrep/?rev=129&view=rev
Author: denis_arnaud
Date: 2009-07-15 14:45:43 +0000 (Wed, 15 Jul 2009)
Log Message:
-----------
[Indexer] Fixed a bug in the indexer (where terms were inserted with spaces).
Modified Paths:
--------------
trunk/opentrep/opentrep/command/IndexBuilder.cpp
trunk/opentrep/refdata/data/ref_place_names.csv
trunk/opentrep/test/xapian/simple_search.cpp
trunk/opentrep/test/xapian/string_search.cpp
Modified: trunk/opentrep/opentrep/command/IndexBuilder.cpp
===================================================================
--- trunk/opentrep/opentrep/command/IndexBuilder.cpp 2009-07-14 22:34:14 UTC (rev 128)
+++ trunk/opentrep/opentrep/command/IndexBuilder.cpp 2009-07-15 14:45:43 UTC (rev 129)
@@ -7,6 +7,8 @@
#include <string>
#include <vector>
#include <exception>
+// Boost
+#include <boost/tokenizer.hpp>
// OpenTrep
#include <opentrep/bom/World.hpp>
#include <opentrep/bom/Place.hpp>
@@ -21,6 +23,30 @@
#include <xapian.h>
namespace OPENTREP {
+
+ // //////////////////////////////////////////////////////////////////////
+ void tokeniseAndAddToDocument (const std::string& iPhrase,
+ Xapian::Document& ioDocument,
+ Xapian::WritableDatabase& ioDatabase) {
+
+ // Boost Tokeniser
+ typedef boost::tokenizer<boost::char_separator<char> > Tokeniser_T;
+
+ // Define the separators
+ boost::char_separator<char> lSepatorList(" .,;:|+-*/_=!@#$%`~^&(){}[]?'<>\"");
+
+ // Initialise the phrase to be tokenised
+ Tokeniser_T lTokens (iPhrase, lSepatorList);
+ for (Tokeniser_T::const_iterator tok_iter = lTokens.begin();
+ tok_iter != lTokens.end(); ++tok_iter) {
+ const std::string& lTerm = *tok_iter;
+
+ ioDatabase.add_spelling (lTerm);
+ ioDocument.add_term (lTerm);
+
+ OPENTREP_LOG_DEBUG ("Added term: " << lTerm);
+ }
+ }
// //////////////////////////////////////////////////////////////////////
void IndexBuilder::
@@ -90,8 +116,9 @@
// extended, alternate, etc.)
if (lName.empty() == false) {
// OPENTREP_LOG_DEBUG ("Added name: " << lName);
- lDocument.add_term (lName); ++idx;
- ioDatabase.add_spelling (lName);
+ // lDocument.add_term (lName); ++idx;
+ // ioDatabase.add_spelling (lName);
+ tokeniseAndAddToDocument (lName, lDocument, ioDatabase);
}
}
}
Modified: trunk/opentrep/refdata/data/ref_place_names.csv
===================================================================
--- trunk/opentrep/refdata/data/ref_place_names.csv 2009-07-14 22:34:14 UTC (rev 128)
+++ trunk/opentrep/refdata/data/ref_place_names.csv 2009-07-15 14:45:43 UTC (rev 129)
@@ -1826,7 +1826,7 @@
en,jbt,bethel jbt,bethel jbt,bethel/ak/us:city landing
en,jca,cannes jca,cannes jca,cannes/fr:croisette hpt
en,jcb,joacaba,joacaba,joacaba/sc/br
-en,jcc,sanfrancisco jcc,sanfrancisco jc,san francisco/ca/us:china hpt
+en,jcc,san francisco jcc,san francisco jc,san francisco/ca/us:china hpt
en,jcd,st croix is jcd,st croix is jcd,st croix is/vi:downtown hpt
en,jce,convention,convention,convention/ca/us:heliport
en,jch,qasigiannguit,qasigiannguit,qasigiannguit/gl
Modified: trunk/opentrep/test/xapian/simple_search.cpp
===================================================================
--- trunk/opentrep/test/xapian/simple_search.cpp 2009-07-14 22:34:14 UTC (rev 128)
+++ trunk/opentrep/test/xapian/simple_search.cpp 2009-07-15 14:45:43 UTC (rev 129)
@@ -1,5 +1,6 @@
// STL
#include <iostream>
+#include <string>
// Xapian
#include <xapian.h>
@@ -7,46 +8,59 @@
int main (int argc, char* argv[]) {
// Simplest possible options parsing: we just require two or more
- // parameters.
- if (argc < 3) {
- std::cout << "Usage: " << argv[0]
- << " <path to database> <search terms>" << std::endl;
- return -1;
- }
+ // parameters.
+ if (argc < 3) {
+ std::cout << "Usage: " << argv[0]
+ << " <path to database> <search terms>" << std::endl;
+ return -1;
+ }
- // Catch any Xapian::Error exceptions thrown
- try {
+ // Catch any Xapian::Error exceptions thrown
+ try {
- // Make the database
- Xapian::Database db (argv[1]);
+ // Open the database for searching.
+ Xapian::Database db (argv[1]);
- // Start an enquire session
- Xapian::Enquire enquire (db);
+ // Start an enquire session
+ Xapian::Enquire enquire (db);
- // Build the query object
- Xapian::Query query (Xapian::Query::OP_AND, argv + 2, argv + argc);
- std::cout << "Performing query `" << query.get_description() << "'"
- << std::endl;
-
- // Give the query object to the enquire session
- enquire.set_query (query);
+ // Combine the rest of the command line arguments with spaces between
+ // them, so that simple queries don't have to be quoted at the shell
+ // level.
+ std::string query_string (argv[2]);
+ argv += 3;
+ while (*argv) {
+ query_string += ' ';
+ query_string += *argv++;
+ }
- // Get the top 10 results of the query
- Xapian::MSet matches = enquire.get_mset (0, 10);
+ // Parse the query string to produce a Xapian::Query object.
+ Xapian::QueryParser qp;
+ Xapian::Stem stemmer ("english");
+ qp.set_stemmer (stemmer);
+ qp.set_database (db);
+ qp.set_stemming_strategy (Xapian::QueryParser::STEM_SOME);
+ Xapian::Query query = qp.parse_query (query_string);
+ std::cout << "Parsed query is: " << query.get_description() << std::endl;
- // Display the results
- std::cout << matches.size() << " results found" << std::endl;
+ // Find the top 10 results for the query.
+ enquire.set_query (query);
+ Xapian::MSet matches = enquire.get_mset(0, 10);
- for (Xapian::MSetIterator i = matches.begin(); i != matches.end(); ++i) {
- Xapian::Document doc = i.get_document();
- std::cout << "Document ID " << *i << "\t" <<
- i.get_percent() << "% [" <<
- doc.get_data() << "]" << std::endl;
- }
+ // Display the results.
+ std::cout << matches.get_matches_estimated() << " results found."
+ << std::endl;
+ std::cout << "Matches 1-" << matches.size() << ":" << std::endl << std::endl;
+
+ for (Xapian::MSetIterator i = matches.begin(); i != matches.end(); ++i) {
+ std::cout << i.get_rank() + 1 << ": " << i.get_percent() << "% docid="
+ << *i << " [" << i.get_document().get_data() << "]"
+ << std::endl << std::endl;
+ }
- } catch (const Xapian::Error& error) {
- std::cerr << "Exception: " << error.get_msg() << std::endl;
- }
+ } catch (const Xapian::Error& error) {
+ std::cerr << "Exception: " << error.get_msg() << std::endl;
+ }
- return 0;
+ return 0;
}
Modified: trunk/opentrep/test/xapian/string_search.cpp
===================================================================
--- trunk/opentrep/test/xapian/string_search.cpp 2009-07-14 22:34:14 UTC (rev 128)
+++ trunk/opentrep/test/xapian/string_search.cpp 2009-07-15 14:45:43 UTC (rev 129)
@@ -41,7 +41,11 @@
// Build the query object
Xapian::QueryParser lQueryParser;
lQueryParser.set_database (db);
- lQueryParser.set_default_op (Xapian::Query::OP_NEAR);
+ // As explained in http://www.xapian.org/docs/queryparser.html,
+ // Xapian::Query::OP_ADJ is better than Xapian::Query::OP_PHRASE,
+ // but only available from version 1.0.13 of Xapian
+ // lQueryParser.set_default_op (Xapian::Query::OP_ADJ);
+ lQueryParser.set_default_op (Xapian::Query::OP_PHRASE);
std::cout << "Query parser `" << lQueryParser.get_description() << "'"
<< std::endl;
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-14 22:34:18
|
Revision: 128
http://opentrep.svn.sourceforge.net/opentrep/?rev=128&view=rev
Author: denis_arnaud
Date: 2009-07-14 22:34:14 +0000 (Tue, 14 Jul 2009)
Log Message:
-----------
[Test] Tested a few variations for Xapian string search.
Modified Paths:
--------------
trunk/opentrep/test/xapian/string_search.cpp
Property Changed:
----------------
trunk/opentrep/ternary_tree/
Property changes on: trunk/opentrep/ternary_tree
___________________________________________________________________
Modified: svn:ignore
- .libs
.deps
Makefile
Makefile.in
+ .libs
.deps
Makefile
Makefile.in
simple_tst
Modified: trunk/opentrep/test/xapian/string_search.cpp
===================================================================
--- trunk/opentrep/test/xapian/string_search.cpp 2009-07-14 14:07:29 UTC (rev 127)
+++ trunk/opentrep/test/xapian/string_search.cpp 2009-07-14 22:34:14 UTC (rev 128)
@@ -28,20 +28,20 @@
for (int idx=2; idx != argc; ++idx) {
if (idx != 2) {
oStr << " ";
-// oStr << " AND ";
}
const std::string lWord (argv[idx]);
- const std::string lSuggestedWord =
- db.get_spelling_suggestion (lWord, 3);
+ const std::string lSuggestedWord = db.get_spelling_suggestion(lWord, 3);
std::cout << "Word `" << lWord << "' ==> Suggested word `"
<< lSuggestedWord << "'" << std::endl;
oStr << lWord;
}
const std::string lQueryString = oStr.str();
+ std::cout << "QueryString `" << lQueryString << "'" << std::endl;
// Build the query object
Xapian::QueryParser lQueryParser;
lQueryParser.set_database (db);
+ lQueryParser.set_default_op (Xapian::Query::OP_NEAR);
std::cout << "Query parser `" << lQueryParser.get_description() << "'"
<< std::endl;
@@ -53,10 +53,11 @@
| Xapian::QueryParser::FLAG_LOVEHATE
| Xapian::QueryParser::FLAG_SPELLING_CORRECTION);
Xapian::Query lCorrectedQuery = lQueryParser.get_corrected_query_string();
-
+
std::cout << "Query `" << lQuery.get_description()
<< "', Corrected query `" << lCorrectedQuery.get_description()
- << "'" << std::endl;
+ << "' "
+ << std::endl;
// Give the query object to the enquire session
enquire.set_query (lQuery);
@@ -65,12 +66,19 @@
Xapian::MSet matches = enquire.get_mset (0, 10);
// Display the results
- int nbMatches = matches.size();
+ const int nbMatches = matches.size();
std::cout << nbMatches << " results found" << std::endl;
+
+ // if (true) {
if (nbMatches == 0) {
enquire.set_query (lCorrectedQuery);
matches = enquire.get_mset (0, 10);
+ //const Xapian::MSet matchesAll = enquire.get_mset ();
+ if (matches.size() == matches.max_size()) {
+ std::cout << "Corrected string matches all the documents"
+ << std::endl;
+ }
}
const Xapian::Query& lActualQuery = enquire.get_query();
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-14 14:07:33
|
Revision: 127
http://opentrep.svn.sourceforge.net/opentrep/?rev=127&view=rev
Author: denis_arnaud
Date: 2009-07-14 14:07:29 +0000 (Tue, 14 Jul 2009)
Log Message:
-----------
[TST] Updated the Ternary Structure Tree (TST). It still does not compile.
Modified Paths:
--------------
trunk/opentrep/configure.ac
trunk/opentrep/opentrep/Makefile.am
trunk/opentrep/ternary_tree/examples.cpp
trunk/opentrep/ternary_tree/iterator_compile_test.cpp
trunk/opentrep/ternary_tree/iterator_wrapper.hpp
trunk/opentrep/ternary_tree/readme.txt
trunk/opentrep/ternary_tree/structured_map.hpp
trunk/opentrep/ternary_tree/structured_set.hpp
trunk/opentrep/ternary_tree/ternary_tree.hpp
trunk/opentrep/ternary_tree/tst_concept_checks.cpp
trunk/opentrep/ternary_tree/tst_detail/iteration_impl.hpp
trunk/opentrep/ternary_tree/tst_detail/new_iterator_base.ipp
trunk/opentrep/ternary_tree/tst_detail/tst_implementation.ipp
trunk/opentrep/ternary_tree/tst_detail/tst_iterator_base.ipp
trunk/opentrep/ternary_tree/tst_detail/tst_iterator_facade.hpp
trunk/opentrep/ternary_tree/tst_detail/tst_node.hpp
trunk/opentrep/ternary_tree/tst_detail/tst_search_results.ipp
Added Paths:
-----------
trunk/opentrep/ternary_tree/Makefile.am
trunk/opentrep/ternary_tree/fill_dictionary.hpp
trunk/opentrep/ternary_tree/simple_tst.cpp
trunk/opentrep/ternary_tree/sources.mk
Removed Paths:
-------------
trunk/opentrep/ternary_tree/fill_dictionary.cpp
Property Changed:
----------------
trunk/opentrep/ternary_tree/
Modified: trunk/opentrep/configure.ac
===================================================================
--- trunk/opentrep/configure.ac 2009-07-14 10:45:19 UTC (rev 126)
+++ trunk/opentrep/configure.ac 2009-07-14 14:07:29 UTC (rev 127)
@@ -211,6 +211,7 @@
opentrep.pc
opentrep.spec
opentrep.m4
+ ternary_tree/Makefile
opentrep/Makefile
opentrep/basic/Makefile
opentrep/bom/Makefile
Modified: trunk/opentrep/opentrep/Makefile.am
===================================================================
--- trunk/opentrep/opentrep/Makefile.am 2009-07-14 10:45:19 UTC (rev 126)
+++ trunk/opentrep/opentrep/Makefile.am 2009-07-14 14:07:29 UTC (rev 127)
@@ -3,8 +3,6 @@
## Source directory
-DISTCLEANFILES = @PACKAGE@-paths.h
-
MAINTAINERCLEANFILES = Makefile.in
SUBDIRS = basic bom factory dbadaptor command service core config batches
Property changes on: trunk/opentrep/ternary_tree
___________________________________________________________________
Added: svn:ignore
+ .libs
.deps
Makefile
Makefile.in
Added: trunk/opentrep/ternary_tree/Makefile.am
===================================================================
--- trunk/opentrep/ternary_tree/Makefile.am (rev 0)
+++ trunk/opentrep/ternary_tree/Makefile.am 2009-07-14 14:07:29 UTC (rev 127)
@@ -0,0 +1,46 @@
+include $(top_srcdir)/Makefile.common
+include $(srcdir)/sources.mk
+
+##
+# Source directory
+
+DISTCLEANFILES =
+
+MAINTAINERCLEANFILES = Makefile.in
+
+SUBDIRS =
+
+EXTRA_DIST =
+
+
+##
+# Library
+lib_LTLIBRARIES = libtst.la
+
+libtst_la_SOURCES = $(tst_h_sources) $(tst_cc_sources)
+#libtst_la_LIBADD =
+libtst_la_LDFLAGS = -version-info $(GENERIC_LIBRARY_VERSION)
+
+# Header files
+nobase_pkginclude_HEADERS = $(ttree_h_sources)
+#nobase_nodist_pkginclude_HEADERS = $(top_builddir)/@PACKAGE@/config.h
+
+
+##
+# Binaries (batches)
+bin_PROGRAMS = simple_tst
+
+simple_tst_SOURCES = simple_tst.cpp
+#simple_tst_CXXFLAGS =
+#simple_tst_LDADD =
+#simple_tst_LDFLAGS =
+
+##
+# Test binaries
+#check_PROGRAMS = iterator_compile_test tst_concept_checks
+
+#iterator_compile_test_SOURCES = iterator_compile_test.cpp
+#iterator_compile_test_LDFLAGS =
+
+#tst_concept_checks_SOURCES = tst_concept_checks.cpp
+#tst_concept_checks_LDFLAGS =
Modified: trunk/opentrep/ternary_tree/examples.cpp
===================================================================
--- trunk/opentrep/ternary_tree/examples.cpp 2009-07-14 10:45:19 UTC (rev 126)
+++ trunk/opentrep/ternary_tree/examples.cpp 2009-07-14 14:07:29 UTC (rev 127)
@@ -1,231 +1,231 @@
-/** \file
- * Usage examples for Structured Containers.
- */
-
-#include <iostream>
-#include "structured_set.hpp"
-#include "structured_map.hpp"
-
-#include <iostream>
-#include <string>
-#include <set>
-#include <functional>
-//#include <boost/scoped_ptr.hpp>
-
-//
-// Basic use of structured_set
-//
-
-void basic()
-{
- typedef containers::structured_set<std::string> Set;
- typedef Set::iterator SetIter;
- typedef std::pair<SetIter, SetIter> IterPair;
-
- Set names;
- names.insert("apps");
- names.insert("applets");
- names.insert("banana");
-
- std::cout << "The set contains\n\t";
- for (SetIter it = names.begin(); it != names.end(); ++it)
- std::cout << *it << ", ";
-
- IterPair p = names.prefix_range("app");
- std::cout << "\nprefix_range(\"app\") returns:\n\t";
- while (p.first != p.second) {
- std::cout << *p.first++ << ", ";
- }
- std::cout << "\np.second points to " << *p.second;
-
- std::cout << "\nequal_range(\"app\") returns:\n\t";
- p = names.equal_range("app");
- if (p.first == p.second)
- std::cout << "empty range";
- std::cout << "\np.second points to " << *p.second;
-}
-
-
-//#############################################################################
-//
-// prefix_range example (compile-only, does not run as is)
-//
-typedef containers::structured_set<std::string> SymbolSet;
-SymbolSet symbols;
-bool is_defined_in_scope(std::string scope, std::string name)
-{
- typedef std::pair<SymbolSet::iterator, SymbolSet::iterator> Range;
- Range r = symbols.prefix_range(scope + "::");
- SymbolSet::iterator n = symbols.find(name);
- return n != symbols.end() && *n >= *r.first && *n < *r.second;
-}
-
-
-//#############################################################################
-//
-// Case-insensitive structured containers
-//
-template<class CharT>
-struct nocase_less : public std::binary_function<CharT, CharT, bool>
-{
- bool operator()(CharT a, CharT b) const { return tolower(a) < tolower(b); }
-};
-
-void caseless_set()
-{
- typedef containers::structured_multiset<std::string, nocase_less<char> > CaselessSet;
- typedef containers::structured_multimap<std::string, double, nocase_less<char> > CaselessMap;
-
- CaselessMap uncased;
- uncased.insert(std::make_pair("NoCase", 0.1));
- CaselessSet caseless;
- caseless.insert("NoCase");
- caseless.insert("nocase");
- caseless.insert("noCase");
- caseless.insert("NOCASE");
-
- std::cout << "nocase = " << (int)caseless.count("nocase");
-
- CaselessSet::const_iterator endit = caseless.end();
- for(CaselessSet::const_iterator it = caseless.begin(); it != endit; ++it) {
- std::cout << ", " << *it;
- }
-
-}
-
-//#############################################################################
-//
-// Localization comparator
-//
-
-#include "examples/locale_less.hpp"
-
-void localized_comparator()
-{
- typedef containers::structured_set<std::string, utility::locale_less<char> > LocalSet;
-
- typedef containers::structured_set<std::string> DefaultSet;
-
- if (utility::swedish_locale_name == "C")
- std::cout << "No locale to test\n";
- else
- std::cout << "Attempt to set Swedish locale \"" << utility::swedish_locale_name << "\"\n";
-
- try {
- // use comparator constructor, create Swedish locale
- LocalSet se_names(utility::locale_less<char>::locale_less(utility::swedish_locale_name));
- DefaultSet anynames;
-
- se_names.insert("\xC4ska");
- se_names.insert("\xC5m\xE5l");
- se_names.insert("\xD6dla");
- se_names.insert("Adam");
-
- anynames.insert("\xC4ska");
- anynames.insert("\xC5m\xE5l");
- anynames.insert("\xD6dla");
- anynames.insert("Adam");
-
-
- for(LocalSet::iterator sit = se_names.begin(); sit != se_names.end(); ++sit) {
- std::cout << *sit << ", ";
- }
- std::cout << "not:\n";
- for(DefaultSet::iterator dit = anynames.begin(); dit != anynames.end(); ++dit) {
- std::cout << *dit << ", ";
- }
- } catch(std::exception& x) {
- std::cout << "...failed - skip test\n" << x.what() << "\n";
- }
-}
-
-//#############################################################################
-//
-// longest_match example
-//
-#include <fstream>
-
-typedef containers::structured_map<std::string, int, nocase_less<char> > Vocabulary;
-
-void fill_wordlist(const char* filename, Vocabulary& wordlist)
-{
- std::ifstream wordstream(filename);
- if (!wordstream.is_open()) {
- std::cerr << "Could not open dictionary " << filename << "\n";
- return;
- }
- char buf[300];
- int linecount = 0;
- while(wordstream.getline(buf, 300, '\n').good())
- wordlist[buf] = ++linecount;
-}
-
-std::streamsize get_filesize(std::ifstream& str)
-{
- std::streamsize pos = str.tellg();
- str.seekg(0, std::ios_base::end);
- std::streamsize result = str.tellg();
- str.seekg(pos, std::ios_base::beg);
- return result;
-}
-
-namespace {
- template<class T>
- struct scoped_array
- {
- scoped_array(size_t count) : buf(new T[count]) {}
- ~scoped_array() { delete[] buf; }
- T* get() { return buf; }
- private:
- T* buf;
- };
-}
-
-void longest_match_example(const char* dictfile, const char* parsefile)
-{
- Vocabulary english;
- // Read dictionary from disk
- fill_wordlist(dictfile, english);
- if (english.empty())
- return;
-
- std::ifstream infile(parsefile);
- if (!infile.is_open())
- return;
-
- // longest_match does not work with istream_iterator, so must fill buffer
- size_t filesize = (size_t)get_filesize(infile);
- // instead of boost::scoped_array
- scoped_array<char> bytes(filesize);
- infile.read(bytes.get(), filesize);
-
- const char *first = bytes.get();
- const char *last = first + infile.gcount();
-
- while (first != last)
- {
- Vocabulary::iterator word = english.longest_match(first, last);
- if (word != english.end())
- std::cout << (*word).first << " "; //= " << (*word).second << "\n";
- else {
- // No key; try next char
- ++first;
- }
- }
-}
-
-//#############################################################################
-
-int main()
-{
- std::cout << "*** basic usage ***\n";
- basic();
- std::cout << "\n\n*** custom comparator ***\n";
- caseless_set();
- std::cout << "\n\n*** locale comparator ***\n";
- localized_comparator();
- std::cout << "\n\n*** longest_match ***\n";
- // You need to supply files, not included in ternary_tree distribution
- longest_match_example("../english-150k.txt", "../shakequotes.txt");
- return 0;
-}
+/** \file
+ * Usage examples for Structured Containers.
+ */
+
+#include <iostream>
+#include "structured_set.hpp"
+#include "structured_map.hpp"
+
+#include <iostream>
+#include <string>
+#include <set>
+#include <functional>
+//#include <boost/scoped_ptr.hpp>
+
+//
+// Basic use of structured_set
+//
+
+void basic()
+{
+ typedef containers::structured_set<std::string> Set;
+ typedef Set::iterator SetIter;
+ typedef std::pair<SetIter, SetIter> IterPair;
+
+ Set names;
+ names.insert("apps");
+ names.insert("applets");
+ names.insert("banana");
+
+ std::cout << "The set contains\n\t";
+ for (SetIter it = names.begin(); it != names.end(); ++it)
+ std::cout << *it << ", ";
+
+ IterPair p = names.prefix_range("app");
+ std::cout << "\nprefix_range(\"app\") returns:\n\t";
+ while (p.first != p.second) {
+ std::cout << *p.first++ << ", ";
+ }
+ std::cout << "\np.second points to " << *p.second;
+
+ std::cout << "\nequal_range(\"app\") returns:\n\t";
+ p = names.equal_range("app");
+ if (p.first == p.second)
+ std::cout << "empty range";
+ std::cout << "\np.second points to " << *p.second;
+}
+
+
+//#############################################################################
+//
+// prefix_range example (compile-only, does not run as is)
+//
+typedef containers::structured_set<std::string> SymbolSet;
+SymbolSet symbols;
+bool is_defined_in_scope(std::string scope, std::string name)
+{
+ typedef std::pair<SymbolSet::iterator, SymbolSet::iterator> Range;
+ Range r = symbols.prefix_range(scope + "::");
+ SymbolSet::iterator n = symbols.find(name);
+ return n != symbols.end() && *n >= *r.first && *n < *r.second;
+}
+
+
+//#############################################################################
+//
+// Case-insensitive structured containers
+//
+template<class CharT>
+struct nocase_less : public std::binary_function<CharT, CharT, bool>
+{
+ bool operator()(CharT a, CharT b) const { return tolower(a) < tolower(b); }
+};
+
+void caseless_set()
+{
+ typedef containers::structured_multiset<std::string, nocase_less<char> > CaselessSet;
+ typedef containers::structured_multimap<std::string, double, nocase_less<char> > CaselessMap;
+
+ CaselessMap uncased;
+ uncased.insert(std::make_pair("NoCase", 0.1));
+ CaselessSet caseless;
+ caseless.insert("NoCase");
+ caseless.insert("nocase");
+ caseless.insert("noCase");
+ caseless.insert("NOCASE");
+
+ std::cout << "nocase = " << (int)caseless.count("nocase");
+
+ CaselessSet::const_iterator endit = caseless.end();
+ for(CaselessSet::const_iterator it = caseless.begin(); it != endit; ++it) {
+ std::cout << ", " << *it;
+ }
+
+}
+
+//#############################################################################
+//
+// Localization comparator
+//
+
+#include "examples/locale_less.hpp"
+
+void localized_comparator()
+{
+ typedef containers::structured_set<std::string, utility::locale_less<char> > LocalSet;
+
+ typedef containers::structured_set<std::string> DefaultSet;
+
+ if (utility::swedish_locale_name == "C")
+ std::cout << "No locale to test\n";
+ else
+ std::cout << "Attempt to set Swedish locale \"" << utility::swedish_locale_name << "\"\n";
+
+ try {
+ // use comparator constructor, create Swedish locale
+ LocalSet se_names(utility::locale_less<char>::locale_less(utility::swedish_locale_name));
+ DefaultSet anynames;
+
+ se_names.insert("\xC4ska");
+ se_names.insert("\xC5m\xE5l");
+ se_names.insert("\xD6dla");
+ se_names.insert("Adam");
+
+ anynames.insert("\xC4ska");
+ anynames.insert("\xC5m\xE5l");
+ anynames.insert("\xD6dla");
+ anynames.insert("Adam");
+
+
+ for(LocalSet::iterator sit = se_names.begin(); sit != se_names.end(); ++sit) {
+ std::cout << *sit << ", ";
+ }
+ std::cout << "not:\n";
+ for(DefaultSet::iterator dit = anynames.begin(); dit != anynames.end(); ++dit) {
+ std::cout << *dit << ", ";
+ }
+ } catch(std::exception& x) {
+ std::cout << "...failed - skip test\n" << x.what() << "\n";
+ }
+}
+
+//#############################################################################
+//
+// longest_match example
+//
+#include <fstream>
+
+typedef containers::structured_map<std::string, int, nocase_less<char> > Vocabulary;
+
+void fill_wordlist(const char* filename, Vocabulary& wordlist)
+{
+ std::ifstream wordstream(filename);
+ if (!wordstream.is_open()) {
+ std::cerr << "Could not open dictionary " << filename << "\n";
+ return;
+ }
+ char buf[300];
+ int linecount = 0;
+ while(wordstream.getline(buf, 300, '\n').good())
+ wordlist[buf] = ++linecount;
+}
+
+std::streamsize get_filesize(std::ifstream& str)
+{
+ std::streamsize pos = str.tellg();
+ str.seekg(0, std::ios_base::end);
+ std::streamsize result = str.tellg();
+ str.seekg(pos, std::ios_base::beg);
+ return result;
+}
+
+namespace {
+ template<class T>
+ struct scoped_array
+ {
+ scoped_array(size_t count) : buf(new T[count]) {}
+ ~scoped_array() { delete[] buf; }
+ T* get() { return buf; }
+ private:
+ T* buf;
+ };
+}
+
+void longest_match_example(const char* dictfile, const char* parsefile)
+{
+ Vocabulary english;
+ // Read dictionary from disk
+ fill_wordlist(dictfile, english);
+ if (english.empty())
+ return;
+
+ std::ifstream infile(parsefile);
+ if (!infile.is_open())
+ return;
+
+ // longest_match does not work with istream_iterator, so must fill buffer
+ size_t filesize = (size_t)get_filesize(infile);
+ // instead of boost::scoped_array
+ scoped_array<char> bytes(filesize);
+ infile.read(bytes.get(), filesize);
+
+ const char *first = bytes.get();
+ const char *last = first + infile.gcount();
+
+ while (first != last)
+ {
+ Vocabulary::iterator word = english.longest_match(first, last);
+ if (word != english.end())
+ std::cout << (*word).first << " "; //= " << (*word).second << "\n";
+ else {
+ // No key; try next char
+ ++first;
+ }
+ }
+}
+
+//#############################################################################
+
+int main()
+{
+ std::cout << "*** basic usage ***\n";
+ basic();
+ std::cout << "\n\n*** custom comparator ***\n";
+ caseless_set();
+ std::cout << "\n\n*** locale comparator ***\n";
+ localized_comparator();
+ std::cout << "\n\n*** longest_match ***\n";
+ // You need to supply files, not included in ternary_tree distribution
+ longest_match_example("../english-150k.txt", "../shakequotes.txt");
+ return 0;
+}
Deleted: trunk/opentrep/ternary_tree/fill_dictionary.cpp
===================================================================
--- trunk/opentrep/ternary_tree/fill_dictionary.cpp 2009-07-14 10:45:19 UTC (rev 126)
+++ trunk/opentrep/ternary_tree/fill_dictionary.cpp 2009-07-14 14:07:29 UTC (rev 127)
@@ -1,66 +0,0 @@
-#include <vector>
-#include <string>
-#include <iostream>
-#include <fstream>
-#include <algorithm>
-#include <stdexcept>
-
-typedef std::vector<std::string> Dictionary;
-
-//template<class Container>
-size_t fill_dictionary(const char* filename, Dictionary& dictionary, size_t maxsize, size_t line_length = 0)
-{
- std::ifstream input(filename);
- size_t longest_in_file = 0;
- size_t linecount = 0;
- if (!input.is_open()) {
- std::cerr << filename << ": file open fail\n";
- throw std::runtime_error("fill_dictionary failed");
- }
- if (input.is_open() && !line_length)
- {
- std::vector<char> next;
- next.resize(std::max(line_length, size_t(300)));
- while(input.good() && linecount < maxsize) {
- input.getline(&next[0], next.capacity());
- std::string s(next.begin(), next.begin() + (size_t)input.gcount());
- if (!s.empty()) {
- dictionary.push_back(s.c_str());
- ++linecount;
- if (s.size() > longest_in_file)
- longest_in_file = s.size();
- //std::cerr << s.c_str() << "\n";
- }
- //next.clear();
- }
- //std::cerr << "Read " << dictionary.size() << " lines from wordlist.txt\n";
- }
-
- // If file not long enough, fill up with some random alphabetic strings
- if (line_length || (dictionary.size() < maxsize && (maxsize < size_t(-1)) ) )
- {
- std::string next;
- if (!line_length) {
- line_length = linecount? longest_in_file : 10;
- std::cerr << "zero-length line, we'll have trouble";
- }
- next.reserve(line_length + 1);
- for (size_t i = dictionary.size(); i < maxsize; ++i)
- {
- size_t length = 1 + (rand() % line_length);
- next.resize(length--);
- while(length--) {
- next[length] = rand() % (127-' ') + ' ';
- }
- //std::cerr << next << '\n';
- dictionary.push_back(next.c_str());
- }
- if (line_length > longest_in_file)
- longest_in_file = line_length;
- }
- return longest_in_file;
-}
-
-
-
-
Copied: trunk/opentrep/ternary_tree/fill_dictionary.hpp (from rev 126, trunk/opentrep/ternary_tree/fill_dictionary.cpp)
===================================================================
--- trunk/opentrep/ternary_tree/fill_dictionary.hpp (rev 0)
+++ trunk/opentrep/ternary_tree/fill_dictionary.hpp 2009-07-14 14:07:29 UTC (rev 127)
@@ -0,0 +1,66 @@
+#include <vector>
+#include <string>
+#include <iostream>
+#include <fstream>
+#include <algorithm>
+#include <stdexcept>
+
+typedef std::vector<std::string> Dictionary;
+
+//template<class Container>
+size_t fill_dictionary(const char* filename, Dictionary& dictionary, size_t maxsize, size_t line_length = 0)
+{
+ std::ifstream input(filename);
+ size_t longest_in_file = 0;
+ size_t linecount = 0;
+ if (!input.is_open()) {
+ std::cerr << filename << ": file open fail\n";
+ throw std::runtime_error("fill_dictionary failed");
+ }
+ if (input.is_open() && !line_length)
+ {
+ std::vector<char> next;
+ next.resize(std::max(line_length, size_t(300)));
+ while(input.good() && linecount < maxsize) {
+ input.getline(&next[0], next.capacity());
+ std::string s(next.begin(), next.begin() + (size_t)input.gcount());
+ if (!s.empty()) {
+ dictionary.push_back(s.c_str());
+ ++linecount;
+ if (s.size() > longest_in_file)
+ longest_in_file = s.size();
+ //std::cerr << s.c_str() << "\n";
+ }
+ //next.clear();
+ }
+ //std::cerr << "Read " << dictionary.size() << " lines from wordlist.txt\n";
+ }
+
+ // If file not long enough, fill up with some random alphabetic strings
+ if (line_length || (dictionary.size() < maxsize && (maxsize < size_t(-1)) ) )
+ {
+ std::string next;
+ if (!line_length) {
+ line_length = linecount? longest_in_file : 10;
+ std::cerr << "zero-length line, we'll have trouble";
+ }
+ next.reserve(line_length + 1);
+ for (size_t i = dictionary.size(); i < maxsize; ++i)
+ {
+ size_t length = 1 + (rand() % line_length);
+ next.resize(length--);
+ while(length--) {
+ next[length] = rand() % (127-' ') + ' ';
+ }
+ //std::cerr << next << '\n';
+ dictionary.push_back(next.c_str());
+ }
+ if (line_length > longest_in_file)
+ longest_in_file = line_length;
+ }
+ return longest_in_file;
+}
+
+
+
+
Modified: trunk/opentrep/ternary_tree/iterator_compile_test.cpp
===================================================================
--- trunk/opentrep/ternary_tree/iterator_compile_test.cpp 2009-07-14 10:45:19 UTC (rev 126)
+++ trunk/opentrep/ternary_tree/iterator_compile_test.cpp 2009-07-14 14:07:29 UTC (rev 127)
@@ -1,163 +1,163 @@
-/** Pure compilation/header test
- * \file
- * This file checks interoperability requirements for iterator_wrapper.hpp
- * The problem cases are those that should fail: they cannot be checked
- * automatically by compiler.
- * So to use, you must define TEST_COMPILATION_FAILURE or
- * CHECK_SPECIAL_COMP_FAILURE below, and then inspect compiler warnings
- * to see that you get an error for every SHOULD_FAIL line (#1-F, #20-21).
- * To simplify this, uncomment one statement at a time in the function
- * iterator_interop_checks_main()
- * at the end of this file, then try to compile.
- *
- * Construction/assignment to reverse_iterator from const_reverse_iterator
- * generates more complicated error messages from compilers,
- * so were broken out to allow separate runs.
- * Define the macro CHECK_SPECIAL_COMP_FAILURE and look for
- * SPECIAL_FAIL_1 and _2 in the compiler output.
- *
- * This (un)works as required with MSVC and Comeau online tryitout compiler.
- */
-
-#ifdef _MSC_VER
-# pragma warning(disable: 4245 4127 4189 4700)
-#endif
-
-
-//#define TEST_COMPILATION_FAILURE
-// Two cases must be inspected in references on MSVC
-//#define CHECK_SPECIAL_COMP_FAILURE
-
-#ifdef TEST_COMPILATION_FAILURE
-# define SHOULD_FAIL( Pred ) Pred
-#else
-# define SHOULD_FAIL( Pred )
-#endif
-
-#ifdef CHECK_SPECIAL_COMP_FAILURE
-# define SPECIAL_FAIL_1( Pred ) Pred
-# define SPECIAL_FAIL_2( Pred ) Pred
-#else
-# define SPECIAL_FAIL_1( Pred )
-# define SPECIAL_FAIL_2( Pred )
-#endif
-
-
-
-template<class Container>
-void iterator_interop_checks()
-{
- typedef Container C;
- typedef typename C::const_iterator c_t;
- typedef typename C::iterator i_t;
- typedef typename C::const_reverse_iterator cr_t;
- typedef typename C::reverse_iterator r_t;
- c_t c;
- i_t i;
- cr_t cr;
- r_t r;
-///// COPY-CONSTRUCTORS
- // Should work
- c_t i1( i );
- c_t i2( cr.base() );
- c_t i3( r.base() );
- i_t i4( r.base() );
- cr_t i5( r );
- cr_t i6( c );
- cr_t i7( i );
- r_t i8( i );
-
- SHOULD_FAIL( c_t i101( cr ); ) // #1
- SHOULD_FAIL( c_t i102( r ); ) // #2
- SHOULD_FAIL( i_t i103( c ); ) // #3
- SHOULD_FAIL( i_t i104( cr ); ) // #4
- SHOULD_FAIL( i_t i105( r ); ) // #5
- SHOULD_FAIL( r_t i106( c ); ) // #6
-
-///// ASSIGNMENT
- // Should work
- c = i;
- c = cr.base();
- c = r.base();
- i = r.base();
- cr = r;
-
- SHOULD_FAIL( c = cr; ) // #7
- SHOULD_FAIL( c = r; ) // #8
- SHOULD_FAIL( i = c; ) // #9
- SHOULD_FAIL( i = cr; ) // #A
- SHOULD_FAIL( i = r; ) // #B
- SHOULD_FAIL( cr = c; ) // #C
- SHOULD_FAIL( cr = i; ) // #D
- SHOULD_FAIL( r = c; ) // #E
- SHOULD_FAIL( r = i; ) // #F
-
-// these fail in 2nd pass or something, compile separately
- SPECIAL_FAIL_1( r_t i107( cr ); ) // #10
- SPECIAL_FAIL_2( r = cr; ) // #11
-
-///// ADVANCE
- ++c; --c;
- ++i; --i;
- ++cr; --cr;
- ++r; --r;
- c++; c--;
- i++; i--;
- cr++; cr--;
- r++; r--;
-
-///// COMPARE
- if (c == i && i == c) c = c;
- if (c == cr.base() && cr.base() == c) c = c;
- if (i == r.base() && r.base() == i) c = c;
- // Should this fail? - Dinkum nor StlPort don't seem to prevent it
- //if (r == cr && cr == r) c = c;
- if (r.base() == cr.base() && cr.base() == r.base()) c = c;
-
-///// DEREFERENCE
- typedef typename C::value_type val_t;
- typedef typename C::pointer ptr_t;
- typedef typename C::reference ref_t;
- typedef typename C::const_reference cref_t;
-
- val_t val1 = *c;
- val_t val2 = *i;
- val_t val3 = *cr;
- val_t val4 = *r;
-
- ref_t ref2 = *i;
- ref_t ref4 = *r;
-
- cref_t cref1 = *c;
- cref_t cref2 = *i;
- cref_t cref3 = *cr;
- cref_t cref4 = *r;
-
- SHOULD_FAIL( ref_t ref1 = *c ); // #20
- SHOULD_FAIL( ref_t ref3 = *cr ); // #21
-
-}
-
-#include <vector>
-#include "ternary_tree.hpp"
-#include "structured_set.hpp"
-#include "structured_map.hpp"
-
-void iterator_interop_checks_main()
-{
- typedef std::vector<int> Cont;
-// iterator_interop_checks<Cont>();
-/* typedef containers::ternary_tree<std::string, int> Tst;
- iterator_interop_checks<Tst>();
- typedef containers::structured_set<std::string> StrucSet;
- iterator_interop_checks<StrucSet>();
-*/ typedef containers::structured_multiset<std::string> MStrucSet;
- iterator_interop_checks<MStrucSet>();
-/* typedef containers::structured_map<std::string, int> StrucMap;
- iterator_interop_checks<StrucMap>();
- typedef containers::structured_multimap<std::string, int> MStrucMap;
- iterator_interop_checks<MStrucMap>(); */
-}
-
-
-
+/** Pure compilation/header test
+ * \file
+ * This file checks interoperability requirements for iterator_wrapper.hpp
+ * The problem cases are those that should fail: they cannot be checked
+ * automatically by compiler.
+ * So to use, you must define TEST_COMPILATION_FAILURE or
+ * CHECK_SPECIAL_COMP_FAILURE below, and then inspect compiler warnings
+ * to see that you get an error for every SHOULD_FAIL line (#1-F, #20-21).
+ * To simplify this, uncomment one statement at a time in the function
+ * iterator_interop_checks_main()
+ * at the end of this file, then try to compile.
+ *
+ * Construction/assignment to reverse_iterator from const_reverse_iterator
+ * generates more complicated error messages from compilers,
+ * so were broken out to allow separate runs.
+ * Define the macro CHECK_SPECIAL_COMP_FAILURE and look for
+ * SPECIAL_FAIL_1 and _2 in the compiler output.
+ *
+ * This (un)works as required with MSVC and Comeau online tryitout compiler.
+ */
+
+#ifdef _MSC_VER
+# pragma warning(disable: 4245 4127 4189 4700)
+#endif
+
+
+//#define TEST_COMPILATION_FAILURE
+// Two cases must be inspected in references on MSVC
+//#define CHECK_SPECIAL_COMP_FAILURE
+
+#ifdef TEST_COMPILATION_FAILURE
+# define SHOULD_FAIL( Pred ) Pred
+#else
+# define SHOULD_FAIL( Pred )
+#endif
+
+#ifdef CHECK_SPECIAL_COMP_FAILURE
+# define SPECIAL_FAIL_1( Pred ) Pred
+# define SPECIAL_FAIL_2( Pred ) Pred
+#else
+# define SPECIAL_FAIL_1( Pred )
+# define SPECIAL_FAIL_2( Pred )
+#endif
+
+
+
+template<class Container>
+void iterator_interop_checks()
+{
+ typedef Container C;
+ typedef typename C::const_iterator c_t;
+ typedef typename C::iterator i_t;
+ typedef typename C::const_reverse_iterator cr_t;
+ typedef typename C::reverse_iterator r_t;
+ c_t c;
+ i_t i;
+ cr_t cr;
+ r_t r;
+///// COPY-CONSTRUCTORS
+ // Should work
+ c_t i1( i );
+ c_t i2( cr.base() );
+ c_t i3( r.base() );
+ i_t i4( r.base() );
+ cr_t i5( r );
+ cr_t i6( c );
+ cr_t i7( i );
+ r_t i8( i );
+
+ SHOULD_FAIL( c_t i101( cr ); ) // #1
+ SHOULD_FAIL( c_t i102( r ); ) // #2
+ SHOULD_FAIL( i_t i103( c ); ) // #3
+ SHOULD_FAIL( i_t i104( cr ); ) // #4
+ SHOULD_FAIL( i_t i105( r ); ) // #5
+ SHOULD_FAIL( r_t i106( c ); ) // #6
+
+///// ASSIGNMENT
+ // Should work
+ c = i;
+ c = cr.base();
+ c = r.base();
+ i = r.base();
+ cr = r;
+
+ SHOULD_FAIL( c = cr; ) // #7
+ SHOULD_FAIL( c = r; ) // #8
+ SHOULD_FAIL( i = c; ) // #9
+ SHOULD_FAIL( i = cr; ) // #A
+ SHOULD_FAIL( i = r; ) // #B
+ SHOULD_FAIL( cr = c; ) // #C
+ SHOULD_FAIL( cr = i; ) // #D
+ SHOULD_FAIL( r = c; ) // #E
+ SHOULD_FAIL( r = i; ) // #F
+
+// these fail in 2nd pass or something, compile separately
+ SPECIAL_FAIL_1( r_t i107( cr ); ) // #10
+ SPECIAL_FAIL_2( r = cr; ) // #11
+
+///// ADVANCE
+ ++c; --c;
+ ++i; --i;
+ ++cr; --cr;
+ ++r; --r;
+ c++; c--;
+ i++; i--;
+ cr++; cr--;
+ r++; r--;
+
+///// COMPARE
+ if (c == i && i == c) c = c;
+ if (c == cr.base() && cr.base() == c) c = c;
+ if (i == r.base() && r.base() == i) c = c;
+ // Should this fail? - Dinkum nor StlPort don't seem to prevent it
+ //if (r == cr && cr == r) c = c;
+ if (r.base() == cr.base() && cr.base() == r.base()) c = c;
+
+///// DEREFERENCE
+ typedef typename C::value_type val_t;
+ typedef typename C::pointer ptr_t;
+ typedef typename C::reference ref_t;
+ typedef typename C::const_reference cref_t;
+
+ val_t val1 = *c;
+ val_t val2 = *i;
+ val_t val3 = *cr;
+ val_t val4 = *r;
+
+ ref_t ref2 = *i;
+ ref_t ref4 = *r;
+
+ cref_t cref1 = *c;
+ cref_t cref2 = *i;
+ cref_t cref3 = *cr;
+ cref_t cref4 = *r;
+
+ SHOULD_FAIL( ref_t ref1 = *c ); // #20
+ SHOULD_FAIL( ref_t ref3 = *cr ); // #21
+
+}
+
+#include <vector>
+#include "ternary_tree.hpp"
+#include "structured_set.hpp"
+#include "structured_map.hpp"
+
+void iterator_interop_checks_main()
+{
+ typedef std::vector<int> Cont;
+// iterator_interop_checks<Cont>();
+/* typedef containers::ternary_tree<std::string, int> Tst;
+ iterator_interop_checks<Tst>();
+ typedef containers::structured_set<std::string> StrucSet;
+ iterator_interop_checks<StrucSet>();
+*/ typedef containers::structured_multiset<std::string> MStrucSet;
+ iterator_interop_checks<MStrucSet>();
+/* typedef containers::structured_map<std::string, int> StrucMap;
+ iterator_interop_checks<StrucMap>();
+ typedef containers::structured_multimap<std::string, int> MStrucMap;
+ iterator_interop_checks<MStrucMap>(); */
+}
+
+
+
Modified: trunk/opentrep/ternary_tree/iterator_wrapper.hpp
===================================================================
--- trunk/opentrep/ternary_tree/iterator_wrapper.hpp 2009-07-14 10:45:19 UTC (rev 126)
+++ trunk/opentrep/ternary_tree/iterator_wrapper.hpp 2009-07-14 14:07:29 UTC (rev 127)
@@ -1,233 +1,233 @@
-// Created Mon Feb 06 13:20:01 2006
-#ifndef ITERATOR_WRAPPER_HPP_INCLUDED
-#define ITERATOR_WRAPPER_HPP_INCLUDED
-
-#include <iterator>
-
-namespace iterators {
-
- // This is mostly a lame ripoff from Boost.Iterator, to avoid the dependency...
-
- //! Standard type traits for const_iterators. \see iterator_wrapper
- template <class T>
- struct const_traits {
- typedef T value_type;
- typedef const T* pointer;
- typedef const T* const_pointer;
- typedef const T& reference;
- typedef const T& const_reference;
- };
-
- //! Standard type traits for (non-const) iterators. \see iterator_wrapper
- template <class T>
- struct nonconst_traits {
- typedef T value_type;
- typedef T* pointer;
- typedef const T* const_pointer;
- typedef T& reference;
- typedef const T& const_reference;
- };
-
- /** Creates a bidirectional iterator from a base implementation,
- * which is required to supply the interface \code
- * struct iter_impl_sample
- * {
- * typedef /impl-defined/ reference;
- * iter_impl_sample();
- * iter_impl_sample(/args/);
- * void increment();
- * void decrement();
- * reference dereference() const;
- * template<class OtherIter> bool equal(const OtherIter& rhs);
- * void swap(this_type& rhs);
- * }; \endcode
- * (This class is meant for iterators you control - if you need to adapt an existing iterator
- * with different interface, something like boost::iterator_facade is needed.)
- *
- * The first template parameter is the iterator implementation class.
- * iterator_wrapper does not inherit from this. The second parameter is either const_traits <T>
- * or nonconst_traits <T>, which provide the basic value_type related definitions.
- *
- * Note that Boost.Iterator will do the same job better, this was provided to avoid the dependency.
- * Future versions may move to Boost instead.
- *
- * \ingroup utilities
- */
- template< class BaseIterT
- , class TraitsT
- , class IterCatT = std::bidirectional_iterator_tag
- >
- struct iterator_wrapper
- {
- typedef BaseIterT base_iter;
- typedef TraitsT traits_type;
- typedef iterator_wrapper<BaseIterT, TraitsT, IterCatT> this_type;
-
- typedef typename TraitsT::value_type value_type;
- typedef typename TraitsT::pointer pointer;
- typedef typename TraitsT::reference reference;
- typedef typename TraitsT::const_reference const_reference;
-
- typedef IterCatT iterator_category;
- typedef ptrdiff_t difference_type;
- typedef size_t size_type;
-
- iterator_wrapper() {}
-
- //! Copy constructor for iterator and constructor from (non-const) iterator for const_iterator
- template<class SameBase>
- iterator_wrapper(const iterator_wrapper<SameBase, nonconst_traits<value_type>, IterCatT>& it)
- : m_iter(it.iter_base())
- {}
-
- iterator_wrapper(const iterator_wrapper& it) : m_iter(it.iter_base()) {}
-
- iterator_wrapper(const base_iter& it) : m_iter(it) {}
-
- reference operator*() const { return m_iter.dereference(); }
-
- pointer operator->() const { return &m_iter.dereference(); }
-
- this_type& operator++() { m_iter.increment(); return *this; }
-
- this_type operator++(int) {
- this_type tmp(*this);
- this->operator++();
- return tmp;
- }
-
- this_type& operator--() { m_iter.decrement(); return *this; }
-
- this_type operator--(int) {
- this_type tmp(*this);
- this->operator--();
- return tmp;
- }
-
- //! Assignment from non-const to const_iterator
- template<class SameBase>
- this_type& operator=(const iterator_wrapper<SameBase, nonconst_traits<value_type> >& rhs) {
- this_type(rhs).swap(*this);
- return *this;
- }
-
- this_type& operator=(const iterator_wrapper<BaseIterT, TraitsT>& rhs) {
- this_type(rhs).swap(*this);
- return *this;
- }
-
- template<class Base, class Constness>
- void swap(iterator_wrapper<Base, Constness>& other) {
- iter_base().swap(other.iter_base());
- }
-
- base_iter& iter_base() { return m_iter; }
- const base_iter& iter_base() const { return m_iter; }
-
- private:
- base_iter m_iter;
- };
-
- // \relates iterator_wrapper
- template<class Base, class Val, class Val2, class Cat>
- bool operator== (const iterator_wrapper<Base, Val, Cat>& lhs, const iterator_wrapper<Base, Val2, Cat>& rhs) {
- return lhs.iter_base().equal(rhs.iter_base());
- }
-
- // \relates iterator_wrapper
- template<class Base, class Val, class Val2, class Cat>
- bool operator!= (const iterator_wrapper<Base, Val, Cat>& lhs, const iterator_wrapper<Base, Val2, Cat>& rhs) {
- return ! (lhs == rhs);
- }
-
-//! \def provide equality operator for reverse const/nonconst iterators \relates iterator_wrapper
-#define INTEROPERABLE_REVERSE_ITERATOR_WRAPPERS(ConstTraits, NonConstTraits) \
- template<class Base, class Val, class Cat> \
- bool operator== (const std::reverse_iterator<iterator_wrapper<Base, ConstTraits<Val>, Cat> >& lhs, \
- const std::reverse_iterator<iterator_wrapper<Base, NonConstTraits<Val>, Cat> >& rhs) { \
- return lhs.base() == rhs.base(); \
- } \
- template<class Base, class Val, class Cat> \
- bool operator!= (const std::reverse_iterator<iterator_wrapper<Base, NonConstTraits<Val>, Cat> >& lhs, \
- const std::reverse_iterator<iterator_wrapper<Base, ConstTraits<Val>, Cat> >& rhs) { \
- return !(lhs.base() == rhs.base()); \
- }
-
-INTEROPERABLE_REVERSE_ITERATOR_WRAPPERS(const_traits, nonconst_traits)
-INTEROPERABLE_REVERSE_ITERATOR_WRAPPERS(nonconst_traits, const_traits)
-
-#undef INTEROPERABLE_REVERSE_ITERATOR_WRAPPERS
-
-
-/** \relates iterator_wrapper
- * @{
- */
-template<class Base, class Val, class Val2, class Cat>
-bool operator< (const iterator_wrapper<Base, Val, Cat>& lhs, const iterator_wrapper<Base, Val2, Cat>& rhs) {
- return lhs.iter_base().less(rhs.iter_base());
-}
-
-
-template<class Base, class Val, class Val2, class Cat>
-bool operator> (const iterator_wrapper<Base, Val, Cat>& lhs, const iterator_wrapper<Base, Val2, Cat>& rhs) {
- return rhs < lhs;
-}
-
-
-template<class Base, class Val, class Val2, class Cat>
-bool operator>= (const iterator_wrapper<Base, Val, Cat>& lhs, const iterator_wrapper<Base, Val2, Cat>& rhs) {
- return ! (lhs < rhs);
-}
-
-template<class Base, class Val, class Val2, class Cat>
-bool operator<= (const iterator_wrapper<Base, Val, Cat>& lhs, const iterator_wrapper<Base, Val2, Cat>& rhs) {
- return ! (rhs > lhs);
-}
-
-// random access iter operations (+= -= etc)
-
-template<class Base, class Val, class Dist>
-iterator_wrapper<Base, Val, std::random_access_iterator_tag>&
-operator+= (iterator_wrapper<Base, Val, std::random_access_iterator_tag>& it, Dist n) {
- it.iter_base().advance(n);
- return it;
-}
-
-template<class Base, class Val, class Dist>
-iterator_wrapper<Base, Val, std::random_access_iterator_tag>&
-operator-= (iterator_wrapper<Base, Val, std::random_access_iterator_tag>& it, Dist n) {
- it.iter_base().advance(-n);
- return it;
-}
-
-template<class Base, class Val, class Dist>
-iterator_wrapper<Base, Val, std::random_access_iterator_tag>
-operator+ (const iterator_wrapper<Base, Val, std::random_access_iterator_tag>& it, Dist n) {
- iterator_wrapper<Base, Val, std::random_access_iterator_tag> tmp(it);
- tmp += n;
- return tmp;
-}
-
-template<class Base, class Val, class Dist>
-iterator_wrapper<Base, Val, std::random_access_iterator_tag>
-operator- (const iterator_wrapper<Base, Val, std::random_access_iterator_tag>& it, Dist n) {
- iterator_wrapper<Base, Val, std::random_access_iterator_tag> tmp(it);
- tmp -= n;
- return tmp;
-}
-
-template<class Base, class Val, class Dist>
-Dist
-operator- (const iterator_wrapper<Base, Val, std::random_access_iterator_tag>& lhs,
- const iterator_wrapper<Base, Val, std::random_access_iterator_tag>& rhs) {
- return lhs - rhs;
-}
-
-
-/*@}*/
-
-} // namespace iterators
-
-
-
-#endif // ITERATOR_WRAPPER_HPP_INCLUDED
+// Created Mon Feb 06 13:20:01 2006
+#ifndef ITERATOR_WRAPPER_HPP_INCLUDED
+#define ITERATOR_WRAPPER_HPP_INCLUDED
+
+#include <iterator>
+
+namespace iterators {
+
+ // This is mostly a lame ripoff from Boost.Iterator, to avoid the dependency...
+
+ //! Standard type traits for const_iterators. \see iterator_wrapper
+ template <class T>
+ struct const_traits {
+ typedef T value_type;
+ typedef const T* pointer;
+ typedef const T* const_pointer;
+ typedef const T& reference;
+ typedef const T& const_reference;
+ };
+
+ //! Standard type traits for (non-const) iterators. \see iterator_wrapper
+ template <class T>
+ struct nonconst_traits {
+ typedef T value_type;
+ typedef T* pointer;
+ typedef const T* const_pointer;
+ typedef T& reference;
+ typedef const T& const_reference;
+ };
+
+ /** Creates a bidirectional iterator from a base implementation,
+ * which is required to supply the interface \code
+ * struct iter_impl_sample
+ * {
+ * typedef /impl-defined/ reference;
+ * iter_impl_sample();
+ * iter_impl_sample(/args/);
+ * void increment();
+ * void decrement();
+ * reference dereference() const;
+ * template<class OtherIter> bool equal(const OtherIter& rhs);
+ * void swap(this_type& rhs);
+ * }; \endcode
+ * (This class is meant for iterators you control - if you need to adapt an existing iterator
+ * with different interface, something like boost::iterator_facade is needed.)
+ *
+ * The first template parameter is the iterator implementation class.
+ * iterator_wrapper does not inherit from this. The second parameter is either const_traits <T>
+ * or nonconst_traits <T>, which provide the basic value_type related definitions.
+ *
+ * Note that Boost.Iterator will do the same job better, this was provided to avoid the dependency.
+ * Future versions may move to Boost instead.
+ *
+ * \ingroup utilities
+ */
+ template< class BaseIterT
+ , class TraitsT
+ , class IterCatT = std::bidirectional_iterator_tag
+ >
+ struct iterator_wrapper
+ {
+ typedef BaseIterT base_iter;
+ typedef TraitsT traits_type;
+ typedef iterator_wrapper<BaseIterT, TraitsT, IterCatT> this_type;
+
+ typedef typename TraitsT::value_type value_type;
+ typedef typename TraitsT::pointer pointer;
+ typedef typename TraitsT::reference reference;
+ typedef typename TraitsT::const_reference const_reference;
+
+ typedef IterCatT iterator_category;
+ typedef ptrdiff_t difference_type;
+ typedef size_t size_type;
+
+ iterator_wrapper() {}
+
+ //! Copy constructor for iterator and constructor from (non-const) iterator for const_iterator
+ template<class SameBase>
+ iterator_wrapper(const iterator_wrapper<SameBase, nonconst_traits<value_type>, IterCatT>& it)
+ : m_iter(it.iter_base())
+ {}
+
+ iterator_wrapper(const iterator_wrapper& it) : m_iter(it.iter_base()) {}
+
+ iterator_wrapper(const base_iter& it) : m_iter(it) {}
+
+ reference operator*() const { return m_iter.dereference(); }
+
+ pointer operator->() const { return &m_iter.dereference(); }
+
+ this_type& operator++() { m_iter.increment(); return *this; }
+
+ this_type operator++(int) {
+ this_type tmp(*this);
+ this->operator++();
+ return tmp;
+ }
+
+ this_type& operator--() { m_iter.decrement(); return *this; }
+
+ this_type operator--(int) {
+ this_type tmp(*this);
+ this->operator--();
+ return tmp;
+ }
+
+ //! Assignment from non-const to const_iterator
+ template<class SameBase>
+ this_type& operator=(const iterator_wrapper<SameBase, nonconst_traits<value_type> >& rhs) {
+ this_type(rhs).swap(*this);
+ return *this;
+ }
+
+ this_type& operator=(const iterator_wrapper<BaseIterT, TraitsT>& rhs) {
+ this_type(rhs).swap(*this);
+ return *this;
+ }
+
+ template<class Base, class Constness>
+ void swap(iterator_wrapper<Base, Constness>& other) {
+ iter_base().swap(other.iter_base());
+ }
+
+ base_iter& iter_base() { return m_iter; }
+ const base_iter& iter_base() const { return m_iter; }
+
+ private:
+ base_iter m_iter;
+ };
+
+ // \relates iterator_wrapper
+ template<class Base, class Val, class Val2, class Cat>
+ bool operator== (const iterator_wrapper<Base, Val, Cat>& lhs, const iterator_wrapper<Base, Val2, Cat>& rhs) {
+ return lhs.iter_base().equal(rhs.iter_base());
+ }
+
+ // \relates iterator_wrapper
+ template<class Base, class Val, class Val2, class Cat>
+ bool operator!= (const iterator_wrapper<Base, Val, Cat>& lhs, const iterator_wrapper<Base, Val2, Cat>& rhs) {
+ return ! (lhs == rhs);
+ }
+
+//! \def provide equality operator for reverse const/nonconst iterators \relates iterator_wrapper
+#define INTEROPERABLE_REVERSE_ITERATOR_WRAPPERS(ConstTraits, NonConstTraits) \
+ template<class Base, class Val, class Cat> \
+ bool operator== (const std::reverse_iterator<iterator_wrapper<Base, ConstTraits<Val>, Cat> >& lhs, \
+ const std::reverse_iterator<iterator_wrapper<Base, NonConstTraits<Val>, Cat> >& rhs) { \
+ return lhs.base() == rhs.base(); \
+ } \
+ template<class Base, class Val, class Cat> \
+ bool operator!= (const std::reverse_iterator<iterator_wrapper<Base, NonConstTraits<Val>, Cat> >& lhs, \
+ const std::reverse_iterator<iterator_wrapper<Base, ConstTraits<Val>, Cat> >& rhs) { \
+ return !(lhs.base() == rhs.base()); \
+ }
+
+INTEROPERABLE_REVERSE_ITERATOR_WRAPPERS(const_traits, nonconst_traits)
+INTEROPERABLE_REVERSE_ITERATOR_WRAPPERS(nonconst_traits, const_traits)
+
+#undef INTEROPERABLE_REVERSE_ITERATOR_WRAPPERS
+
+
+/** \relates iterator_wrapper
+ * @{
+ */
+template<class Base, class Val, class Val2, class Cat>
+bool operator< (const iterator_wrapper<Base, Val, Cat>& lhs, const iterator_wrapper<Base, Val2, Cat>& rhs) {
+ return lhs.iter_base().less(rhs.iter_base());
+}
+
+
+template<class Base, class Val, class Val2, class Cat>
+bool operator> (const iterator_wrapper<Base, Val, Cat>& lhs, const iterator_wrapper<Base, Val2, Cat>& rhs) {
+ return rhs < lhs;
+}
+
+
+template<class Base, class Val, class Val2, class Cat>
+bool operator>= (const iterator_wrapper<Base, Val, Cat>& lhs, const iterator_wrapper<Base, Val2, Cat>& rhs) {
+ return ! (lhs < rhs);
+}
+
+template<class Base, class Val, class Val2, class Cat>
+bool operator<= (const iterator_wrapper<Base, Val, Cat>& lhs, const iterator_wrapper<Base, Val2, Cat>& rhs) {
+ return ! (rhs > lhs);
+}
+
+// random access iter operations (+= -= etc)
+
+template<class Base, class Val, class Dist>
+iterator_wrapper<Base, Val, std::random_access_iterator_tag>&
+operator+= (iterator_wrapper<Base, Val, std::random_access_iterator_tag>& it, Dist n) {
+ it.iter_base().advance(n);
+ return it;
+}
+
+template<class Base, class Val, class Dist>
+iterator_wrapper<Base, Val, std::random_access_iterator_tag>&
+operator-= (iterator_wrapper<Base, Val, std::random_access_iterator_tag>& it, Dist n) {
+ it.iter_base().advance(-n);
+ return it;
+}
+
+template<class Base, class Val, class Dist>
+iterator_wrapper<Base, Val, std::random_access_iterator_tag>
+operator+ (const iterator_wrapper<Base, Val, std::random_access_iterator_tag>& it, Dist n) {
+ iterator_wrapper<Base, Val, std::random_access_iterator_tag> tmp(it);
+ tmp += n;
+ return tmp;
+}
+
+template<class Base, class Val, class Dist>
+iterator_wrapper<Base, Val, std::random_access_iterator_tag>
+operator- (const iterator_wrapper<Base, Val, std::random_access_iterator_tag>& it, Dist n) {
+ iterator_wrapper<Base, Val, std::random_access_iterator_tag> tmp(it);
+ tmp -= n;
+ return tmp;
+}
+
+template<class Base, class Val, class Dist>
+Dist
+operator- (const iterator_wrapper<Base, Val, std::random_access_iterator_tag>& lhs,
+ const iterator_wrapper<Base, Val, std::random_access_iterator_tag>& rhs) {
+ return lhs - rhs;
+}
+
+
+/*@}*/
+
+} // namespace iterators
+
+
+
+#endif // ITERATOR_WRAPPER_HPP_INCLUDED
Modified: trunk/opentrep/ternary_tree/readme.txt
===================================================================
--- trunk/opentrep/ternary_tree/readme.txt 2009-07-14 10:45:19 UTC (rev 126)
+++ trunk/opentrep/ternary_tree/readme.txt 2009-07-14 14:07:29 UTC (rev 127)
@@ -1,71 +1,71 @@
-LIBRARY
-Ternary Search Tree C++ implementation by rasmus ekman
-A header-only library of fast string containers with advanced search features.
-
-version 0.67, 14 May 2006
-
-Please send bug reports, suggestions or questions to ras...@ab...
-Get latest version at http://abc.se/~re/code/tst/
-
-REQUIREMENTS
-Library files tested with g++ 3.4.3 and MSVC 7.1 (Visual Studio 2003).
-Visual Studio 6 will not work, but may only need moving the template
-methods of ternary_tree and structured_* classes inline.
-To generate documentation, you need Doxygen. See http://doxygen.org.
-
-
-USAGE
-Container classes:
- structured_set<Key [, Comp, Alloc]>
- structured_multiset<Key [, Comp, Alloc]>
- structured_map<Key, Value [, Comp, Alloc]>
- structured_multimap<Key, Value [, Comp, Alloc]>
-- Key is a std::string-like type (a Forward Container),
-- Value is any type,
-- Comp is a less-like sort operation on Key::value_type (eg char/wchar_t)
-- Alloc is std::allocator<Key [, Value]> or has same interface.
-
-These containers can be used as nearly drop-in replacments for std::set,
-multiset, map, multimap or unordered_* containers on string-like types.
-
-There is one difference in interface:
-If you used non-default comparator template argument with a set or map type,
-it must be changed to operate on character type, not string.
-
-To use standard set, map features: See documentation of these classes.
-See included documentation for information about the advanced key search
-facilities in all structured_* containers.
-
-
-FILES
---- l i b r a r y c o d e ---
-structured_map.hpp - classes structured_map and -multimap.
-structured_set.hpp - classes structured_set and multiset.
-ternary_tree.hpp - implementation backend class.
-./tst_detail/.* ternary_tree implementation files
-iterator_wrapper.hpp - iterator interface, included by all containers.
-
---- d o c s ---
-tst_public.doxy - doxygen config file, generates public interface of library.
-tst.doxy - doxygen config file, generates public and private interface docs.
-./doxygen_input/* - extra documentation sources used by Doxygen.
-
-./html/* - generated public and private documentation
-full-docs-index.html - redirects to html directory
-index.html - redirects to tst_docs directory, only useful if doxygen is used
- with tst_public.doxy
-
---- t e s t s ---
-tst_concept_checks.cpp - requires Boost concept_check header, portable.
-iterator_compile_test.cpp - checks iterator_wrapper interoperability, portable.
-
-test_tst.cpp - test suite; relies on non-portable wstring.hpp (but see below)
-fill_dictionary.cpp - (sloppy old support file for test_tst)
- fills a vector with strings from file.
-wstring.hpp - string/wstring conversion, uses Windows API
- if the MultiByteToWideChar and WideCharToMultiByte API calls are replaced,
- test_tst may run on your platform...
-
----
-rasmus ekman
-May 14, 2006
+LIBRARY
+Ternary Search Tree C++ implementation by rasmus ekman
+A header-only library of fast string containers with advanced search features.
+
+version 0.67, 14 May 2006
+
+Please send bug reports, suggestions or questions to ras...@ab...
+Get latest version at http://abc.se/~re/code/tst/
+
+REQUIREMENTS
+Library files tested with g++ 3.4.3 and MSVC 7.1 (Visual Studio 2003).
+Visual Studio 6 will not work, but may only need moving the template
+methods of ternary_tree and structured_* classes inline.
+To generate documentation, you need Doxygen. See http://doxygen.org.
+
+
+USAGE
+Container classes:
+ structured_set<Key [, Comp, Alloc]>
+ structured_multiset<Key [, Comp, Alloc]>
+ structured_map<Key, Value [, Comp, Alloc]>
+ structured_multimap<Key, Value [, Comp, Alloc]>
+- Key is a std::string-like type (a Forward Container),
+- Value is any type,
+- Comp is a less-like sort operation on Key::value_type (eg char/wchar_t)
+- Alloc is std::allocator<Key [, Value]> or has same interface.
+
+These containers can be used as nearly drop-in replacments for std::set,
+multiset, map, multimap or unordered_* containers on string-like types.
+
+There is one difference in interface:
+If you used non-default comparator template argument with a set or map type,
+it must be changed to operate on character type, not string.
+
+To use standard set, map features: See documentation of these classes.
+See included documentation for information about the advanced key search
+facilities in all structured_* containers.
+
+
+FILES
+--- l i b r a r y c o d e ---
+structured_map.hpp - classes structured_map and -multimap.
+structured_set.hpp - classes structured_set and multiset.
+ternary_tree.hpp - implementation backend class.
+./tst_detail/.* ternary_tree implementation files
+iterator_wrapper.hpp - iterator interface, included by all containers.
+
+--- d o c s ---
+tst_public.doxy - doxygen config file, generates public interface of library.
+tst.doxy - doxygen config file, generates public and private interface docs.
+./doxygen_input/* - extra documentation sources used by Doxygen.
+
+./html/* - generated public and private documentation
+full-docs-index.html - redirects to html directory
+index.html - redirects to tst_docs directory, only useful if doxygen is used
+ with tst_public.doxy
+
+--- t e s t s ---
+tst_concept_checks.cpp - requires Boost concept_check header, portable.
+iterator_compile_test.cpp - checks iterator_wrapper interoperability, portable.
+
+test_tst.cpp - test suite; relies on non-portable wstring.hpp (but see below)
+fill_dictionary.cpp - (sloppy old support file for test_tst)
+ fills a vector with strings from file.
+wstring.hpp - string/wstring conversion, uses Windows API
+ if the MultiByteToWideChar and WideCharToMultiByte API calls are replaced,
+ test_tst may run on your platform...
+
+---
+rasmus ekman
+May 14, 2006
Added: trunk/opentrep/ternary_tree/simple_tst.cpp
===================================================================
--- trunk/opentrep/ternary_tree/simple_tst.cpp (rev 0)
+++ trunk/opentrep/ternary_tree/simple_tst.cpp 2009-07-14 14:07:29 UTC (rev 127)
@@ -0,0 +1,15 @@
+// C
+#include <cassert>
+// STL
+#include <iostream>
+// Ternary Tree Structure (TST)
+//#include <structured_set.hpp>
+
+// /////////////// M A I N ///////////////
+int main (int argc, char* argv[]) {
+
+ std::cout << "Hello TST!" << std::endl;
+
+ return 0;
+}
+
Added: trunk/opentrep/ternary_tree/sources.mk
===================================================================
--- trunk/opentrep/ternary_tree/sources.mk (rev 0)
+++ trunk/opentrep/ternary_tree/sources.mk 2009-07-14 14:07:29 UTC (rev 127)
@@ -0,0 +1,6 @@
+tst_h_sources = \
+ $(top_srcdir)/ternary_tree/ternary_tree.hpp \
+ $(top_srcdir)/ternary_tree/structured_set.hpp \
+ $(top_srcdir)/ternary_tree/structured_map.hpp \
+ $(top_srcdir)/ternary_tree/iterator_wrapper.hpp
+tst_cc_sources =
Modified: trunk/opentrep/ternary_tree/structured_map.hpp
===================================================================
--- trunk/opentrep/ternary_tree/structured_map.hpp 2009-07-14 10:45:19 UTC (rev 126)
+++ trunk/opentrep/ternary_tree/structured_map.hpp 2009-07-14 14:07:29 UTC (rev 127)
@@ -1,938 +1,938 @@
-#ifndef STRUCTURED_MAP_HPP_INCLUDED
-#define STRUCTURED_MAP_HPP_INCLUDED
-
-#define TST_NO_STANDALONE_ITERATOR_FACADE
-#include "ternary_tree.hpp"
-#undef TST_NO_STANDALONE_ITERATOR_FACADE
-// note: also #include <list> in structured_multimap section
-
-namespace containers {
-
-
-/** Structured Map is a Sorted Associative Container that stores objects of type pair<Key, Data>.
- * Structured Map is a Structured Container, meaning that its key type is required to be
- * a Forward Container, and that the map uses a comparator to establish
- * a strict weak ordering among key::value_type elements (rather than on whole keys).
- * This allows searches in the set involving parts of keys, ie with shared prefix
- * or with shared middle parts.
- *
- * Structured Map is a Pair Associative Container, meaning that its value type
- * is pair<const Key, T>.
- * It is also a Unique Associative Container, meaning that no two elements are the same.
- *
- * A std::map is normally backed by a binary tree. A structured map is instead backed
- * by a ternary_tree, which manages structured ordering of keys.
- * For string-like keys, a ternary tree is typically as fast as an unordered_map,
- * and several times faster than most std::map implementations.
- * \ingroup containers
- */
-template<class KeyT, class DataT, class CompT = std::less<typename KeyT::value_type>,
- class AllocT = std::allocator<std::pair<const KeyT, DataT> > >
-class structured_map
-{
-public:
- typedef KeyT key_type;
- typedef DataT mapped_type;
- typedef std::pair<const KeyT, DataT> value_type;
- typedef typename KeyT::value_type char_type;
- typedef CompT char_compare;
-
- typedef AllocT allocator_type;
- typedef typename AllocT::difference_type difference_type;
- typedef typename AllocT::size_type size_type;
- typedef typename AllocT::pointer pointer;
- typedef typename AllocT::const_pointer const_pointer;
- typedef typename AllocT::reference reference;
- typedef typename AllocT::const_reference const_reference;
-
-private:
- // Internal value type is pair<Key, Value> (non-const Key).
- typedef ternary_tree< KeyT, value_type, CompT,
- typename AllocT::template rebind<value_type>::other
- > ternary_tree;
- typedef typename ternary_tree::iterator tst_iterator;
- typedef typename ternary_tree::iterator::base_iter tst_iterator_base;
- typedef typename ternary_tree::const_iterator tst_const_iterator;
-
- enum { invalid_index = size_type(-1) };
-
-public:
-
- typedef typename ternary_tree::key_compare key_compare;
-
-
- typedef iterators::iterator_wrapper < tst_iterator_base
- , iterators::nonconst_traits<value_type>
- > iterator;
-
- typedef iterators::iterator_wrapper < tst_iterator_base
- , iterators::const_traits<value_type>
- > const_iterator;
-
- typedef std::reverse_iterator<iterator> reverse_iterator;
- typedef std::reverse_iterator<const_iterator> const_reverse_iterator;
-
-
- /** \name Construct, copy, destroy
- * @{
- */
- structured_map() : m_tree(char_compare(), allocator_type()) {}
-
- explicit structured_map(const char_compare& comp)
- : m_tree(comp, allocator_type())
- {}
-
- structured_map(const char_compare& comp, const allocator_type& alloc)
- : m_tree(comp, alloc)
- {}
-
- template<class InputIterator>
- structured_map( InputIterator first, InputIterator last,
- const char_compare& comp = char_compare(),
- const allocator_type& alloc = allocator_type())
- : m_tree(comp, alloc)
- {
- insert(first, last);
- }
-
- structured_map(const structured_map& other)
- : m_tree(other.m_tree)
- {}
-
- ~structured_map() {}
-
- structured_map& operator= (const structured_map& other) {
- structured_map(other).swap(*this);
- return *this;
- }
-
- allocator_type get_allocator() const { return m_tree.get_allocator(); }
- /* @} */
-
- /** \name Iterators
- * Includes C++0x methods cbegin, cend, crbegin, crend to make it easier
- * to access const iterators.
- * @{
- */
- iterator begin() { return iterator(m_tree.begin()); }
- const_iterator begin() const { return const_iterator(m_tree.begin()); }
- iterator end() { return iterator(m_tree.end()); }
- const_iterator end() const { return const_iterator(m_tree.end()); }
-
- reverse_iterator rbegin() { return reverse_iterator(end()); }
- const_reverse_iterator rbegin() const { return const_reverse_iterator(end()); }
- reverse_iterator rend() { return reverse_iterator(begin()); }
- const_reverse_iterator rend() const { return const_reverse_iterator(begin()); }
-
- // C++0x additions
- const_iterator cbegin() const { return const_iterator(m_tree.begin()); }
- const_iterator cend() const { return const_iterator(m_...
[truncated message content] |
|
From: <den...@us...> - 2009-07-14 10:45:31
|
Revision: 126
http://opentrep.svn.sourceforge.net/opentrep/?rev=126&view=rev
Author: denis_arnaud
Date: 2009-07-14 10:45:19 +0000 (Tue, 14 Jul 2009)
Log Message:
-----------
[Ternary Trees] Added the ternary trees structure.
Added Paths:
-----------
trunk/opentrep/ternary_tree/
trunk/opentrep/ternary_tree/README
trunk/opentrep/ternary_tree/doxygen_input/
trunk/opentrep/ternary_tree/doxygen_input/blather.hpp
trunk/opentrep/ternary_tree/doxygen_input/concepts.txt
trunk/opentrep/ternary_tree/doxygen_input/doxygen-old.css
trunk/opentrep/ternary_tree/doxygen_input/doxygen.css
trunk/opentrep/ternary_tree/doxygen_input/external.png
trunk/opentrep/ternary_tree/doxygen_input/featuretable.html
trunk/opentrep/ternary_tree/doxygen_input/footer_inc.html
trunk/opentrep/ternary_tree/doxygen_input/header_inc.html
trunk/opentrep/ternary_tree/doxygen_input/performancetable.html
trunk/opentrep/ternary_tree/doxygen_input/tree - trie concepts.txt
trunk/opentrep/ternary_tree/doxygen_input/usage.hpp
trunk/opentrep/ternary_tree/examples/
trunk/opentrep/ternary_tree/examples/examples.vcproj
trunk/opentrep/ternary_tree/examples/locale_less.hpp
trunk/opentrep/ternary_tree/examples.cpp
trunk/opentrep/ternary_tree/fill_dictionary.cpp
trunk/opentrep/ternary_tree/full-docs-index.html
trunk/opentrep/ternary_tree/html/
trunk/opentrep/ternary_tree/html/annotated.html
trunk/opentrep/ternary_tree/html/blather_8hpp.html
trunk/opentrep/ternary_tree/html/class_data_t.html
trunk/opentrep/ternary_tree/html/class_data_t_01_5.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1search__results__list-members.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1search__results__list.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1search__results__list_1_1iterator-members.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1search__results__list_1_1iterator.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1structured__map-members.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1structured__map.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1structured__map_1_1value__compare-members.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1structured__map_1_1value__compare.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1structured__multimap-members.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1structured__multimap.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1structured__multimap_1_1value__compare-members.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1structured__multimap_1_1value__compare.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1structured__multiset-members.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1structured__multiset.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1structured__set-members.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1structured__set.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1ternary__tree-members.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1ternary__tree.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1ternary__tree_1_1key__compare-members.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1ternary__tree_1_1key__compare.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1tst__detail_1_1_base_t.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1tst__detail_1_1tst__iterator__base-members.html
trunk/opentrep/ternary_tree/html/classcontainers_1_1tst__detail_1_1tst__iterator__base.html
trunk/opentrep/ternary_tree/html/classstd_1_1back__insert__iterator.html
trunk/opentrep/ternary_tree/html/classstd_1_1binary__function.html
trunk/opentrep/ternary_tree/html/dir_0df55976ff011c1ef61da79183e9e28f.html
trunk/opentrep/ternary_tree/html/dir_59457a7c227558cb0e28f31428e14f54.html
trunk/opentrep/ternary_tree/html/dirs.html
trunk/opentrep/ternary_tree/html/doxygen.css
trunk/opentrep/ternary_tree/html/doxygen.png
trunk/opentrep/ternary_tree/html/files.html
trunk/opentrep/ternary_tree/html/functions.html
trunk/opentrep/ternary_tree/html/functions_0x62.html
trunk/opentrep/ternary_tree/html/functions_0x63.html
trunk/opentrep/ternary_tree/html/functions_0x64.html
trunk/opentrep/ternary_tree/html/functions_0x65.html
trunk/opentrep/ternary_tree/html/functions_0x66.html
trunk/opentrep/ternary_tree/html/functions_0x67.html
trunk/opentrep/ternary_tree/html/functions_0x68.html
trunk/opentrep/ternary_tree/html/functions_0x69.html
trunk/opentrep/ternary_tree/html/functions_0x6b.html
trunk/opentrep/ternary_tree/html/functions_0x6c.html
trunk/opentrep/ternary_tree/html/functions_0x6d.html
trunk/opentrep/ternary_tree/html/functions_0x6e.html
trunk/opentrep/ternary_tree/html/functions_0x6f.html
trunk/opentrep/ternary_tree/html/functions_0x70.html
trunk/opentrep/ternary_tree/html/functions_0x72.html
trunk/opentrep/ternary_tree/html/functions_0x73.html
trunk/opentrep/ternary_tree/html/functions_0x74.html
trunk/opentrep/ternary_tree/html/functions_0x75.html
trunk/opentrep/ternary_tree/html/functions_0x76.html
trunk/opentrep/ternary_tree/html/functions_0x77.html
trunk/opentrep/ternary_tree/html/functions_0x7e.html
trunk/opentrep/ternary_tree/html/functions_enum.html
trunk/opentrep/ternary_tree/html/functions_eval.html
trunk/opentrep/ternary_tree/html/functions_func.html
trunk/opentrep/ternary_tree/html/functions_func_0x62.html
trunk/opentrep/ternary_tree/html/functions_func_0x63.html
trunk/opentrep/ternary_tree/html/functions_func_0x64.html
trunk/opentrep/ternary_tree/html/functions_func_0x65.html
trunk/opentrep/ternary_tree/html/functions_func_0x66.html
trunk/opentrep/ternary_tree/html/functions_func_0x67.html
trunk/opentrep/ternary_tree/html/functions_func_0x68.html
trunk/opentrep/ternary_tree/html/functions_func_0x69.html
trunk/opentrep/ternary_tree/html/functions_func_0x6b.html
trunk/opentrep/ternary_tree/html/functions_func_0x6c.html
trunk/opentrep/ternary_tree/html/functions_func_0x6d.html
trunk/opentrep/ternary_tree/html/functions_func_0x6e.html
trunk/opentrep/ternary_tree/html/functions_func_0x6f.html
trunk/opentrep/ternary_tree/html/functions_func_0x70.html
trunk/opentrep/ternary_tree/html/functions_func_0x72.html
trunk/opentrep/ternary_tree/html/functions_func_0x73.html
trunk/opentrep/ternary_tree/html/functions_func_0x74.html
trunk/opentrep/ternary_tree/html/functions_func_0x75.html
trunk/opentrep/ternary_tree/html/functions_func_0x76.html
trunk/opentrep/ternary_tree/html/functions_func_0x7e.html
trunk/opentrep/ternary_tree/html/functions_rela.html
trunk/opentrep/ternary_tree/html/functions_type.html
trunk/opentrep/ternary_tree/html/functions_type_0x62.html
trunk/opentrep/ternary_tree/html/functions_type_0x63.html
trunk/opentrep/ternary_tree/html/functions_type_0x64.html
trunk/opentrep/ternary_tree/html/functions_type_0x66.html
trunk/opentrep/ternary_tree/html/functions_type_0x68.html
trunk/opentrep/ternary_tree/html/functions_type_0x69.html
trunk/opentrep/ternary_tree/html/functions_type_0x6b.html
trunk/opentrep/ternary_tree/html/functions_type_0x6c.html
trunk/opentrep/ternary_tree/html/functions_type_0x6d.html
trunk/opentrep/ternary_tree/html/functions_type_0x6e.html
trunk/opentrep/ternary_tree/html/functions_type_0x70.html
trunk/opentrep/ternary_tree/html/functions_type_0x72.html
trunk/opentrep/ternary_tree/html/functions_type_0x73.html
trunk/opentrep/ternary_tree/html/functions_type_0x74.html
trunk/opentrep/ternary_tree/html/functions_type_0x76.html
trunk/opentrep/ternary_tree/html/functions_vars.html
trunk/opentrep/ternary_tree/html/globals.html
trunk/opentrep/ternary_tree/html/globals_defs.html
trunk/opentrep/ternary_tree/html/globals_func.html
trunk/opentrep/ternary_tree/html/graph_legend.dot
trunk/opentrep/ternary_tree/html/graph_legend.html
trunk/opentrep/ternary_tree/html/graph_legend.png
trunk/opentrep/ternary_tree/html/hierarchy.html
trunk/opentrep/ternary_tree/html/index.html
trunk/opentrep/ternary_tree/html/iteration__impl_8hpp.html
trunk/opentrep/ternary_tree/html/iterator__wrapper_8hpp.html
trunk/opentrep/ternary_tree/html/namespacecontainers.html
trunk/opentrep/ternary_tree/html/namespacecontainers_1_1smap__detail.html
trunk/opentrep/ternary_tree/html/namespacecontainers_1_1sset__detail.html
trunk/opentrep/ternary_tree/html/namespacecontainers_1_1tst__detail.html
trunk/opentrep/ternary_tree/html/namespacecontainers_1_1tst__detail_1_1mpl__detail.html
trunk/opentrep/ternary_tree/html/namespacecontainers_1_1tst__erase__impl__detail.html
trunk/opentrep/ternary_tree/html/namespacecontainers_1_1util.html
trunk/opentrep/ternary_tree/html/namespaceiterators.html
trunk/opentrep/ternary_tree/html/namespacemembers.html
trunk/opentrep/ternary_tree/html/namespacemembers_func.html
trunk/opentrep/ternary_tree/html/namespaces.html
trunk/opentrep/ternary_tree/html/namespacestd.html
trunk/opentrep/ternary_tree/html/new__iterator__base_8ipp.html
trunk/opentrep/ternary_tree/html/pages.html
trunk/opentrep/ternary_tree/html/perf_notes.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1smap__detail_1_1multimap__iterator-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1smap__detail_1_1multimap__iterator.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1sset__detail_1_1multiset__iterator-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1sset__detail_1_1multiset__iterator.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1ternary__tree_1_1find__result-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1ternary__tree_1_1find__result.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1always__heap__node-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1always__heap__node.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1back__push__pop-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1back__push__pop.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1dummy__sequence-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1dummy__sequence.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1heap__node-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1heap__node.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1inorder__seek-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1inorder__seek.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1inplace__node-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1inplace__node.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1iter__method__forward-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1iter__method__forward.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1key__access-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1key__access.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1levenshtein__search__info-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1levenshtein__search__info.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1levenshtein__search__info_1_1search-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1levenshtein__search__info_1_1search.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1mpl__detail_1_1if__c-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1mpl__detail_1_1if__c.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1mpl__detail_1_1if__c_3_01false_00_01_t1_00_01_t2_01_4-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1mpl__detail_1_1if__c_3_01false_00_01_t1_00_01_t2_01_4.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1node__base-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1node__base.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1size__policy__node-members.html
trunk/opentrep/ternary_tree/html/structcontainers_1_1tst__detail_1_1size__policy__node.html
trunk/opentrep/ternary_tree/html/structiterators_1_1const__traits-members.html
trunk/opentrep/ternary_tree/html/structiterators_1_1const__traits.html
trunk/opentrep/ternary_tree/html/structiterators_1_1iterator__wrapper-members.html
trunk/opentrep/ternary_tree/html/structiterators_1_1iterator__wrapper.html
trunk/opentrep/ternary_tree/html/structiterators_1_1nonconst__traits-members.html
trunk/opentrep/ternary_tree/html/structiterators_1_1nonconst__traits.html
trunk/opentrep/ternary_tree/html/structured__map_8hpp.html
trunk/opentrep/ternary_tree/html/structured__set_8hpp.html
trunk/opentrep/ternary_tree/html/structured_concept.html
trunk/opentrep/ternary_tree/html/tab_b.gif
trunk/opentrep/ternary_tree/html/tab_l.gif
trunk/opentrep/ternary_tree/html/tab_r.gif
trunk/opentrep/ternary_tree/html/tabs.css
trunk/opentrep/ternary_tree/html/ternary__tree_8hpp.html
trunk/opentrep/ternary_tree/html/todo.html
trunk/opentrep/ternary_tree/html/tst__implementation_8ipp.html
trunk/opentrep/ternary_tree/html/tst__iterator__base_8ipp.html
trunk/opentrep/ternary_tree/html/tst__iterator__facade_8hpp.html
trunk/opentrep/ternary_tree/html/tst__node_8hpp.html
trunk/opentrep/ternary_tree/html/tst__search__results_8ipp.html
trunk/opentrep/ternary_tree/html/tst_impl.html
trunk/opentrep/ternary_tree/html/tst_links.html
trunk/opentrep/ternary_tree/html/tst_reference.html
trunk/opentrep/ternary_tree/html/tst_tests.html
trunk/opentrep/ternary_tree/html/tst_usage.html
trunk/opentrep/ternary_tree/html/usage_8hpp.html
trunk/opentrep/ternary_tree/index.html
trunk/opentrep/ternary_tree/iterator_compile_test.cpp
trunk/opentrep/ternary_tree/iterator_wrapper.hpp
trunk/opentrep/ternary_tree/readme.txt
trunk/opentrep/ternary_tree/structured_map.hpp
trunk/opentrep/ternary_tree/structured_set.hpp
trunk/opentrep/ternary_tree/ternary_tree.hpp
trunk/opentrep/ternary_tree/test/
trunk/opentrep/ternary_tree/test/basic_insertion_test.hpp
trunk/opentrep/ternary_tree/test/check_iteration.hpp
trunk/opentrep/ternary_tree/test/copy_test.hpp
trunk/opentrep/ternary_tree/test/element_range_test.hpp
trunk/opentrep/ternary_tree/test/erase_test.cpp
trunk/opentrep/ternary_tree/test/hamming_search_test.cpp
trunk/opentrep/ternary_tree/test/iterator_test.cpp
trunk/opentrep/ternary_tree/test/localization_test.cpp
trunk/opentrep/ternary_tree/test/longest_match_test.cpp
trunk/opentrep/ternary_tree/test/mapped_value_test.cpp
trunk/opentrep/ternary_tree/test/partial_match_test.cpp
trunk/opentrep/ternary_tree/test/prefix_range_test.cpp
trunk/opentrep/ternary_tree/test/scrabble_search_test.cpp
trunk/opentrep/ternary_tree/test/test.vcproj
trunk/opentrep/ternary_tree/test/test_tst.cpp
trunk/opentrep/ternary_tree/test/tests_common.hpp
trunk/opentrep/ternary_tree/tst.doxy
trunk/opentrep/ternary_tree/tst_concept_checks.cpp
trunk/opentrep/ternary_tree/tst_detail/
trunk/opentrep/ternary_tree/tst_detail/iteration_impl.hpp
trunk/opentrep/ternary_tree/tst_detail/new_iterator_base.ipp
trunk/opentrep/ternary_tree/tst_detail/tst_implementation.ipp
trunk/opentrep/ternary_tree/tst_detail/tst_iterator_base.ipp
trunk/opentrep/ternary_tree/tst_detail/tst_iterator_facade.hpp
trunk/opentrep/ternary_tree/tst_detail/tst_node.hpp
trunk/opentrep/ternary_tree/tst_detail/tst_search_results.ipp
trunk/opentrep/ternary_tree/tst_public.doxy
trunk/opentrep/test/ternary/
Added: trunk/opentrep/ternary_tree/README
===================================================================
--- trunk/opentrep/ternary_tree/README (rev 0)
+++ trunk/opentrep/ternary_tree/README 2009-07-14 10:45:19 UTC (rev 126)
@@ -0,0 +1,3 @@
+
+Source: http://abc.se/~re/code/tst and http://abc.se/~re/code/tst/ternary_tree.zip
+
Added: trunk/opentrep/ternary_tree/doxygen_input/blather.hpp
===================================================================
--- trunk/opentrep/ternary_tree/doxygen_input/blather.hpp (rev 0)
+++ trunk/opentrep/ternary_tree/doxygen_input/blather.hpp 2009-07-14 10:45:19 UTC (rev 126)
@@ -0,0 +1,558 @@
+/** \mainpage Structured Associative Containers
+
+Ternary Search Tree containers to replace \c set<string> and \c map<string, Value> </h2>
+
+<center><table bgcolor="#fbf9e5" style="border: thin dotted #808000;" width="95%" border=0>
+<tr>
+<td>
+<h3>Table of contents</h3>
+<dl>
+ <dt>\ref introduction "Introduction"</dt>
+ <dt>\ref subkey_search_overview "Advanced searches overview"</dt>
+ <dt>\ref tst_usage "Tutorial"</dt>
+ <dt>\ref tst_reference "Reference"</dt> <dd>
+ <dd>\ref structured_concept "Structured Container concept" \n
+ Class \ref containers::structured_set "structured_set" \n
+ Class \ref containers::structured_map "structured_map" \n
+ Class \ref containers::structured_multiset "structured_multiset" \n
+ Class \ref containers::structured_multimap "structured_multimap" \n
+ Implementation class \ref containers::ternary_tree "ternary_tree"
+ </dd></dt>
+ <dt>\ref perf_notes "Performance notes"</dt>
+ <dt>\ref tst_impl "Implementation details"</dt>
+ <dt>\ref tst_links "Links"</dt>
+ <dt>\ref tst_tests "Test Suite"</dt>
+</dl>
+</td>
+</tr></table></center>
+
+Download: Latest version (0.684) http://abc.se/~re/code/tst/ternary_tree.zip\n
+
+Copyleft: <a href="mailto:rasmus%20point%20ekman%20at%20abc%20point%20se?subject=Structured Containers suck/rule">
+rasmus ekman</a> 2007-2009 \n
+Weblink: http://abc.se/~re/code/tst
+
+\anchor introduction <hr>
+<h2>Introduction</h2>
+<b>Structured containers</b> are \c map and \c set -like containers specialized for strings.
+They are commonly used for dictionaries.\n
+Structured containers have two major benefits:
+- They offer near-match searches (wildcard search, partial match etc) that are hard to implement
+ with other containers.
+- Lookup performance is on a par with hashed containers for many common applications,
+and 2-5 times faster than standard maps and sets (with string-like keys).
+
+Of course there is a price to pay: structured containers use much more memory than
+other containers: Around 6-8 bytes <b>per letter</b> inserted (whether \c char or \c wchar_t);
+an English 150 k word dictionary uses eg 7.3 MB to store 1.2 MB words (2.4 MB of \c wchar_t words).
+
+The container classes in this library can be used as drop-in replacements for \c set and \c map
+(or \c unordered_set, \c unordered_map):
+ - \ref containers::structured_set "structured_set": This stores unique keys and allows structured key searches.
+ - \ref containers::structured_multiset "structured_multiset": This stores non-unique keys.
+ - \ref containers::structured_map "structured_map": This is a
+ <a target="sgi" href="http://www.sgi.com/tech/stl/PairAssociativeContainer.html">Pair Associative Container</a>,
+ as it allows associating a value with each key.
+ - \ref containers::structured_multimap "structured_multimap": Technically, a
+ <a target="sgi" href="http://www.sgi.com/tech/stl/MultipleSortedAssociativeContainer.html">Multiple, Sorted,
+ Pair Associative Container</a> - it allows storing several values with each key.
+
+While the STL standard associative containers are normally backed by a binary tree structure,
+Structured Containers are backed by a Ternary Search Tree, as presented by
+\ref note_1 "Jon Bentley and Robert Sedgewick in [1]".
+
+Class \ref containers::ternary_tree "ternary_tree<Key, Value, Comp, Alloc>" provides the implementation backend.
+Due to its internals, its interface cannot easily be made to conform with standard STL concepts,
+so it is used internally by the structured* wrapper classes (much like STL's internal \c rb_tree class).
+
+Basically, if you have code using sets or maps, you have code to use structured containers.
+And with 1-3 lines of code, you're ready to make advanced imprecise searches in your dictionaries.\n
+See \ref tst_usage "the usage section" for examples of how to use these classes.
+
+<table bgcolor="#f0f0ff" style="border: thin dotted #808000;" border=0>
+<tr><th>Library status</th></tr>
+<tr><td valign="top" align="right">Compatibility:</td>
+<td>Note that the file \b tst_concept_checks.cpp is currently broken. Will investigate.\n
+<!-- This used to compile with Mingw GCC 3.4.2 and with MSVC7.1 (with STLport 5). Requires Boost 1.33.
+Not sure what happened in Boost 1.36-37 or if I've mangled something. \n
+Due to recent changes, ternary tree does not support stateful allocators (earlier versions did this by implication) -->
+</td>
+<tr><td valign="top" align="right">version 0.684: (Jan 2009)</td>
+<td>Fix standard-breakage in multimap/multiset return from <code>insert(const value_type&)</code>.<br>
+Added <code>operator-></code> to iterator wrapper for C++0x compatibility.
+Thanks to Geoffrey Noel for reports.</td>
+</tr>
+<tr><td valign="top" align="right">version 0.683: (March 2007)</td>
+<td>Fix portability issues for GCC and non-STLport libraries. Fix longest_match.<br>
+Thanks to Arjen Wagenaar for several reports, fixes and encouragement. Thanks also to Michel Tourn for reports.</td>
+</tr>
+<tr><td valign="top" align="right">version 0.68: (Dec 2006)</td>
+<td>Implement TST_NODE_COUNT_TYPE macro, which can be used to control node size on 64-bit systems.
+ See \ref containers::ternary_tree "class ternary_tree"</td>
+</tr>
+<tr><td valign="top" align="right">version 0.68 (alpha):</td>
+<td>Reimplemented node type. Do proper management of value type (was inconsistent, partly unimplemented - duh!)</td>
+</tr>
+<!--tr>
+<tr><td valign="top" align="right">version 0.676:</td>
+<td>Modified containers to follow C++0x draft standard: \n
+Added \c cbegin, \c cend methods returning \c const_iterator, and \c crbegin, \c crend
+returning \c const_reverse_iterator, to make it easier to code with const-correctness. \n
+\c erase(iterator pos); and \c erase(iterator first, iterator last); methods now return iterators.</td>
+<tr><td valign="top" align="right">version 0.675:</td>
+<td>All Structured Container classes implemented. Structured search interface TBD.
+</td-->
+</table>
+
+
+\anchor subkey_search_overview <hr>
+<h2>Sub-key, or Structure Searches</h2>
+<span style="color:#905050;">(a new interface for these searches will be specified in the future)</span>
+
+Ternary trees allow searches that match parts of keys and ignores mismatches in other parts.\n
+In the current interface we specify a small number of searches facilitated by the tree structure;
+the Partial Match and Hamming searches are defined in several other implementations
+(showcased in \ref note_1 "Bentley and Sedgewick" code).
+The Levenshtein and combinatorial searches are not found in other ternary trees (that I know of).
+
+<table border="1" cellspacing="0">
+ <tr><th bgcolor="#f0f0ff">Name (function name)</th><th bgcolor="#f0f0ff">Description</th></tr>
+ <tr><th>
+ Prefix match (\ref containers::ternary_tree::prefix_range "prefix_range")</th><td>
+ Finds keys sharing a common prefix, returns a pair of iterators.</td></tr>
+ <tr><th>
+ Longest match (\ref containers::ternary_tree::longest_match "longest_match")</th><td>
+ Finds the longest key that matches beginning of search string.
+ A typical application is to tokenize a string using the ternary tree as dictionary.</td></tr>
+ <tr><th>
+ Partial match, or wildcard search (\ref containers::ternary_tree::partial_match_search "partial_match_search")</th><td>
+ Accepts a search string with wildcard characters that will match any letter,
+ eg "b?nd" would match "band", "bend", "bind", "bond" in an English dictionary.</td></tr>
+ <tr><th>
+ Search allowing \c N mismatches,
+ (\ref containers::ternary_tree::hamming_search "hamming_search"<span style="font-weight:normal;"></span>)</th><td>
+ Accepts a search string and an integer \c dist indicating how many non-matching letters are allowed,
+ then finds keys matching search string that have at most \a dist mismatches.
+ This works like a partial match search with all combinations of \a dist
+ wildcards in the search string.\n
+ \c hamming_search("band", 1) matches the wildcard search plus "bald", "bane" and "wand", etc. \n
+ The version here, following DDJ code, extends the strict Hamming search by also allowing shorter and longer
+ strings; a search for "band", \a dist = 1, also finds "ban" and "bandy" etc.\n
+ See also http://wikipedia.org/wiki/Hamming_distance</td></tr>
+ <tr><th>
+ Levenshtein distance search</b> (\ref containers::ternary_tree::levenshtein_search "levenshtein_search"
+ <span style="font-weight:normal;">- consider descriptive name</span>)</th><td>
+
+ Hamming search matches characters in fixed position, allowing substitution of \a dist chars.
+ Levenshtein search also allows shifting parts of the search string by insertion or skipping chars (in \a dist places).
+ So <code>levenshtein_search("band", 1) </code> extends the hamming_search set with "and" and "bland", etc.
+ A typical application is to match mispelt words.\n
+ See also http://wikipedia.org/wiki/Levenshtein_distance</td></tr>
+ <tr><th>
+ Combinatorial or "scrabble" search (\ref containers::ternary_tree::combinatorial_search "combinatorial_search")</th><td>
+ Finds all keys using the characters in search string. \c combinatorial_search("band") finds
+ "ad", "and", "bad", "dab", "nab", etc. A count of wildcards can be added, also allowing
+ nonmatching characters (use with care, values over 10% of average key length
+ may cause the algorithm to traverse a large part of the tree).</td></tr>
+</table>
+
+See \ref usage_imprecise_searches "advanced search overview" in the tutorial.
+
+These searches are defined for all containers in this library.
+But they are also marked as deprecated (to be replaced by generic algorithms with same interface).
+For a relative performance comparison of imprecise searches, see the second table in \ref perf_notes.
+
+<h3>Future directions</h3>
+The searches currently defined are clearly special cases in a sea of search possibilities.
+We have only defined searches that are relatively efficient, compared to other combinations of containers and algorithms.
+But there can be many variations on the available searches: increasing Hamming/Levenshtein distance
+at the end of words, or matching limited ranges of characters (eg allowing mismatches only in vowels), etc.
+
+The next step for this project is to support a more flexible low-level interface for
+traversing and filtering tree nodes.
+The interface for these "structured searches" is open for consideration, but it
+will basically define sub-key iterators, conversion of full-key from sub-key iterators,
+and a small collection of algorithms operating on these sub-key iterators.
+
+At least the following operations are needed:
+
+ - sub-key match: matching a part of a key (prefix, or starting from current char position)
+ - key element range increment: from a sub-key position, match a range of characters
+ in next position (returns a list of sub-key iterators? - or iterator-like operation?)
+ - conversion from sub-key iterator to full-key iterator range (nearest and post-furthest
+ keys in the subtree)
+ - \c is_key(subkey_iterator pos): true if end-of-key exists at iterator position.
+ - \c count_elements(subkey_iterator pos): returns number of available key elements at position.
+ - In all predefined algorithms above, either a specific, or any char is matched,
+ we would also support arbitrary char sets (possibly with special case for char ranges).
+
+ */
+
+/** \page tst_reference Reference
+<center><table bgcolor="#fbf9e5" style="border: thin dotted #808000;" width="95%" border=0>
+<tr>
+<td>
+<dl>
+ <dt>\ref structured_concept "Structured Container concept"</dt>
+ <dt>\ref ref_sethpp "Header < structured_set.hpp >"</dt>
+ <dt>\ref ref_maphpp "Header < structured_map.hpp >"</dt>
+ <dt>\ref ref_tsthpp "Header < ternary_tree.hpp >"</dt>
+ <dt>\ref ref_iterhpp "Header < iterator_wrapper.hpp >"</dt>
+</dl>
+</td>
+</tr></table></center>
+
+<hr>
+
+\anchor ref_sethpp
+<h2>Header < <a href="../structured_set.hpp">%structured_set.hpp</a> > synopsis</h2>
+<pre>
+\b namespace containers {
+ \b template <\b class Key,
+ \b class Comp = std::less<\b typename Key::value_type>,
+ \b class Alloc = std::allocator<Key> >
+ \b class \ref containers::structured_set "structured_set";
+
+ \b template <\b class Key,
+ \b class Comp = std::less<\b typename Key::value_type>,
+ \b class Alloc = std::allocator<Key> >
+ \b class \ref containers::structured_multiset "structured_multiset";
+}
+</pre>
+
+\anchor ref_maphpp
+<h2>Header < <a href="../structured_map.hpp">%structured_map.hpp</a> > synopsis</h2>
+<pre>
+\b namespace containers {
+ \b template <\b class Key,
+ \b class T,
+ \b class Comp = std::less<\b typename Key::value_type>,
+ \b class Alloc = std::allocator<std::pair<\b const Key, T> > >
+ \b class \ref containers::structured_map "structured_map";
+
+ \b template <\b class Key,
+ \b class T,
+ \b class Comp = std::less<\b typename Key::value_type>,
+ \b class Alloc = std::allocator<std::pair<\b const Key, T> > >
+ \b class \ref containers::structured_multimap "structured_multimap";
+}
+</pre>
+
+<hr>
+Supplementary header files needed to support structured_set and -map classes.
+
+
+\anchor ref_tsthpp
+<h2>Header < <a href="../ternary_tree.hpp">%ternary_tree.hpp</a> > synopsis</h2>
+<pre>
+\b namespace containers {
+
+ \b template <\b class Key,
+ \b class T,
+ \b class Comp = std::less<\b typename Key::value_type>,
+ \b class Alloc = std::allocator<std::pair<\b const Key, T> > >
+ \b class \ref containers::ternary_tree "ternary_tree";
+
+ \b template <\b class TreeT, \b class IteratorT>
+ \b class \ref containers::search_results_list "search_results_list";
+
+}
+</pre>
+
+
+\anchor ref_iterhpp
+<h2>Header < <a href="../iterator_wrapper.hpp">%iterator_wrapper.hpp</a> > synopsis</h2>
+<pre>
+\b namespace iterators {
+
+ \b template <\b class T> \b struct const_traits;
+ \b template <\b class T> \b struct nonconst_traits;
+
+ \b template <\b class BaseIterT,
+ \b class TraitsT, // either const_traits<T> or nonconst_traits<T>
+ \b class IterCatT = std::bidirectional_iterator_tag >
+ \b class \ref iterators::iterator_wrapper "iterator_wrapper";
+}
+</pre>
+
+*/
+
+/**
+\page structured_concept Structured Associative Container Concept
+
+<span style="color:#905050;">(a preliminary sketch of the formal technical concept description)</span>
+
+A Structured Associative Container is a specialization of the C++ 98 standard concept
+<a target="sgi" href="http://www.sgi.com/tech/stl/SortedAssociativeContainer.html">Sorted Associative Container</a>,
+with extended interface.
+
+The template parameters are similar to that of the Associated Containers:
+
+<code> structured_set<Key, Comp, Alloc>; </code>\n
+<code> structured_map<Key, Value, Comp, Alloc>; </code>\n
+
+where:
+ - \c <b>Key</b> type is itself a container (eg a \c std::string or \c std::wstring)
+ - \c <b>Comp</b> is a comparison operator that imposes a sort order on \c Key::value_type elements \n
+ (so if \c Key is string, \c Comp compares \c char, if \c Key is \c wstring, \c Comp applies to \c wchar_t).
+ - \c <b>Value</b> can be any Assignable type
+ - \c <b>Alloc</b> is an allocator that manages all memory allocation for the container.
+
+The \c Comp and the \c Alloc types have default template arguments.
+
+In other words Structured containers are like Sorted Associative Containers, BUT
+ - add the requirement on Key template type to be a
+ <a target="sgi" href="http://www.sgi.com/tech/stl/ForwardContainer.html">Forward Container</a>.\n
+ For example, \c std::basic_string<CharT> is compatible with this requirement.
+ - change the requirement on the \c Comp (comparator) template argument to operate on
+ \c key_type::value_type elements (rather than on \c key_type itself).
+ Like Sorted Associative comparator, the \c Comp type shall define a less-like comparison, a
+ <a target="sgi" href="http://www.sgi.com/tech/stl/StrictWeakOrdering.html">Strict Weak Ordering</a>
+ of key-elements.
+
+<b>Associated types</b>
+ - \b char_compare: less-like comparison of key elements (establishing a Strict Weak Ordering).
+ The <a target="sgi" href="http://www.sgi.com/tech/stl/AssociativeContainer.html">Associative Container</a>
+ \c key_compare type is also provided, but is defined in terms of \c char_compare. \n
+ - \b subkey_iterator: Used in structure searches. Convertible to iterator (TBD).
+
+In consequence it allows searches involving subparts of keys, ie with shared prefix and/or
+with shared middle parts.
+
+<hr>
+<h3>Deprecated search interface</h3>
+
+In the first iteration, additional searches are provided as methods on the containers.
+This will be changed to use free functions operating on \c subkey_iterator.
+The deprecated search methods will still be provided as convenience functions;
+to migrate your code from present version to the new interface, will mean moving
+the object name to the first argument, but also to respecify the search_results_list type.
+(This sloppy-hackish type is by itself reason not to keep the method interface)
+
+See \ref subkey_search_overview "Structured search overview"
+and \ref tst_structsearch "ternary_tree Structure search section".
+*/
+
+/*
+
+\b Notation \n
+<table border=0>
+<tr><td>\c X <td>A type that is a model of Associative Container </td></tr>
+<tr><td>\c a <td>Object of type \c X </tr>
+<tr><td>\c k <td>Object of type \c X::key_type </tr>
+<tr><td>\c p, \c q <td>Object of type \c X::char_iterator </tr>
+<tr><td>\c c <td>Object of type \c X::char_type </tr>
+<tr><td>\c o <td>Object modelling output iterator </tr>
+<tr><td>\c i <td>Object of type \c X::size_type </tr>
+</dl>
+
+<table border=1>
+<tr><th>Name</th><th>Expression</th><th>Return value</th>
+<tr><td>Prefix match</td><td><code>a.prefix_range(k)</code></td><td>
+ \c std::pair<iterator, iterator> if \c a is mutable, otherwise <br>\c std::pair<const_iterator, const_iterator></td></tr>
+<tr><td>Longest match</td><td><code>a.longest_match(p, q)</code></td><td>
+ \c iterator if \c a is mutable, otherwise \c const_iterator</td></tr>
+<tr><td>Partial match, or <br>wildcard search</td><td><code>a.partial_match_search(k, o, c)</code></td><td>
+ The output iterator \c o</td></tr>
+<tr><td>Hamming search</td><td><code>a.hamming_search(k, o, i)</code></td><td>
+ The output iterator \c o</td></tr>
+<tr><td>Levenshtein search</td><td><code>a.levenshtein_search(k, o, i)</code></td><td>
+ The output iterator \c o</td></tr>
+<tr><td>Combinatorial or <br>"scrabble" search</td><td><code>a.combinatorial_search(k, o, i)</code></td><td>
+ The output iterator \c o</td></tr>
+</table>
+
+*/
+
+/** \page tst_impl Implementation Details
+ * (In the following, "original" and "DDJ" code refers to the article by Bentley/Sedgewick
+ * published in Dr Dobb's Journal, and the accompanying C source code - see \ref tst_links)
+ *
+ * In most implementations, a ternary tree node has the following members: \code
+ * struct node {
+ * char splitchar; // key letter, or 0 at end of key
+ * node *hikid; // subtree of keys with higher value than splitchar
+ * node *eqkid; // subtree matching splitchar (pointer to mapped value at end-of-key node)
+ * node *lokid; // subtree less than splitchar
+ * node *parent; // necessary for iteration (not needed for insert/find)
+ * }; \endcode
+ *
+ * This means that each node is 1 char plus three or four pointers size.
+ * On many systems, struct member alignment makes the char member consume size of one pointer
+ * as well, so we have 4 (or 5) x sizeof(pointer) per node in the tree.
+ * With several kinds of dictionaries, the node count ends up at around 0.3-0.5 times
+ * total key length (since keys share nodes).
+ * This is even more expensive on 64-bit machines.
+ *
+ * There are several variation points in the node class:
+ * -# the DDJ C code designates an invalid value of zero to indicate end-of-string. We want to
+ * allow any string as key, so the end-of-string representation should change.
+ * We note that on many platforms, C/C++ struct member alignment leaves a "hole"
+ * in the binary representation of the node, between the char and the first pointer ("hikid").
+ * On such systems there is no space cost to use another char-sized value to indicate end node.
+ * This also works for \c wchar_t strings on 32- or 64-bit systems.
+ * -# The original code stores a value for each string in the terminal node's "equal" pointer.
+ * The value in DDJ code is always a pointer to the terminated string. This is used to make
+ * advanced searches work (they return an array of pointers to strings stored in end-nodes).
+ * In reality this means that strings may need to be copied on insertion (not reflected in DDJ timings).
+ * -# Original DDJ code does not support iterating over strings in the tree.
+ * Idiomatic STL-like container style strongly suggests that iteration should be supported.
+ * This is fairly simple to implement if a parent pointer is added to the node struct:
+ * Because when an end-node is reached, the iterator must backtrack to find the previous
+ * branch point.
+ *
+ * The parent pointer also makes it possible to recover the inserted string by walking nodes
+ * backward from a terminal node to the root. Complexity is key length, plus log(tree.size),
+ * but it means inserted keys do not \b have to be copied to the end node.
+ * We opt to cache keys in iterators, at no measurable extra cost in iteration.
+
+ * Instead of the key, an arbitrary value can be associated with endnodes.
+ * However, it should not be allowed to increase node size, since most nodes in the tree are not endnodes.
+ * In this library we store the mapped value directly in end-node if it is <tt> <= \c sizeof(void*). </tt>
+ * Larger objects are allocated on the heap, and a pointer to the copy is stored in end-node
+ * (the copy is managed by the tree).
+ *
+ * <h4>Now for some optimization</h4>
+ * We use a \c vector<node> as pool allocator, and record eq-hi-lo links as vector index instead of pointers.
+ * The pool allocation essentially follows original C code insert2() principle.
+ * For us, it also simplifies reallocation, since pointers do not have to be rebound;
+ * the indices are always valid.
+ * This has the following consequences:
+ * - allow the option of 4-byte indices also on 64-bit systems (with obvious resulting tree size limit)
+ * - When a new key is inserted, the last part (unique to the key) is always allocated in a batch.
+ * This means that one node member, \c "eqkid", becomes redundant, as it is always the next index
+ * (except after terminal nodes of course).
+ * - in DDJ code the end-node value is stored in union with the eqkid. We note that the \c lokid node index
+ * is also unused by end-nodes (as no char should be lower than zero), so all endnode children
+ * are linked to the hi node.
+ *
+ * (In our binary-cognizant version where zero is a regular char value, this still holds,
+ * we just change the end-node test accordingly)
+ *
+ * In the final cut, our node struct data members appear roughly like this: \code
+ * struct node {
+ * CharType splitchar; // key letter, or 0 at end of key (to make sure lokid is never allowed)
+ * CharType endflag; // zero on normal nodes, 1 at end nodes, 2 at erased nodes.
+ * node_index hikid; // subtree of keys with higher value than splitchar
+ * node_index lokid; // subtree less than splitchar
+ * node_index parent; // necessary for iteration (not needed for insert/find)
+ * }; \endcode
+ *
+ * where \c CharType is defined by template \c Key::value_type, and treated as an unsigned type
+ * (so 0 is the lowest value); and \c node_index is a \c size_t -like type used by the node
+ * storage backend (currently \c std::vector).
+ *
+ * This optimization could also be applied to C version, trimming space requirement in DDJ code
+ * to 3-word nodes.
+ */
+
+/** \page perf_notes Performance Notes
+ *
+ * <h3>Space considerations</h3>
+ *
+ * Ternary trees are notably larger than hash maps or most binary tree designs.
+ * Each node holds only one character (instead of a whole key), and use 3-5 pointers.
+ * Our nodes consist of 4 \c size_t values (16 bytes) regardless of platform pointer size,
+ * or char type (if at most 2-byte like \c wchar_t).
+ *
+ * The shared parts of strings save space: In a typical English dictionary,
+ * each key shares over half its nodes with other keys, so the allocated space is about half
+ * of total key-length times 16. In a scrabble dictionary like the one reported below,
+ * which contains all valid word endings, most nodes are shared, so its storage cost is "only"
+ * total key length times 0.35 times 16, or less than 6 bytes per char.
+ * With \c wchar_t type, the storage cost cannot be considered overly large.
+ *
+ * See also \ref tst_impl
+ *
+ * <h3>Lookup speed</h3>
+ *
+ * The complexity of ternary tree operations is basically the same as for binary trees,
+ * (logarithmic in tree size) but with quite different constant factors. See \ref note_1 "[1]".
+ *
+ * Overall lookup and iteration speed depends on application factors - ie
+ * whether strings are inserted in random order or not, etc.
+ *
+ * Rough speed estimates (compared to Stlport hash_map and map).
+ * - insertion is a bit slower (>30% to 0%) than hash_map, ~30% faster than map.
+ * - finding a key is ~0-50% slower than hash_map (equal on failure, with short keys).
+ * - finding a key is 1.5-3 times faster than map (again with short keys).
+ *
+ * Compared to C versions (DDJ and libtst),
+ * - find and insert are slower, by factors ranging from 1.5-4.
+ * - partial_match and neighbour searches are 5-20% faster than published DDJ code -
+ * the code is essentially the same, but our implementation rolls out some recursion.
+ * This is easily back-ported, so in effect they should be considered to run at same speed.
+ * This by itself is good news though, since eg single-key lookup is always slower.
+ *
+ * Since each character in a key is at a separate node in the internal tree,
+ * iterating over values is a little slower than for other tree-based containers.
+ *
+ * For detailed test, see performance table below.
+ *
+ * <hr style="height: 3px; border-top: 0px; background-color: #e09060;">
+ * \htmlinclude performancetable.html
+ */
+
+
+/** \page tst_links Links
+ * ternary_tree by rasmus ekman, see http://abc.se/~re/code/tst <br>
+ * Download: http://abc.se/~re/code/tst/ternary_tree.zip
+ *
+ * Some other TST implementations.
+ * - <b>DDJ code:</b> Original C implementation by Jon Bentley and Robert Sedgewick.
+ * Article in Dr Dobb's Journal, 1998 #4: http://www.ddj.com/documents/s=921/ddj9804a/9804a.htm \anchor note_1 \n
+ * See http://www.cs.princeton.edu/~rs/strings/ for C code and article on TST complexity.
+ * - \b libtst: Worked-out version of DDJ code by Peter A. Friend 2002. Version 1.3. \n
+ * See http://www.octavian.org/cs/software.html \anchor note_2 \n
+ * - \b Boost.Spirit version: C++ reimplementation by Joel de Guzman. \anchor note_3 \n
+ * See http://spirit.sourceforge.net/ internal file ./boost/spirit/symbols/impl/tst.h
+ * - <b>Hartmut Kaiser version:</b> C++ reimplementation intended for generalization of tst.
+ * Currently abandoned, available in Spirit CVS. (interesting for interface design) \n
+ * See http://lists.boost.org/Archives/boost/2005/09/93316.php \n
+ * and http://article.gmane.org/gmane.comp.parsers.spirit.general/6959
+ * - \b pytst: C++ version by Nicolas Lehuen, with SWIG wrappers for use from other languages. Version 0.97. \n
+ * See http://nicolas.lehuen.com/download/pytst/
+ * (not yet tested) \anchor note_4
+ *
+ * <h2>Feature chart</h2>
+ * All versions have insert and plain search, other features available as tabled below:
+ * \htmlinclude featuretable.html
+ */
+
+/** \page tst_tests Test Suite
+
+All tests require the <a href="http://boost.org">Boost library</a> to compile.
+
+<h3>Concept checks</h3>
+
+The file <a href="../tst_concept_checks.cpp">tst_concept_checks.cpp</a>
+performs a compile-time test of structured containers. \n
+A class \c StructuredAssociativeContainer is defined, which contains
+prototypes of all required methods for structured containers (also class ternary_tree).
+Relevant concepts from \c boost/concept_check.hpp are used to check the structured set/map
+containers.
+
+<h3>Correctness tests</h3>
+
+The subdirectory \b test in the distribution contains a bunch of files hacked up during development.
+All these tests are performed by a single main test file <a href="../test/test_tst.cpp">test_tst.cpp</a>.
+This file includes individual .cpp files, since we use a simplified (hacked) version
+of the Boost.Test harness.
+
+Each test prints a single line to \c std::cerr saying whether the test was "OK" or "FAIL".
+A line is added if an exception was thrown.
+
+These are runtime tests, several which require a file name to a dictionary-type file,
+a plain-text file with one word per line.
+The file \c fill_dictionary.cpp must be compiled with test projects,
+it reads a dictionary file and fills a std::vector with strings.
+
+Dictionary files can be found by an internet search (try eg "dictionary file").
+
+<h3>To do</h3>
+
+Proper organization and cleanup of this part of our library will be required before 1.0 release.
+
+*/
+
+
Added: trunk/opentrep/ternary_tree/doxygen_input/concepts.txt
===================================================================
--- trunk/opentrep/ternary_tree/doxygen_input/concepts.txt (rev 0)
+++ trunk/opentrep/ternary_tree/doxygen_input/concepts.txt 2009-07-14 10:45:19 UTC (rev 126)
@@ -0,0 +1,122 @@
+Tutorial
+
+In programming, a Concept is a set of formal requirements on input/output of a subsystem, or pre/post-conditions of
+an operation, of complexity constraints and exceptional behaviour.
+Note ye well the "complexity" bit. Since the specification of C++ STL, the complexity of operations on
+a type have been introduced as a proper feature of its concept, a full-citizenship part of type specification.
+(This diverges from the mathematical roots of programming, which defines types/concepts in purely structural terms
+-- ie as long as an operation does not transgress countability or infinity boundaries, it doesn't matter a damn bit whether
+it requires zero overhead or would enrol half the atoms in the universe to encode intermediate information.
+Maths is not about bean counting.).
+
+Here we will discuss tree concepts in the common sense.
+The following are some definitions of terms as used in documentation of the Structured Containers library.
+The definitions given are stipulative, in that they do not purely document an existing usage, but unless an expert
+tells me otherwise, I believe they should be made into when discussing trees and tries.
+
+
+Tree =df a directed acyclic graph of single-parented, multi-childed Nodes. Usually single-rooted, but this is not essential.
+ Stipulative: tree nodes have a fixed maximum number of children.
+ An implementation constraint that has become ingrained in most programmers' understanding of the concept.
+ All trees can be reduced to (easily and naturally implemented by) a binary tree.
+
+We assume common terminology for the parts of Trees:
+ Root - a node designated as start point, from which other nodes are reachable as children, or children of children etc.
+ Single-rooted Tree - a Tree where all nodes are reachable from a single Root node.
+ Multi-rooted tree, or Forest - a Tree where several start nodes are designated.
+ Level N - the set of nodes that are at the same distance from Root. Every node at level 3 is reachable
+ from the/a Root node by following exactly 3 child-node links.
+ Sibling - relation between any two nodes on the same Level.
+ Leaf - a node without children.
+ Fanout - the number of children that a node can have. This defines the maximum number of nodes at each level.
+
+Trie =df A Tree where the nodes have a "alphabet"-sized (max) number of children, for some alphabet.
+ Typical alphabets are the English letters, Unicode, or the ACGT genetic bases.
+
+Tree nodes represent a full "key" of any [less-comparable] type.
+Tries store string-like keys; a Trie node does not store a full key, only a part of it.
+A full key is represented by a leaf node and its path back to the/a root node.
+The reason for using Tries is that access to string-like keys is very fast - in principle linear in key length.
+Binary Trees over the same key is O(log n) where n is count of keys in the tree, with average key length
+as a constant factor.
+
+
+Ternary Search Tree (TST) is a space optimization for Tries.
+
+Because each path to a child of a Tree node takes up memory space, Trie nodes are very expensive
+if the alphabet is large. From the 3rd or 4th level on, most child-node links are empty.
+A TST constructs exactly the number of links existing at each level of a Trie.
+
+Graphics:
+Tree
+ root==node==node==node
+ \\node==node \\node==node
+
+Trie (6-letter alphabet: 123456)
+ root
+ ______________||______________
+ || || || || || ||
+ node1 (empty) node3 node4 node5 (empty)
+ ______________||_____________
+ || || || || || ||
+ node1 node2 node3 node4 node5 node6
+
+Here we see the root node with 4/6 child links populated. Each child has 6 empty links, except the 5th child.
+The 5th child has 6/6 links filled, and each of its children has 0/6 children.
+In all there are 4+6 = 10 nodes, and 10*6 = 60 links.
+Since a Trie only stores part of a key, the substantial information in each trie node is small, and the
+structure overhead - the links - is very large.
+
+This cries out for optimization. Several kinds of variable-sized nodes have been tried, but they usually
+end up with complex code to use and maintain, and thus squander the search speed which was the rationale
+for constructing Tries in the first place.
+
+
+TST is one such optimization. Here each trie node is constructed from much smaller nodes, but the
+code to use and maintain nodes is still fairly simple, so search speed is not badly compromised.
+Let's see the structure of the above tree:
+The exact runtime layout depends on insert order. If child 3 is inserted before child 1, child1 may become a
+"lower-child" child of child 2.
+Here we assume insertion order 4, 3, 5, 1 for the root
+
+ //node1
+ //node3
+root=node4
+ \\node5==(level2)
+
+level 2: assume insertion order 3,4,5,1,2,6
+
+ //node1
+ // \\node2
+level2.root=node3
+ \\node4
+ \\node5
+ \\node6
+
+
+Here we see 10 nodes, each with 3 child links. This means 30 links, ie half the link count of Trie.
+The space savings are of course even better for larger alphabets.
+(And in the Structured Containers implementation, the middle child link is omitted, so only 2*10=20 links are needed.)
+Given English alphabet Trie with the above sparse population, there would be 26*10 = 260 links in
+the Trie (each Trie node has 26 child links), and still the same count of TST node-links (30, or 20).
+A Unicode Trie would have 10*2^16 links= 10*65536= 655 thousand links. A TST for Unicode again uses 30 (20) links only.
+
+Important points wrt TSTs and TST nodes.
+A. TST nodes have two different kinds of child links:
+ (1) Two same-level sibling links
+ (2) One next-level "proper" child link.
+
+B. TSTs generalizes Trie implementations.
+ Tries with any alphabet can be implemented by the same TST node type - no new type needs to be defined for new alphabets.
+ (However a specialization is still often needed, since there must be a comparison function for the alphabet letters)
+
+C. TSTs are a hybrid of (binary) Tree and Trie.
+ In consequence of (A), TSTs combine the features of binary trees with Tries.
+ TST nodes can be viewed as binary nodes with associated data, where the data is a link to a next-level binary tree
+ (that implements a trie node).
+ - Against this view one may note that the binary treelets have absolute size constraints (defined by
+ count of letters in the implemented alphabet) - this is an "unnatural" constraint on a tree type.
+ - In support of the view one may note that search complexity is more like binary trees than pure Trie implementation.
+
+
+
Added: trunk/opentrep/ternary_tree/doxygen_input/doxygen-old.css
===================================================================
--- trunk/opentrep/ternary_tree/doxygen_input/doxygen-old.css (rev 0)
+++ trunk/opentrep/ternary_tree/doxygen_input/doxygen-old.css 2009-07-14 10:45:19 UTC (rev 126)
@@ -0,0 +1,311 @@
+BODY,H1,H2,H3,H4,H5,H6,P,CENTER,TD,TH,UL,DL,DIV {
+ font-family: Geneva, Arial, Helvetica, sans-serif;
+}
+/*BODY,TD { font-size: 90%; }
+H1 {
+ font-size: 150%;
+ background-color: #eeeeff;
+ width: 100%;
+ border: 1px solid #b00000;
+ margin: 2px;
+ padding: 2px;
+}
+H2 { font-size: 140%; }
+H3 { font-size: 100%; } */
+
+CAPTION { font-weight: bold }
+DIV.qindex {
+ width: 100%;
+ background-color: #eeeeff;
+ border: 1px solid #b0b0b0;
+ text-align: center;
+ margin: 2px;
+ padding: 2px;
+ line-height: 140%;
+}
+DIV.nav {
+ width: 100%;
+ background-color: #eeeeff;
+ border: 1px solid #b0b0b0;
+ text-align: center;
+ margin: 2px;
+ padding: 2px;
+ line-height: 140%;
+}
+DIV.navtab {
+ background-color: #eeeeff;
+ border: 1px solid #b0b0b0;
+ text-align: center;
+ margin: 2px;
+ margin-right: 15px;
+ padding: 2px;
+}
+TD.navtab {
+ font-size: 90%;
+}
+A.qindex {
+ text-decoration: none;
+ font-weight: bold;
+ color: #1A419D;
+}
+A.qindex:visited {
+ text-decoration: none;
+ font-weight: bold;
+ color: #1A419D
+}
+A.qindex:hover {
+ text-decoration: none;
+ background-color: #ddddff;
+}
+A.qindexHL {
+ text-decoration: none;
+ font-weight: bold;
+ background-color: #6666cc;
+ color: #ffffff;
+ border: 1px double #9295C2;
+}
+A.qindexHL:hover {
+ text-decoration: none;
+ background-color: #6666cc;
+ color: #ffffff;
+}
+A.qindexHL:visited { text-decoration: none; background-color: #6666cc; color: #ffffff }
+A.el { text-decoration: none; font-weight: bold }
+A.elRef { font-weight: bold }
+A.code:link { text-decoration: none; font-weight: normal; color: #0000a0 }
+A.code:visited { text-decoration: none; font-weight: normal; color: #0000a0 }
+A.codeRef:link { font-weight: normal; color: #0000FF}
+A.codeRef:visited { font-weight: normal; color: #0000FF}
+A:hover { text-decoration: none; background-color: #f2f2ff }
+DL.el { margin-left: -1cm }
+.fragment {
+ font-family: Fixed, monospace
+ font-size: 100%;
+}
+PRE.fragment {
+ font-size: normal;
+ border: 1px solid #CCCCCC;
+ background-color: #f5f5f5;
+ margin-top: 4px;
+ margin-bottom: 4px;
+ margin-left: 2px;
+ margin-right: 8px;
+ padding-left: 6px;
+ padding-right: 6px;
+ padding-top: 4px;
+ padding-bottom: 4px;
+}
+DIV.ah { background-color: black; font-weight: bold; color: #ffffff; margin-bottom: 3px; margin-top: 3px }
+TD.md { background-color: #F4F4FB; font-weight: bold; }
+TD.mdPrefix {
+ background-color: #F4F4FB;
+ color: #606060;
+ font-size: 90%;
+}
+TD.mdname1 { background-color: #F4F4FB; font-weight: bold; color: #602020; }
+TD.mdname { background-color: #F4F4FB; font-weight: bold; color: #602020; width: 600px; }
+DIV.groupHeader {
+ margin-left: 16px;
+ margin-top: 12px;
+ margin-bottom: 6px;
+ padding: 3px;
+ font-weight: bold;
+ font-size: 110%;
+ background-color: #d0d0ff;
+}
+DIV.groupText { margin-left: 16px; font-style: italic; font-size: 90%; }
+BODY {
+ background: white;
+ color: black;
+ margin-right: 20px;
+ margin-left: 20px;
+}
+TD.indexkey {
+ background-color: #eeeeff;
+ font-weight: bold;
+ padding-right : 10px;
+ padding-top : 2px;
+ padding-left : 10px;
+ padding-bottom : 2px;
+ margin-left : 0px;
+ margin-right : 0px;
+ margin-top : 2px;
+ margin-bottom : 2px;
+ border: 1px solid #CCCCCC;
+}
+TD.indexvalue {
+ background-color: #eeeeff;
+ font-style: italic;
+ padding-right : 10px;
+ padding-top : 2px;
+ padding-left : 10px;
+ padding-bottom : 2px;
+ margin-left : 0px;
+ margin-right : 0px;
+ margin-top : 2px;
+ margin-bottom : 2px;
+ border: 1px solid #CCCCCC;
+}
+TR.memlist {
+ background-color: #f0f0f0;
+}
+P.formulaDsp { text-align: center; }
+IMG.formulaDsp { }
+IMG.formulaInl { vertical-align: middle; }
+SPAN.keyword { color: #0000ff }
+SPAN.keywordtype { color: #0000ff }
+SPAN.keywordflow { color: #0000ff }
+SPAN.comment { color: #008000 }
+SPAN.preprocessor { color: #806020 }
+SPAN.stringliteral { color: #800000 }
+SPAN.charliteral { color: #800000 }
+.mdTable {
+ border: 1px solid #868686;
+ background-color: #F4F4FB;
+}
+.mdRow {
+ padding: 8px 10px;
+}
+.mdescLeft {
+ padding: 0px 8px 4px 8px;
+ font-size: 12px;
+ font-style: italic;
+ background-color: #FAFAFA;
+ border-top: 1px none #E0E0E0;
+ border-right: 1px none #E0E0E0;
+ border-bottom: 1px none #E0E0E0;
+ border-left: 1px none #E0E0E0;
+ margin: 0px;
+}
+.mdescRight {
+ padding: 0px 8px 4px 8px;
+ font-size: 12px;
+ font-style: italic;
+ background-color: #FAFAFA;
+ border-top: 1px none #E0E0E0;
+ border-right: 1px none #E0E0E0;
+ border-bottom: 1px none #E0E0E0;
+ border-left: 1px none #E0E0E0;
+ margin: 0px;
+}
+.memItemLeft {
+ padding: 1px 0px 0px 8px;
+ margin: 4px;
+ border-top-width: 1px;
+ border-right-width: 1px;
+ border-bottom-width: 1px;
+ border-left-width: 1px;
+ border-top-color: #E0E0E0;
+ border-right-color: #E0E0E0;
+ border-bottom-color: #E0E0E0;
+ border-left-color: #E0E0E0;
+ border-top-style: solid;
+ border-right-style: none;
+ border-bottom-style: none;
+ border-left-style: none;
+ background-color: #FAFAFA;
+ font-size: 90%;
+}
+.memItemRight {
+ padding: 1px 8px 0px 8px;
+ margin: 4px;
+ border-top-width: 1px;
+ border-right-width: 1px;
+ border-bottom-width: 1px;
+ border-left-width: 1px;
+ border-top-color: #E0E0E0;
+ border-right-color: #E0E0E0;
+ border-bottom-color: #E0E0E0;
+ border-left-color: #E0E0E0;
+ border-top-style: solid;
+ border-right-style: none;
+ border-bottom-style: none;
+ border-left-style: none;
+ background-color: #FAFAFA;
+ font-size: 100%;
+}
+.memTemplItemLeft {
+ padding: 1px 0px 0px 8px;
+ margin: 4px;
+ border-top-width: 1px;
+ border-right-width: 1px;
+ border-bottom-width: 1px;
+ border-left-width: 1px;
+ border-top-color: #E0E0E0;
+ border-right-color: #E0E0E0;
+ border-bottom-color: #E0E0E0;
+ border-left-color: #E0E0E0;
+ border-top-style: none;
+ border-right-style: none;
+ border-bottom-style: none;
+ border-left-style: none;
+ background-color: #FAFAFA;
+ font-size: 90%;
+}
+.memTemplItemRight {
+ padding: 1px 8px 0px 8px;
+ margin: 4px;
+ border-top-width: 1px;
+ border-right-width: 1px;
+ border-bottom-width: 1px;
+ borde...
[truncated message content] |
|
From: <den...@us...> - 2009-07-13 19:55:07
|
Revision: 125
http://opentrep.svn.sourceforge.net/opentrep/?rev=125&view=rev
Author: denis_arnaud
Date: 2009-07-13 19:55:04 +0000 (Mon, 13 Jul 2009)
Log Message:
-----------
[Indexer] The words are no longer indexed with positions within documents.
Modified Paths:
--------------
trunk/opentrep/opentrep/command/IndexBuilder.cpp
trunk/opentrep/test/xapian/string_search.cpp
Modified: trunk/opentrep/opentrep/command/IndexBuilder.cpp
===================================================================
--- trunk/opentrep/opentrep/command/IndexBuilder.cpp 2009-07-13 18:03:01 UTC (rev 124)
+++ trunk/opentrep/opentrep/command/IndexBuilder.cpp 2009-07-13 19:55:04 UTC (rev 125)
@@ -48,15 +48,18 @@
(lCityCode.empty())?lPlaceCode:lCityCode;
const std::string& lStateCode = ioPlace.getStateCode();
const std::string lDBStateCode = (lStateCode.empty())?"NA":lStateCode;
+
+ // Word index/position within the Xapian document
+ unsigned short idx = 1;
// Add indexing terms
- lDocument.add_posting (lPlaceCode, 1);
- lDocument.add_posting (lDBCityCode, 2);
- lDocument.add_posting (lDBStateCode, 3);
- lDocument.add_posting (ioPlace.getCountryCode(), 4);
- lDocument.add_posting (ioPlace.getRegionCode(), 5);
- lDocument.add_posting (ioPlace.getContinentCode(), 6);
- lDocument.add_posting (ioPlace.getTimeZoneGroup(), 7);
+ lDocument.add_term (lPlaceCode); ++idx;
+ lDocument.add_term (lDBCityCode); ++idx;
+ lDocument.add_term (lDBStateCode); ++idx;
+ lDocument.add_term (ioPlace.getCountryCode()); ++idx;
+ lDocument.add_term (ioPlace.getRegionCode()); ++idx;
+ lDocument.add_term (ioPlace.getContinentCode()); ++idx;
+ lDocument.add_term (ioPlace.getTimeZoneGroup()); ++idx;
// Add terms to the spelling dictionnary
ioDatabase.add_spelling (lPlaceCode);
@@ -65,18 +68,20 @@
ioDatabase.add_spelling (lStateCode);
}
- // Retrieve the map of name lists
- unsigned int i = 1;
- const NameMatrix_T& lNameMatrix = ioPlace.getNameMatrix();
+ // Retrieve the place names in all the available languages
+ const NameMatrix_T& lNameMatrix = ioPlace.getNameMatrix ();
for (NameMatrix_T::const_iterator itNameList = lNameMatrix.begin();
- itNameList != lNameMatrix.end(); ++itNameList, ++i) {
+ itNameList != lNameMatrix.end(); ++itNameList) {
+ // Retrieve the language code and locale
+ const Language::EN_Language& lLanguage = itNameList->first;
const Names& lNames = itNameList->second;
- // Add the language code (e.g., en_US)
- lDocument.add_posting (lNames.describeShortKey(), 7+i);
- ++i;
+ // Add that language code and locale to the Xapian document
+ lDocument.add_term (Language::getLongLabel (lLanguage)); ++idx;
+ // For a given language, retrieve the list of place names
const NameList_T& lNameList = lNames.getNameList();
+
for (NameList_T::const_iterator itName = lNameList.begin();
itName != lNameList.end(); ++itName) {
const std::string& lName = *itName;
@@ -84,9 +89,9 @@
// Add the place name (it can be the classical one, or
// extended, alternate, etc.)
if (lName.empty() == false) {
- lDocument.add_posting (lName, 8+i);
+ // OPENTREP_LOG_DEBUG ("Added name: " << lName);
+ lDocument.add_term (lName); ++idx;
ioDatabase.add_spelling (lName);
- ++i;
}
}
}
Modified: trunk/opentrep/test/xapian/string_search.cpp
===================================================================
--- trunk/opentrep/test/xapian/string_search.cpp 2009-07-13 18:03:01 UTC (rev 124)
+++ trunk/opentrep/test/xapian/string_search.cpp 2009-07-13 19:55:04 UTC (rev 125)
@@ -28,6 +28,7 @@
for (int idx=2; idx != argc; ++idx) {
if (idx != 2) {
oStr << " ";
+// oStr << " AND ";
}
const std::string lWord (argv[idx]);
const std::string lSuggestedWord =
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-13 18:03:11
|
Revision: 124
http://opentrep.svn.sourceforge.net/opentrep/?rev=124&view=rev
Author: denis_arnaud
Date: 2009-07-13 18:03:01 +0000 (Mon, 13 Jul 2009)
Log Message:
-----------
[DB] Updated the MySQL database scripts for the user creation.
Modified Paths:
--------------
trunk/opentrep/refdata/mysql/create_opentrep_user.sh
trunk/opentrep/refdata/mysql/create_opentrep_user.sql
Modified: trunk/opentrep/refdata/mysql/create_opentrep_user.sh
===================================================================
--- trunk/opentrep/refdata/mysql/create_opentrep_user.sh 2009-07-13 00:11:02 UTC (rev 123)
+++ trunk/opentrep/refdata/mysql/create_opentrep_user.sh 2009-07-13 18:03:01 UTC (rev 124)
@@ -43,8 +43,10 @@
function createOpenTrepUser() {
echo "Creating the opentrep user within the database:"
mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} ${DB_NAME} < ${SQL_FILE}
+ mysql -u ${DB_USER} --password=${DB_PASSWD} -P ${DB_PORT} -h ${DB_HOST} -e "flush privileges"
}
# Creating the opentrep user
SQL_FILE="create_opentrep_user.sql"
createOpenTrepUser
+
Modified: trunk/opentrep/refdata/mysql/create_opentrep_user.sql
===================================================================
--- trunk/opentrep/refdata/mysql/create_opentrep_user.sql 2009-07-13 00:11:02 UTC (rev 123)
+++ trunk/opentrep/refdata/mysql/create_opentrep_user.sql 2009-07-13 18:03:01 UTC (rev 124)
@@ -1,3 +1,7 @@
+
insert into `user` (`Host`, `User`, `Password`, `Select_priv`, `Insert_priv`, `Update_priv`, `Delete_priv`, `Create_priv`, `Drop_priv`, `Reload_priv`, `Shutdown_priv`, `Process_priv`, `File_priv`, `Grant_priv`, `References_priv`, `Index_priv`, `Alter_priv`, `Show_db_priv`, `Super_priv`, `Create_tmp_table_priv`, `Lock_tables_priv`, `Execute_priv`, `Repl_slave_priv`, `Repl_client_priv`, `Create_view_priv`, `Show_view_priv`, `Create_routine_priv`, `Alter_routine_priv`, `Create_user_priv`, `ssl_type`, `ssl_cipher`, `x509_issuer`, `x509_subject`, `max_questions`, `max_updates`, `max_connections`, `max_user_connections`) values
('%', 'opentrep', '*C21B5F0DB6BBABAA20B5496E75D652982A6AC65C', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'N', 'N', 'N', 'Y', 'N', 'N', 'Y', 'Y', 'N', 'N', 'Y', 'N', 'Y', 'N', 'N', 'Y', 'Y', 'Y', 'Y', 'N', '', '', '', '', 0, 0, 0, 0),
('localhost', 'opentrep', '*C21B5F0DB6BBABAA20B5496E75D652982A6AC65C', 'Y', 'Y', 'Y', 'Y', 'Y', 'Y', 'N', 'N', 'N', 'Y', 'N', 'N', 'Y', 'Y', 'N', 'N', 'Y', 'N', 'Y', 'N', 'N', 'Y', 'Y', 'Y', 'Y', 'N', '', '', '', '', 0, 0, 0, 0);
+
+flush privileges;
+
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-13 00:11:12
|
Revision: 123
http://opentrep.svn.sourceforge.net/opentrep/?rev=123&view=rev
Author: denis_arnaud
Date: 2009-07-13 00:11:02 +0000 (Mon, 13 Jul 2009)
Log Message:
-----------
[Dev] Added the Ternary-Tree project, and a few examples within the parsers sub-directory.
Modified Paths:
--------------
trunk/opentrep/test/parsers/search_string_parser.cpp
Added Paths:
-----------
trunk/opentrep/test/parsers/levenshtein_new.cpp
trunk/opentrep/tools/
trunk/ternary_tree/
trunk/ternary_tree/README
trunk/ternary_tree/doxygen_input/
trunk/ternary_tree/doxygen_input/blather.hpp
trunk/ternary_tree/doxygen_input/concepts.txt
trunk/ternary_tree/doxygen_input/doxygen-old.css
trunk/ternary_tree/doxygen_input/doxygen.css
trunk/ternary_tree/doxygen_input/external.png
trunk/ternary_tree/doxygen_input/featuretable.html
trunk/ternary_tree/doxygen_input/footer_inc.html
trunk/ternary_tree/doxygen_input/header_inc.html
trunk/ternary_tree/doxygen_input/performancetable.html
trunk/ternary_tree/doxygen_input/tree - trie concepts.txt
trunk/ternary_tree/doxygen_input/usage.hpp
trunk/ternary_tree/examples/
trunk/ternary_tree/examples/examples.vcproj
trunk/ternary_tree/examples/locale_less.hpp
trunk/ternary_tree/examples.cpp
trunk/ternary_tree/fill_dictionary.cpp
trunk/ternary_tree/full-docs-index.html
trunk/ternary_tree/html/
trunk/ternary_tree/html/annotated.html
trunk/ternary_tree/html/blather_8hpp.html
trunk/ternary_tree/html/class_data_t.html
trunk/ternary_tree/html/class_data_t_01_5.html
trunk/ternary_tree/html/classcontainers_1_1search__results__list-members.html
trunk/ternary_tree/html/classcontainers_1_1search__results__list.html
trunk/ternary_tree/html/classcontainers_1_1search__results__list_1_1iterator-members.html
trunk/ternary_tree/html/classcontainers_1_1search__results__list_1_1iterator.html
trunk/ternary_tree/html/classcontainers_1_1structured__map-members.html
trunk/ternary_tree/html/classcontainers_1_1structured__map.html
trunk/ternary_tree/html/classcontainers_1_1structured__map_1_1value__compare-members.html
trunk/ternary_tree/html/classcontainers_1_1structured__map_1_1value__compare.html
trunk/ternary_tree/html/classcontainers_1_1structured__multimap-members.html
trunk/ternary_tree/html/classcontainers_1_1structured__multimap.html
trunk/ternary_tree/html/classcontainers_1_1structured__multimap_1_1value__compare-members.html
trunk/ternary_tree/html/classcontainers_1_1structured__multimap_1_1value__compare.html
trunk/ternary_tree/html/classcontainers_1_1structured__multiset-members.html
trunk/ternary_tree/html/classcontainers_1_1structured__multiset.html
trunk/ternary_tree/html/classcontainers_1_1structured__set-members.html
trunk/ternary_tree/html/classcontainers_1_1structured__set.html
trunk/ternary_tree/html/classcontainers_1_1ternary__tree-members.html
trunk/ternary_tree/html/classcontainers_1_1ternary__tree.html
trunk/ternary_tree/html/classcontainers_1_1ternary__tree_1_1key__compare-members.html
trunk/ternary_tree/html/classcontainers_1_1ternary__tree_1_1key__compare.html
trunk/ternary_tree/html/classcontainers_1_1tst__detail_1_1_base_t.html
trunk/ternary_tree/html/classcontainers_1_1tst__detail_1_1tst__iterator__base-members.html
trunk/ternary_tree/html/classcontainers_1_1tst__detail_1_1tst__iterator__base.html
trunk/ternary_tree/html/classstd_1_1back__insert__iterator.html
trunk/ternary_tree/html/classstd_1_1binary__function.html
trunk/ternary_tree/html/dir_0df55976ff011c1ef61da79183e9e28f.html
trunk/ternary_tree/html/dir_59457a7c227558cb0e28f31428e14f54.html
trunk/ternary_tree/html/dirs.html
trunk/ternary_tree/html/doxygen.css
trunk/ternary_tree/html/doxygen.png
trunk/ternary_tree/html/files.html
trunk/ternary_tree/html/functions.html
trunk/ternary_tree/html/functions_0x62.html
trunk/ternary_tree/html/functions_0x63.html
trunk/ternary_tree/html/functions_0x64.html
trunk/ternary_tree/html/functions_0x65.html
trunk/ternary_tree/html/functions_0x66.html
trunk/ternary_tree/html/functions_0x67.html
trunk/ternary_tree/html/functions_0x68.html
trunk/ternary_tree/html/functions_0x69.html
trunk/ternary_tree/html/functions_0x6b.html
trunk/ternary_tree/html/functions_0x6c.html
trunk/ternary_tree/html/functions_0x6d.html
trunk/ternary_tree/html/functions_0x6e.html
trunk/ternary_tree/html/functions_0x6f.html
trunk/ternary_tree/html/functions_0x70.html
trunk/ternary_tree/html/functions_0x72.html
trunk/ternary_tree/html/functions_0x73.html
trunk/ternary_tree/html/functions_0x74.html
trunk/ternary_tree/html/functions_0x75.html
trunk/ternary_tree/html/functions_0x76.html
trunk/ternary_tree/html/functions_0x77.html
trunk/ternary_tree/html/functions_0x7e.html
trunk/ternary_tree/html/functions_enum.html
trunk/ternary_tree/html/functions_eval.html
trunk/ternary_tree/html/functions_func.html
trunk/ternary_tree/html/functions_func_0x62.html
trunk/ternary_tree/html/functions_func_0x63.html
trunk/ternary_tree/html/functions_func_0x64.html
trunk/ternary_tree/html/functions_func_0x65.html
trunk/ternary_tree/html/functions_func_0x66.html
trunk/ternary_tree/html/functions_func_0x67.html
trunk/ternary_tree/html/functions_func_0x68.html
trunk/ternary_tree/html/functions_func_0x69.html
trunk/ternary_tree/html/functions_func_0x6b.html
trunk/ternary_tree/html/functions_func_0x6c.html
trunk/ternary_tree/html/functions_func_0x6d.html
trunk/ternary_tree/html/functions_func_0x6e.html
trunk/ternary_tree/html/functions_func_0x6f.html
trunk/ternary_tree/html/functions_func_0x70.html
trunk/ternary_tree/html/functions_func_0x72.html
trunk/ternary_tree/html/functions_func_0x73.html
trunk/ternary_tree/html/functions_func_0x74.html
trunk/ternary_tree/html/functions_func_0x75.html
trunk/ternary_tree/html/functions_func_0x76.html
trunk/ternary_tree/html/functions_func_0x7e.html
trunk/ternary_tree/html/functions_rela.html
trunk/ternary_tree/html/functions_type.html
trunk/ternary_tree/html/functions_type_0x62.html
trunk/ternary_tree/html/functions_type_0x63.html
trunk/ternary_tree/html/functions_type_0x64.html
trunk/ternary_tree/html/functions_type_0x66.html
trunk/ternary_tree/html/functions_type_0x68.html
trunk/ternary_tree/html/functions_type_0x69.html
trunk/ternary_tree/html/functions_type_0x6b.html
trunk/ternary_tree/html/functions_type_0x6c.html
trunk/ternary_tree/html/functions_type_0x6d.html
trunk/ternary_tree/html/functions_type_0x6e.html
trunk/ternary_tree/html/functions_type_0x70.html
trunk/ternary_tree/html/functions_type_0x72.html
trunk/ternary_tree/html/functions_type_0x73.html
trunk/ternary_tree/html/functions_type_0x74.html
trunk/ternary_tree/html/functions_type_0x76.html
trunk/ternary_tree/html/functions_vars.html
trunk/ternary_tree/html/globals.html
trunk/ternary_tree/html/globals_defs.html
trunk/ternary_tree/html/globals_func.html
trunk/ternary_tree/html/graph_legend.dot
trunk/ternary_tree/html/graph_legend.html
trunk/ternary_tree/html/graph_legend.png
trunk/ternary_tree/html/hierarchy.html
trunk/ternary_tree/html/index.html
trunk/ternary_tree/html/iteration__impl_8hpp.html
trunk/ternary_tree/html/iterator__wrapper_8hpp.html
trunk/ternary_tree/html/namespacecontainers.html
trunk/ternary_tree/html/namespacecontainers_1_1smap__detail.html
trunk/ternary_tree/html/namespacecontainers_1_1sset__detail.html
trunk/ternary_tree/html/namespacecontainers_1_1tst__detail.html
trunk/ternary_tree/html/namespacecontainers_1_1tst__detail_1_1mpl__detail.html
trunk/ternary_tree/html/namespacecontainers_1_1tst__erase__impl__detail.html
trunk/ternary_tree/html/namespacecontainers_1_1util.html
trunk/ternary_tree/html/namespaceiterators.html
trunk/ternary_tree/html/namespacemembers.html
trunk/ternary_tree/html/namespacemembers_func.html
trunk/ternary_tree/html/namespaces.html
trunk/ternary_tree/html/namespacestd.html
trunk/ternary_tree/html/new__iterator__base_8ipp.html
trunk/ternary_tree/html/pages.html
trunk/ternary_tree/html/perf_notes.html
trunk/ternary_tree/html/structcontainers_1_1smap__detail_1_1multimap__iterator-members.html
trunk/ternary_tree/html/structcontainers_1_1smap__detail_1_1multimap__iterator.html
trunk/ternary_tree/html/structcontainers_1_1sset__detail_1_1multiset__iterator-members.html
trunk/ternary_tree/html/structcontainers_1_1sset__detail_1_1multiset__iterator.html
trunk/ternary_tree/html/structcontainers_1_1ternary__tree_1_1find__result-members.html
trunk/ternary_tree/html/structcontainers_1_1ternary__tree_1_1find__result.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1always__heap__node-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1always__heap__node.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1back__push__pop-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1back__push__pop.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1dummy__sequence-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1dummy__sequence.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1heap__node-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1heap__node.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1inorder__seek-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1inorder__seek.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1inplace__node-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1inplace__node.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1iter__method__forward-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1iter__method__forward.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1key__access-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1key__access.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1levenshtein__search__info-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1levenshtein__search__info.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1levenshtein__search__info_1_1search-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1levenshtein__search__info_1_1search.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1mpl__detail_1_1if__c-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1mpl__detail_1_1if__c.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1mpl__detail_1_1if__c_3_01false_00_01_t1_00_01_t2_01_4-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1mpl__detail_1_1if__c_3_01false_00_01_t1_00_01_t2_01_4.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1node__base-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1node__base.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1size__policy__node-members.html
trunk/ternary_tree/html/structcontainers_1_1tst__detail_1_1size__policy__node.html
trunk/ternary_tree/html/structiterators_1_1const__traits-members.html
trunk/ternary_tree/html/structiterators_1_1const__traits.html
trunk/ternary_tree/html/structiterators_1_1iterator__wrapper-members.html
trunk/ternary_tree/html/structiterators_1_1iterator__wrapper.html
trunk/ternary_tree/html/structiterators_1_1nonconst__traits-members.html
trunk/ternary_tree/html/structiterators_1_1nonconst__traits.html
trunk/ternary_tree/html/structured__map_8hpp.html
trunk/ternary_tree/html/structured__set_8hpp.html
trunk/ternary_tree/html/structured_concept.html
trunk/ternary_tree/html/tab_b.gif
trunk/ternary_tree/html/tab_l.gif
trunk/ternary_tree/html/tab_r.gif
trunk/ternary_tree/html/tabs.css
trunk/ternary_tree/html/ternary__tree_8hpp.html
trunk/ternary_tree/html/todo.html
trunk/ternary_tree/html/tst__implementation_8ipp.html
trunk/ternary_tree/html/tst__iterator__base_8ipp.html
trunk/ternary_tree/html/tst__iterator__facade_8hpp.html
trunk/ternary_tree/html/tst__node_8hpp.html
trunk/ternary_tree/html/tst__search__results_8ipp.html
trunk/ternary_tree/html/tst_impl.html
trunk/ternary_tree/html/tst_links.html
trunk/ternary_tree/html/tst_reference.html
trunk/ternary_tree/html/tst_tests.html
trunk/ternary_tree/html/tst_usage.html
trunk/ternary_tree/html/usage_8hpp.html
trunk/ternary_tree/index.html
trunk/ternary_tree/iterator_compile_test.cpp
trunk/ternary_tree/iterator_wrapper.hpp
trunk/ternary_tree/readme.txt
trunk/ternary_tree/structured_map.hpp
trunk/ternary_tree/structured_set.hpp
trunk/ternary_tree/ternary_tree.hpp
trunk/ternary_tree/test/
trunk/ternary_tree/test/basic_insertion_test.hpp
trunk/ternary_tree/test/check_iteration.hpp
trunk/ternary_tree/test/copy_test.hpp
trunk/ternary_tree/test/element_range_test.hpp
trunk/ternary_tree/test/erase_test.cpp
trunk/ternary_tree/test/hamming_search_test.cpp
trunk/ternary_tree/test/iterator_test.cpp
trunk/ternary_tree/test/localization_test.cpp
trunk/ternary_tree/test/longest_match_test.cpp
trunk/ternary_tree/test/mapped_value_test.cpp
trunk/ternary_tree/test/partial_match_test.cpp
trunk/ternary_tree/test/prefix_range_test.cpp
trunk/ternary_tree/test/scrabble_search_test.cpp
trunk/ternary_tree/test/test.vcproj
trunk/ternary_tree/test/test_tst.cpp
trunk/ternary_tree/test/tests_common.hpp
trunk/ternary_tree/tst.doxy
trunk/ternary_tree/tst_concept_checks.cpp
trunk/ternary_tree/tst_detail/
trunk/ternary_tree/tst_detail/iteration_impl.hpp
trunk/ternary_tree/tst_detail/new_iterator_base.ipp
trunk/ternary_tree/tst_detail/tst_implementation.ipp
trunk/ternary_tree/tst_detail/tst_iterator_base.ipp
trunk/ternary_tree/tst_detail/tst_iterator_facade.hpp
trunk/ternary_tree/tst_detail/tst_node.hpp
trunk/ternary_tree/tst_detail/tst_search_results.ipp
trunk/ternary_tree/tst_public.doxy
Added: trunk/opentrep/test/parsers/levenshtein_new.cpp
===================================================================
--- trunk/opentrep/test/parsers/levenshtein_new.cpp (rev 0)
+++ trunk/opentrep/test/parsers/levenshtein_new.cpp 2009-07-13 00:11:02 UTC (rev 123)
@@ -0,0 +1,316 @@
+//
+// Source: http://www.codeproject.com/KB/recipes/Levenshtein.aspx
+//
+using System;
+using System.Collections.Generic;
+using System.Text;
+
+namespace Test
+{
+ class Program
+ {
+ static void Main(string[] args)
+ {
+ string sNew = @"GAMBOL
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+";
+ string sOld = @"GUMBO
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+";
+
+
+ try
+ {
+ Levenshtein l = new Levenshtein();
+
+
+ while (true)
+ {
+ /// Original version
+ ///
+ ///
+ /* Read the initial time. */
+ DateTime startTime = DateTime.Now;
+
+ l.LD(sNew, sOld);
+
+ /* Read the end time. */
+ DateTime stopTime = DateTime.Now;
+
+ /* Compute the duration between the initial and the end time. */
+ TimeSpan duration = stopTime - startTime;
+ Console.WriteLine("Original:"+duration);
+
+
+ /// New version
+ ///
+ ///
+ /* Read the initial time. */
+ startTime = DateTime.Now;
+
+ l.iLD(sNew, sOld);
+
+ /* Read the end time. */
+ stopTime = DateTime.Now;
+
+ /* Compute the duration between the initial and the end time. */
+ duration = stopTime - startTime;
+ Console.WriteLine("New :"+duration);
+
+ Console.WriteLine("----------------");
+ }
+ }
+ catch (Exception e)
+ {
+ System.Console.WriteLine( e.ToString());
+ }
+
+ }
+ }
+
+ public class Levenshtein
+ {
+
+
+ ///*****************************
+ /// Compute Levenshtein distance
+ /// Memory efficient version
+ ///*****************************
+ public int iLD(String sRow, String sCol)
+ {
+ int RowLen = sRow.Length; // length of sRow
+ int ColLen = sCol.Length; // length of sCol
+ int RowIdx; // iterates through sRow
+ int ColIdx; // iterates through sCol
+ char Row_i; // ith character of sRow
+ char Col_j; // jth character of sCol
+ int cost; // cost
+
+ /// Test string length
+ if (Math.Max(sRow.Length, sCol.Length) > Math.Pow(2, 31))
+ throw (new Exception("\nMaximum string length in Levenshtein.iLD is " + Math.Pow(2, 31) + ".\nYours is " + Math.Max(sRow.Length, sCol.Length) + "."));
+
+ // Step 1
+
+ if (RowLen == 0)
+ {
+ return ColLen;
+ }
+
+ if (ColLen == 0)
+ {
+ return RowLen;
+ }
+
+ /// Create the two vectors
+ int[] v0 = new int[RowLen + 1];
+ int[] v1 = new int[RowLen + 1];
+ int[] vTmp;
+
+
+
+ /// Step 2
+ /// Initialize the first vector
+ for (RowIdx = 1; RowIdx <= RowLen; RowIdx++)
+ {
+ v0[RowIdx] = RowIdx;
+ }
+
+ // Step 3
+
+ /// Fore each column
+ for (ColIdx = 1; ColIdx <= ColLen; ColIdx++)
+ {
+ /// Set the 0'th element to the column number
+ v1[0] = ColIdx;
+
+ Col_j = sCol[ColIdx - 1];
+
+
+ // Step 4
+
+ /// Fore each row
+ for (RowIdx = 1; RowIdx <= RowLen; RowIdx++)
+ {
+ Row_i = sRow[RowIdx - 1];
+
+
+ // Step 5
+
+ if (Row_i == Col_j)
+ {
+ cost = 0;
+ }
+ else
+ {
+ cost = 1;
+ }
+
+ // Step 6
+
+ /// Find minimum
+ int m_min = v0[RowIdx] + 1;
+ int b = v1[RowIdx - 1] + 1;
+ int c = v0[RowIdx - 1] + cost;
+
+ if (b < m_min)
+ {
+ m_min = b;
+ }
+ if (c < m_min)
+ {
+ m_min = c;
+ }
+
+ v1[RowIdx] = m_min;
+ }
+
+ /// Swap the vectors
+ vTmp = v0;
+ v0 = v1;
+ v1 = vTmp;
+
+ }
+
+
+ // Step 7
+
+ /// Value between 0 - 100
+ /// 0==perfect match 100==totaly different
+ ///
+ /// The vectors where swaped one last time at the end of the last loop,
+ /// that is why the result is now in v0 rather than in v1
+ System.Console.WriteLine("iDist=" + v0[RowLen]);
+ int max = System.Math.Max(RowLen, ColLen);
+ return ((100 * v0[RowLen]) / max);
+ }
+
+
+
+
+
+ ///*****************************
+ /// Compute the min
+ ///*****************************
+
+ private int Minimum(int a, int b, int c)
+ {
+ int mi = a;
+
+ if (b < mi)
+ {
+ mi = b;
+ }
+ if (c < mi)
+ {
+ mi = c;
+ }
+
+ return mi;
+ }
+
+ ///*****************************
+ /// Compute Levenshtein distance
+ ///*****************************
+
+ public int LD(String sNew, String sOld)
+ {
+ int[,] matrix; // matrix
+ int sNewLen = sNew.Length; // length of sNew
+ int sOldLen = sOld.Length; // length of sOld
+ int sNewIdx; // iterates through sNew
+ int sOldIdx; // iterates through sOld
+ char sNew_i; // ith character of sNew
+ char sOld_j; // jth character of sOld
+ int cost; // cost
+
+ /// Test string length
+ if (Math.Max(sNew.Length, sOld.Length) > Math.Pow(2, 31))
+ throw (new Exception("\nMaximum string length in Levenshtein.LD is " + Math.Pow(2, 31) + ".\nYours is " + Math.Max(sNew.Length, sOld.Length) + "."));
+
+ // Step 1
+
+ if (sNewLen == 0)
+ {
+ return sOldLen;
+ }
+
+ if (sOldLen == 0)
+ {
+ return sNewLen;
+ }
+
+ matrix = new int[sNewLen + 1, sOldLen + 1];
+
+ // Step 2
+
+ for (sNewIdx = 0; sNewIdx <= sNewLen; sNewIdx++)
+ {
+ matrix[sNewIdx, 0] = sNewIdx;
+ }
+
+ for (sOldIdx = 0; sOldIdx <= sOldLen; sOldIdx++)
+ {
+ matrix[0, sOldIdx] = sOldIdx;
+ }
+
+ // Step 3
+
+ for (sNewIdx = 1; sNewIdx <= sNewLen; sNewIdx++)
+ {
+ sNew_i = sNew[sNewIdx - 1];
+
+ // Step 4
+
+ for (sOldIdx = 1; sOldIdx <= sOldLen; sOldIdx++)
+ {
+ sOld_j = sOld[sOldIdx - 1];
+
+ // Step 5
+
+ if (sNew_i == sOld_j)
+ {
+ cost = 0;
+ }
+ else
+ {
+ cost = 1;
+ }
+
+ // Step 6
+
+ matrix[sNewIdx, sOldIdx] = Minimum(matrix[sNewIdx - 1, sOldIdx] + 1, matrix[sNewIdx, sOldIdx - 1] + 1, matrix[sNewIdx - 1, sOldIdx - 1] + cost);
+
+ }
+ }
+
+ // Step 7
+
+ /// Value between 0 - 100
+ /// 0==perfect match 100==totaly different
+ System.Console.WriteLine("Dist=" + matrix[sNewLen, sOldLen]);
+ int max = System.Math.Max(sNewLen, sOldLen);
+ return (100 * matrix[sNewLen, sOldLen]) / max;
+ }
+ }
+
+
+}
Modified: trunk/opentrep/test/parsers/search_string_parser.cpp
===================================================================
--- trunk/opentrep/test/parsers/search_string_parser.cpp 2009-07-12 16:27:02 UTC (rev 122)
+++ trunk/opentrep/test/parsers/search_string_parser.cpp 2009-07-13 00:11:02 UTC (rev 123)
@@ -81,7 +81,9 @@
Airline_T () : _isPreferred (true), _name(""), _code("") {}
/* Display. */
void display() const {
- std::cout << "Airline: " << _name << " (" << _code << ")" << std::endl;
+ const std::string isPreferredStr = (_isPreferred)?"+":"-";
+ std::cout << "Airline: " << isPreferredStr << _name << " (" << _code << ")"
+ << std::endl;
}
};
@@ -210,7 +212,7 @@
: _searchString (ioSearchString) {}
void operator() (bool iAirlineSign) const {
- _searchString._tmpAirline._isPreferred = iAirlineSign;
+ _searchString._tmpAirline._isPreferred = !iAirlineSign;
// std::cout << "Airline is preferred: " << iAirlineSign << std::endl;
}
@@ -281,6 +283,40 @@
SearchString_T& _searchString;
};
+ /** Store the parsed passenger type. */
+ struct store_child_passenger_type {
+ store_child_passenger_type (SearchString_T& ioSearchString)
+ : _searchString (ioSearchString) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lPassengerType (iStr, iStrEnd);
+ _searchString._tmpPassenger._type = Passenger_T::CHILD;
+ // std::cout << "Passenger type: " << lPassengerType << std::endl;
+
+ // Add the parsed passenger to the list
+ _searchString._passengerList.push_back (_searchString._tmpPassenger);
+ }
+
+ SearchString_T& _searchString;
+ };
+
+ /** Store the parsed passenger type. */
+ struct store_pet_passenger_type {
+ store_pet_passenger_type (SearchString_T& ioSearchString)
+ : _searchString (ioSearchString) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lPassengerType (iStr, iStrEnd);
+ _searchString._tmpPassenger._type = Passenger_T::PET;
+ // std::cout << "Passenger type: " << lPassengerType << std::endl;
+
+ // Add the parsed passenger to the list
+ _searchString._passengerList.push_back (_searchString._tmpPassenger);
+ }
+
+ SearchString_T& _searchString;
+ };
+
}
// /////////// Utilities /////////////
@@ -288,8 +324,12 @@
boost::spirit::int_parser<unsigned int, 10, 1, 1> int1_p;
/** 1-digit-integer parser */
boost::spirit::uint_parser<unsigned int, 10, 1, 1> uint1_p;
+/** Up-to-2-digit-integer parser */
+boost::spirit::uint_parser<unsigned int, 10, 1, 2> uint1_2_p;
/** 2-digit-integer parser */
boost::spirit::uint_parser<int, 10, 2, 2> uint2_p;
+/** Up-to-4-digit-integer parser */
+boost::spirit::uint_parser<int, 10, 2, 4> uint2_4_p;
/** 4-digit-integer parser */
boost::spirit::uint_parser<int, 10, 4, 4> uint4_p;
/** Up-to-4-digit-integer parser */
@@ -360,13 +400,26 @@
;
date =
- lexeme_d[ limit_d(1u,12u)[uint2_p][assign_a(self._searchString._tmpDate._month)]
- >> boost::spirit::chset_p("/-")
- >> limit_d(1u,31u)[uint2_p][assign_a(self._searchString._tmpDate._day)]
- >> boost::spirit::chset_p("/-")
- >> limit_d(2000u,2099u)[uint4_p][assign_a(self._searchString._tmpDate._year)]]
+ ( month | day )
+ >> boost::spirit::chset_p("/-")
+ >> ( day | month )
+ >> ! ( boost::spirit::chset_p("/-")
+ >> year )
;
+ day =
+ lexeme_d[ limit_d(1u,31u)[uint1_2_p][assign_a(self._searchString._tmpDate._day)] ]
+ ;
+
+ month =
+ lexeme_d[ limit_d(1u,12u)[uint1_2_p][assign_a(self._searchString._tmpDate._month)] ]
+ ;
+
+ year =
+ lexeme_d[ limit_d(2000u,2099u)[uint4_p][assign_a(self._searchString._tmpDate._year)] ]
+ | lexeme_d[ limit_d(0u,99u)[uint2_p][assign_a(self._searchString._tmpDate._year)] ]
+ ;
+
preferred_airlines =
!(boost::spirit::sign_p)[store_airline_sign(self._searchString)]
>> airline_code | airline_name
@@ -389,25 +442,47 @@
;
passenger_type =
- lexeme_d[ (repeat_p(3,20)[chset_p("0-9a-z")])[store_adult_passenger_type(self._searchString)] ]
+ passenger_adult_type[store_adult_passenger_type(self._searchString)]
+ | passenger_child_type[store_child_passenger_type(self._searchString)]
+ | passenger_pet_type[store_pet_passenger_type(self._searchString)]
;
+ passenger_adult_type =
+ lexeme_d[ as_lower_d [ str_p("adult") >> !ch_p('s') ] ]
+ ;
+
+ passenger_child_type =
+ lexeme_d[ as_lower_d [ str_p("child") >> !str_p("ren") ] ]
+ ;
+
+ passenger_pet_type =
+ lexeme_d[ as_lower_d [ str_p("dog") | str_p("cat") >> !ch_p('s') ] ]
+ ;
+
BOOST_SPIRIT_DEBUG_NODE (search_string);
BOOST_SPIRIT_DEBUG_NODE (places);
BOOST_SPIRIT_DEBUG_NODE (place_element);
BOOST_SPIRIT_DEBUG_NODE (dates);
BOOST_SPIRIT_DEBUG_NODE (date);
+ BOOST_SPIRIT_DEBUG_NODE (day);
+ BOOST_SPIRIT_DEBUG_NODE (month);
+ BOOST_SPIRIT_DEBUG_NODE (year);
BOOST_SPIRIT_DEBUG_NODE (preferred_airlines);
BOOST_SPIRIT_DEBUG_NODE (airline_code);
BOOST_SPIRIT_DEBUG_NODE (airline_name);
BOOST_SPIRIT_DEBUG_NODE (passengers);
BOOST_SPIRIT_DEBUG_NODE (passenger_number);
BOOST_SPIRIT_DEBUG_NODE (passenger_type);
+ BOOST_SPIRIT_DEBUG_NODE (passenger_adult_type);
+ BOOST_SPIRIT_DEBUG_NODE (passenger_child_type);
+ BOOST_SPIRIT_DEBUG_NODE (passenger_pet_type);
}
boost::spirit::rule<ScannerT> search_string, places, place_element,
- dates, date, preferred_airlines, airline_code, airline_name,
- passengers, passenger_number, passenger_type;
+ dates, date, month, day, year,
+ preferred_airlines, airline_code, airline_name,
+ passengers, passenger_number, passenger_type, passenger_adult_type,
+ passenger_child_type, passenger_pet_type;
boost::spirit::rule<ScannerT> const& start() const { return search_string; }
};
@@ -415,42 +490,93 @@
SearchString_T& _searchString;
};
+// //////////////////////////////////////////////////////////
+bool parseSearchString (std::ostream& oStream,
+ const std::string& iSearchString,
+ SearchString_T& ioSearchStringStruct) {
+ bool oHasBeenParsingSuccessful = false;
+
+ // Read the search string
+ iterator_t lStringIterator = iSearchString.c_str();
+
+ // Instantiate the structure that will hold the result of the parsing.
+ SearchStringParser lSearchStringParser (ioSearchStringStruct);
+ boost::spirit::parse_info<iterator_t> info =
+ boost::spirit::parse (lStringIterator, lSearchStringParser,
+ boost::spirit::space_p);
+
+ oStream << "-------------------------" << std::endl;
+
+ oHasBeenParsingSuccessful = info.full;
+ if (oHasBeenParsingSuccessful == true) {
+ oStream << "Parsing succeeded" << std::endl;
+
+ } else {
+ oStream << "Parsing failed" << std::endl;
+ }
+ oStream << "-------------------------" << std::endl;
+
+ return oHasBeenParsingSuccessful;
+}
+
// /////////////// M A I N /////////////////
int main (int argc, char* argv[]) {
try {
- // File to be parsed
- std::string lSearchString ("guadeloupe rio de janeiro 07/22/2009 +aa -ua 2 adults 1 dog");
-
- // Read the command-line parameters
- if (argc >= 1 && argv[1] != NULL) {
- std::istringstream istr (argv[1]);
- istr >> lSearchString;
- }
+ // String to be parsed
+ const std::string lSearchString1 ("guadeloupe rio de janeiro 07/22/2009 +aa -ua 2 adults 3 children 1 dog");
+ const std::string lSearchString2 ("nce rekjavik 07-22-2009 +aa -ua 2 ADULTS 3 CHILDREN 1 DOG");
+ const std::string lSearchString3 ("sna francisco ods 07.22.2009 +aa");
+ const std::string lSearchString4 ("nce ods 07/22 +aa");
+ const std::string lSearchString5 ("nce ods 22/07 +aa");
- // Read the search string
- iterator_t lStringIterator = lSearchString.c_str();
+ // Parse the search string
+ SearchString_T lSearchStringStruct1;
+ bool hasBeenParsingSuccessful = parseSearchString (std::cout,
+ lSearchString1,
+ lSearchStringStruct1);
- // Instantiate the structure that will hold the result of the parsing.
- SearchString_T lSearchStringStruct;
- SearchStringParser lSearchStringParser (lSearchStringStruct);
- boost::spirit::parse_info<iterator_t> info =
- boost::spirit::parse (lStringIterator, lSearchStringParser,
- boost::spirit::space_p);
+ // Parse the search string
+ SearchString_T lSearchStringStruct2;
+ hasBeenParsingSuccessful = parseSearchString (std::cout, lSearchString2,
+ lSearchStringStruct2);
+ // Parse the search string
+ SearchString_T lSearchStringStruct3;
+ hasBeenParsingSuccessful = parseSearchString (std::cout, lSearchString3,
+ lSearchStringStruct3);
+
+ // Parse the search string
+ SearchString_T lSearchStringStruct4;
+ hasBeenParsingSuccessful = parseSearchString (std::cout, lSearchString4,
+ lSearchStringStruct4);
+
+ // Parse the search string
+ SearchString_T lSearchStringStruct5;
+ hasBeenParsingSuccessful = parseSearchString (std::cout, lSearchString1,
+ lSearchStringStruct5);
+
// DEBUG
- std::cout << "Search string:" << std::endl;
- lSearchStringStruct.display();
+ std::cout << "--------------------------------------------" << std::endl;
+ std::cout << "Search string1: " << lSearchString1 << std::endl;
+ lSearchStringStruct1.display();
- std::cout << "-------------------------" << std::endl;
- if (info.full) {
- std::cout << "Parsing succeeded" << std::endl;
-
- } else {
- std::cout << "Parsing failed" << std::endl;
- }
- std::cout << "-------------------------" << std::endl;
+ std::cout << "--------------------------------------------" << std::endl;
+ std::cout << "Search string2: " << lSearchString2 << std::endl;
+ lSearchStringStruct2.display();
+ std::cout << "--------------------------------------------" << std::endl;
+ std::cout << "Search string3: " << lSearchString3 << std::endl;
+ lSearchStringStruct3.display();
+
+ std::cout << "--------------------------------------------" << std::endl;
+ std::cout << "Search string4: " << lSearchString4 << std::endl;
+ lSearchStringStruct4.display();
+
+ std::cout << "--------------------------------------------" << std::endl;
+ std::cout << "Search string5: " << lSearchString5 << std::endl;
+ lSearchStringStruct5.display();
+
} catch (const std::exception& stde) {
std::cerr << "Standard exception: " << stde.what() << std::endl;
return -1;
Added: trunk/ternary_tree/README
===================================================================
--- trunk/ternary_tree/README (rev 0)
+++ trunk/ternary_tree/README 2009-07-13 00:11:02 UTC (rev 123)
@@ -0,0 +1,3 @@
+
+Source: http://abc.se/~re/code/tst and http://abc.se/~re/code/tst/ternary_tree.zip
+
Added: trunk/ternary_tree/doxygen_input/blather.hpp
===================================================================
--- trunk/ternary_tree/doxygen_input/blather.hpp (rev 0)
+++ trunk/ternary_tree/doxygen_input/blather.hpp 2009-07-13 00:11:02 UTC (rev 123)
@@ -0,0 +1,558 @@
+/** \mainpage Structured Associative Containers
+
+Ternary Search Tree containers to replace \c set<string> and \c map<string, Value> </h2>
+
+<center><table bgcolor="#fbf9e5" style="border: thin dotted #808000;" width="95%" border=0>
+<tr>
+<td>
+<h3>Table of contents</h3>
+<dl>
+ <dt>\ref introduction "Introduction"</dt>
+ <dt>\ref subkey_search_overview "Advanced searches overview"</dt>
+ <dt>\ref tst_usage "Tutorial"</dt>
+ <dt>\ref tst_reference "Reference"</dt> <dd>
+ <dd>\ref structured_concept "Structured Container concept" \n
+ Class \ref containers::structured_set "structured_set" \n
+ Class \ref containers::structured_map "structured_map" \n
+ Class \ref containers::structured_multiset "structured_multiset" \n
+ Class \ref containers::structured_multimap "structured_multimap" \n
+ Implementation class \ref containers::ternary_tree "ternary_tree"
+ </dd></dt>
+ <dt>\ref perf_notes "Performance notes"</dt>
+ <dt>\ref tst_impl "Implementation details"</dt>
+ <dt>\ref tst_links "Links"</dt>
+ <dt>\ref tst_tests "Test Suite"</dt>
+</dl>
+</td>
+</tr></table></center>
+
+Download: Latest version (0.684) http://abc.se/~re/code/tst/ternary_tree.zip\n
+
+Copyleft: <a href="mailto:rasmus%20point%20ekman%20at%20abc%20point%20se?subject=Structured Containers suck/rule">
+rasmus ekman</a> 2007-2009 \n
+Weblink: http://abc.se/~re/code/tst
+
+\anchor introduction <hr>
+<h2>Introduction</h2>
+<b>Structured containers</b> are \c map and \c set -like containers specialized for strings.
+They are commonly used for dictionaries.\n
+Structured containers have two major benefits:
+- They offer near-match searches (wildcard search, partial match etc) that are hard to implement
+ with other containers.
+- Lookup performance is on a par with hashed containers for many common applications,
+and 2-5 times faster than standard maps and sets (with string-like keys).
+
+Of course there is a price to pay: structured containers use much more memory than
+other containers: Around 6-8 bytes <b>per letter</b> inserted (whether \c char or \c wchar_t);
+an English 150 k word dictionary uses eg 7.3 MB to store 1.2 MB words (2.4 MB of \c wchar_t words).
+
+The container classes in this library can be used as drop-in replacements for \c set and \c map
+(or \c unordered_set, \c unordered_map):
+ - \ref containers::structured_set "structured_set": This stores unique keys and allows structured key searches.
+ - \ref containers::structured_multiset "structured_multiset": This stores non-unique keys.
+ - \ref containers::structured_map "structured_map": This is a
+ <a target="sgi" href="http://www.sgi.com/tech/stl/PairAssociativeContainer.html">Pair Associative Container</a>,
+ as it allows associating a value with each key.
+ - \ref containers::structured_multimap "structured_multimap": Technically, a
+ <a target="sgi" href="http://www.sgi.com/tech/stl/MultipleSortedAssociativeContainer.html">Multiple, Sorted,
+ Pair Associative Container</a> - it allows storing several values with each key.
+
+While the STL standard associative containers are normally backed by a binary tree structure,
+Structured Containers are backed by a Ternary Search Tree, as presented by
+\ref note_1 "Jon Bentley and Robert Sedgewick in [1]".
+
+Class \ref containers::ternary_tree "ternary_tree<Key, Value, Comp, Alloc>" provides the implementation backend.
+Due to its internals, its interface cannot easily be made to conform with standard STL concepts,
+so it is used internally by the structured* wrapper classes (much like STL's internal \c rb_tree class).
+
+Basically, if you have code using sets or maps, you have code to use structured containers.
+And with 1-3 lines of code, you're ready to make advanced imprecise searches in your dictionaries.\n
+See \ref tst_usage "the usage section" for examples of how to use these classes.
+
+<table bgcolor="#f0f0ff" style="border: thin dotted #808000;" border=0>
+<tr><th>Library status</th></tr>
+<tr><td valign="top" align="right">Compatibility:</td>
+<td>Note that the file \b tst_concept_checks.cpp is currently broken. Will investigate.\n
+<!-- This used to compile with Mingw GCC 3.4.2 and with MSVC7.1 (with STLport 5). Requires Boost 1.33.
+Not sure what happened in Boost 1.36-37 or if I've mangled something. \n
+Due to recent changes, ternary tree does not support stateful allocators (earlier versions did this by implication) -->
+</td>
+<tr><td valign="top" align="right">version 0.684: (Jan 2009)</td>
+<td>Fix standard-breakage in multimap/multiset return from <code>insert(const value_type&)</code>.<br>
+Added <code>operator-></code> to iterator wrapper for C++0x compatibility.
+Thanks to Geoffrey Noel for reports.</td>
+</tr>
+<tr><td valign="top" align="right">version 0.683: (March 2007)</td>
+<td>Fix portability issues for GCC and non-STLport libraries. Fix longest_match.<br>
+Thanks to Arjen Wagenaar for several reports, fixes and encouragement. Thanks also to Michel Tourn for reports.</td>
+</tr>
+<tr><td valign="top" align="right">version 0.68: (Dec 2006)</td>
+<td>Implement TST_NODE_COUNT_TYPE macro, which can be used to control node size on 64-bit systems.
+ See \ref containers::ternary_tree "class ternary_tree"</td>
+</tr>
+<tr><td valign="top" align="right">version 0.68 (alpha):</td>
+<td>Reimplemented node type. Do proper management of value type (was inconsistent, partly unimplemented - duh!)</td>
+</tr>
+<!--tr>
+<tr><td valign="top" align="right">version 0.676:</td>
+<td>Modified containers to follow C++0x draft standard: \n
+Added \c cbegin, \c cend methods returning \c const_iterator, and \c crbegin, \c crend
+returning \c const_reverse_iterator, to make it easier to code with const-correctness. \n
+\c erase(iterator pos); and \c erase(iterator first, iterator last); methods now return iterators.</td>
+<tr><td valign="top" align="right">version 0.675:</td>
+<td>All Structured Container classes implemented. Structured search interface TBD.
+</td-->
+</table>
+
+
+\anchor subkey_search_overview <hr>
+<h2>Sub-key, or Structure Searches</h2>
+<span style="color:#905050;">(a new interface for these searches will be specified in the future)</span>
+
+Ternary trees allow searches that match parts of keys and ignores mismatches in other parts.\n
+In the current interface we specify a small number of searches facilitated by the tree structure;
+the Partial Match and Hamming searches are defined in several other implementations
+(showcased in \ref note_1 "Bentley and Sedgewick" code).
+The Levenshtein and combinatorial searches are not found in other ternary trees (that I know of).
+
+<table border="1" cellspacing="0">
+ <tr><th bgcolor="#f0f0ff">Name (function name)</th><th bgcolor="#f0f0ff">Description</th></tr>
+ <tr><th>
+ Prefix match (\ref containers::ternary_tree::prefix_range "prefix_range")</th><td>
+ Finds keys sharing a common prefix, returns a pair of iterators.</td></tr>
+ <tr><th>
+ Longest match (\ref containers::ternary_tree::longest_match "longest_match")</th><td>
+ Finds the longest key that matches beginning of search string.
+ A typical application is to tokenize a string using the ternary tree as dictionary.</td></tr>
+ <tr><th>
+ Partial match, or wildcard search (\ref containers::ternary_tree::partial_match_search "partial_match_search")</th><td>
+ Accepts a search string with wildcard characters that will match any letter,
+ eg "b?nd" would match "band", "bend", "bind", "bond" in an English dictionary.</td></tr>
+ <tr><th>
+ Search allowing \c N mismatches,
+ (\ref containers::ternary_tree::hamming_search "hamming_search"<span style="font-weight:normal;"></span>)</th><td>
+ Accepts a search string and an integer \c dist indicating how many non-matching letters are allowed,
+ then finds keys matching search string that have at most \a dist mismatches.
+ This works like a partial match search with all combinations of \a dist
+ wildcards in the search string.\n
+ \c hamming_search("band", 1) matches the wildcard search plus "bald", "bane" and "wand", etc. \n
+ The version here, following DDJ code, extends the strict Hamming search by also allowing shorter and longer
+ strings; a search for "band", \a dist = 1, also finds "ban" and "bandy" etc.\n
+ See also http://wikipedia.org/wiki/Hamming_distance</td></tr>
+ <tr><th>
+ Levenshtein distance search</b> (\ref containers::ternary_tree::levenshtein_search "levenshtein_search"
+ <span style="font-weight:normal;">- consider descriptive name</span>)</th><td>
+
+ Hamming search matches characters in fixed position, allowing substitution of \a dist chars.
+ Levenshtein search also allows shifting parts of the search string by insertion or skipping chars (in \a dist places).
+ So <code>levenshtein_search("band", 1) </code> extends the hamming_search set with "and" and "bland", etc.
+ A typical application is to match mispelt words.\n
+ See also http://wikipedia.org/wiki/Levenshtein_distance</td></tr>
+ <tr><th>
+ Combinatorial or "scrabble" search (\ref containers::ternary_tree::combinatorial_search "combinatorial_search")</th><td>
+ Finds all keys using the characters in search string. \c combinatorial_search("band") finds
+ "ad", "and", "bad", "dab", "nab", etc. A count of wildcards can be added, also allowing
+ nonmatching characters (use with care, values over 10% of average key length
+ may cause the algorithm to traverse a large part of the tree).</td></tr>
+</table>
+
+See \ref usage_imprecise_searches "advanced search overview" in the tutorial.
+
+These searches are defined for all containers in this library.
+But they are also marked as deprecated (to be replaced by generic algorithms with same interface).
+For a relative performance comparison of imprecise searches, see the second table in \ref perf_notes.
+
+<h3>Future directions</h3>
+The searches currently defined are clearly special cases in a sea of search possibilities.
+We have only defined searches that are relatively efficient, compared to other combinations of containers and algorithms.
+But there can be many variations on the available searches: increasing Hamming/Levenshtein distance
+at the end of words, or matching limited ranges of characters (eg allowing mismatches only in vowels), etc.
+
+The next step for this project is to support a more flexible low-level interface for
+traversing and filtering tree nodes.
+The interface for these "structured searches" is open for consideration, but it
+will basically define sub-key iterators, conversion of full-key from sub-key iterators,
+and a small collection of algorithms operating on these sub-key iterators.
+
+At least the following operations are needed:
+
+ - sub-key match: matching a part of a key (prefix, or starting from current char position)
+ - key element range increment: from a sub-key position, match a range of characters
+ in next position (returns a list of sub-key iterators? - or iterator-like operation?)
+ - conversion from sub-key iterator to full-key iterator range (nearest and post-furthest
+ keys in the subtree)
+ - \c is_key(subkey_iterator pos): true if end-of-key exists at iterator position.
+ - \c count_elements(subkey_iterator pos): returns number of available key elements at position.
+ - In all predefined algorithms above, either a specific, or any char is matched,
+ we would also support arbitrary char sets (possibly with special case for char ranges).
+
+ */
+
+/** \page tst_reference Reference
+<center><table bgcolor="#fbf9e5" style="border: thin dotted #808000;" width="95%" border=0>
+<tr>
+<td>
+<dl>
+ <dt>\ref structured_concept "Structured Container concept"</dt>
+ <dt>\ref ref_sethpp "Header < structured_set.hpp >"</dt>
+ <dt>\ref ref_maphpp "Header < structured_map.hpp >"</dt>
+ <dt>\ref ref_tsthpp "Header < ternary_tree.hpp >"</dt>
+ <dt>\ref ref_iterhpp "Header < iterator_wrapper.hpp >"</dt>
+</dl>
+</td>
+</tr></table></center>
+
+<hr>
+
+\anchor ref_sethpp
+<h2>Header < <a href="../structured_set.hpp">%structured_set.hpp</a> > synopsis</h2>
+<pre>
+\b namespace containers {
+ \b template <\b class Key,
+ \b class Comp = std::less<\b typename Key::value_type>,
+ \b class Alloc = std::allocator<Key> >
+ \b class \ref containers::structured_set "structured_set";
+
+ \b template <\b class Key,
+ \b class Comp = std::less<\b typename Key::value_type>,
+ \b class Alloc = std::allocator<Key> >
+ \b class \ref containers::structured_multiset "structured_multiset";
+}
+</pre>
+
+\anchor ref_maphpp
+<h2>Header < <a href="../structured_map.hpp">%structured_map.hpp</a> > synopsis</h2>
+<pre>
+\b namespace containers {
+ \b template <\b class Key,
+ \b class T,
+ \b class Comp = std::less<\b typename Key::value_type>,
+ \b class Alloc = std::allocator<std::pair<\b const Key, T> > >
+ \b class \ref containers::structured_map "structured_map";
+
+ \b template <\b class Key,
+ \b class T,
+ \b class Comp = std::less<\b typename Key::value_type>,
+ \b class Alloc = std::allocator<std::pair<\b const Key, T> > >
+ \b class \ref containers::structured_multimap "structured_multimap";
+}
+</pre>
+
+<hr>
+Supplementary header files needed to support structured_set and -map classes.
+
+
+\anchor ref_tsthpp
+<h2>Header < <a href="../ternary_tree.hpp">%ternary_tree.hpp</a> > synopsis</h2>
+<pre>
+\b namespace containers {
+
+ \b template <\b class Key,
+ \b class T,
+ \b class Comp = std::less<\b typename Key::value_type>,
+ \b class Alloc = std::allocator<std::pair<\b const Key, T> > >
+ \b class \ref containers::ternary_tree "ternary_tree";
+
+ \b template <\b class TreeT, \b class IteratorT>
+ \b class \ref containers::search_results_list "search_results_list";
+
+}
+</pre>
+
+
+\anchor ref_iterhpp
+<h2>Header < <a href="../iterator_wrapper.hpp">%iterator_wrapper.hpp</a> > synopsis</h2>
+<pre>
+\b namespace iterators {
+
+ \b template <\b class T> \b struct const_traits;
+ \b template <\b class T> \b struct nonconst_traits;
+
+ \b template <\b class BaseIterT,
+ \b class TraitsT, // either const_traits<T> or nonconst_traits<T>
+ \b class IterCatT = std::bidirectional_iterator_tag >
+ \b class \ref iterators::iterator_wrapper "iterator_wrapper";
+}
+</pre>
+
+*/
+
+/**
+\page structured_concept Structured Associative Container Concept
+
+<span style="color:#905050;">(a preliminary sketch of the formal technical concept description)</span>
+
+A Structured Associative Container is a specialization of the C++ 98 standard concept
+<a target="sgi" href="http://www.sgi.com/tech/stl/SortedAssociativeContainer.html">Sorted Associative Container</a>,
+with extended interface.
+
+The template parameters are similar to that of the Associated Containers:
+
+<code> structured_set<Key, Comp, Alloc>; </code>\n
+<code> structured_map<Key, Value, Comp, Alloc>; </code>\n
+
+where:
+ - \c <b>Key</b> type is itself a container (eg a \c std::string or \c std::wstring)
+ - \c <b>Comp</b> is a comparison operator that imposes a sort order on \c Key::value_type elements \n
+ (so if \c Key is string, \c Comp compares \c char, if \c Key is \c wstring, \c Comp applies to \c wchar_t).
+ - \c <b>Value</b> can be any Assignable type
+ - \c <b>Alloc</b> is an allocator that manages all memory allocation for the container.
+
+The \c Comp and the \c Alloc types have default template arguments.
+
+In other words Structured containers are like Sorted Associative Containers, BUT
+ - add the requirement on Key template type to be a
+ <a target="sgi" href="http://www.sgi.com/tech/stl/ForwardContainer.html">Forward Container</a>.\n
+ For example, \c std::basic_string<CharT> is compatible with this requirement.
+ - change the requirement on the \c Comp (comparator) template argument to operate on
+ \c key_type::value_type elements (rather than on \c key_type itself).
+ Like Sorted Associative comparator, the \c Comp type shall define a less-like comparison, a
+ <a target="sgi" href="http://www.sgi.com/tech/stl/StrictWeakOrdering.html">Strict Weak Ordering</a>
+ of key-elements.
+
+<b>Associated types</b>
+ - \b char_compare: less-like comparison of key elements (establishing a Strict Weak Ordering).
+ The <a target="sgi" href="http://www.sgi.com/tech/stl/AssociativeContainer.html">Associative Container</a>
+ \c key_compare type is also provided, but is defined in terms of \c char_compare. \n
+ - \b subkey_iterator: Used in structure searches. Convertible to iterator (TBD).
+
+In consequence it allows searches involving subparts of keys, ie with shared prefix and/or
+with shared middle parts.
+
+<hr>
+<h3>Deprecated search interface</h3>
+
+In the first iteration, additional searches are provided as methods on the containers.
+This will be changed to use free functions operating on \c subkey_iterator.
+The deprecated search methods will still be provided as convenience functions;
+to migrate your code from present version to the new interface, will mean moving
+the object name to the first argument, but also to respecify the search_results_list type.
+(This sloppy-hackish type is by itself reason not to keep the method interface)
+
+See \ref subkey_search_overview "Structured search overview"
+and \ref tst_structsearch "ternary_tree Structure search section".
+*/
+
+/*
+
+\b Notation \n
+<table border=0>
+<tr><td>\c X <td>A type that is a model of Associative Container </td></tr>
+<tr><td>\c a <td>Object of type \c X </tr>
+<tr><td>\c k <td>Object of type \c X::key_type </tr>
+<tr><td>\c p, \c q <td>Object of type \c X::char_iterator </tr>
+<tr><td>\c c <td>Object of type \c X::char_type </tr>
+<tr><td>\c o <td>Object modelling output iterator </tr>
+<tr><td>\c i <td>Object of type \c X::size_type </tr>
+</dl>
+
+<table border=1>
+<tr><th>Name</th><th>Expression</th><th>Return value</th>
+<tr><td>Prefix match</td><td><code>a.prefix_range(k)</code></td><td>
+ \c std::pair<iterator, iterator> if \c a is mutable, otherwise <br>\c std::pair<const_iterator, const_iterator></td></tr>
+<tr><td>Longest match</td><td><code>a.longest_match(p, q)</code></td><td>
+ \c iterator if \c a is mutable, otherwise \c const_iterator</td></tr>
+<tr><td>Partial match, or <br>wildcard search</td><td><code>a.partial_match_search(k, o, c)</code></td><td>
+ The output iterator \c o</td></tr>
+<tr><td>Hamming search</td><td><code>a.hamming_search(k, o, i)</code></td><td>
+ The output iterator \c o</td></tr>
+<tr><td>Levenshtein search</td><td><code>a.levenshtein_search(k, o, i)</code></td><td>
+ The output iterator \c o</td></tr>
+<tr><td>Combinatorial or <br>"scrabble" search</td><td><code>a.combinatorial_search(k, o, i)</code></td><td>
+ The output iterator \c o</td></tr>
+</table>
+
+*/
+
+/** \page tst_impl Implementation Details
+ * (In the following, "original" and "DDJ" code refers to the article by Bentley/Sedgewick
+ * published in Dr Dobb's Journal, and the accompanying C source code - see \ref tst_links)
+ *
+ * In most implementations, a ternary tree node has the following members: \code
+ * struct node {
+ * char splitchar; // key letter, or 0 at end of key
+ * node *hikid; // subtree of keys with higher value than splitchar
+ * node *eqkid; // subtree matching splitchar (pointer to mapped value at end-of-key node)
+ * node *lokid; // subtree less than splitchar
+ * node *parent; // necessary for iteration (not needed for insert/find)
+ * }; \endcode
+ *
+ * This means that each node is 1 char plus three or four pointers size.
+ * On many systems, struct member alignment makes the char member consume size of one pointer
+ * as well, so we have 4 (or 5) x sizeof(pointer) per node in the tree.
+ * With several kinds of dictionaries, the node count ends up at around 0.3-0.5 times
+ * total key length (since keys share nodes).
+ * This is even more expensive on 64-bit machines.
+ *
+ * There are several variation points in the node class:
+ * -# the DDJ C code designates an invalid value of zero to indicate end-of-string. We want to
+ * allow any string as key, so the end-of-string representation should change.
+ * We note that on many platforms, C/C++ struct member alignment leaves a "hole"
+ * in the binary representation of the node, between the char and the first pointer ("hikid").
+ * On such systems there is no space cost to use another char-sized value to indicate end node.
+ * This also works for \c wchar_t strings on 32- or 64-bit systems.
+ * -# The original code stores a value for each string in the terminal node's "equal" pointer.
+ * The value in DDJ code is always a pointer to the terminated string. This is used to make
+ * advanced searches work (they return an array of pointers to strings stored in end-nodes).
+ * In reality this means that strings may need to be copied on insertion (not reflected in DDJ timings).
+ * -# Original DDJ code does not support iterating over strings in the tree.
+ * Idiomatic STL-like container style strongly suggests that iteration should be supported.
+ * This is fairly simple to implement if a parent pointer is added to the node struct:
+ * Because when an end-node is reached, the iterator must backtrack to find the previous
+ * branch point.
+ *
+ * The parent pointer also makes it possible to recover the inserted string by walking nodes
+ * backward from a terminal node to the root. Complexity is key length, plus log(tree.size),
+ * but it means inserted keys do not \b have to be copied to the end node.
+ * We opt to cache keys in iterators, at no measurable extra cost in iteration.
+
+ * Instead of the key, an arbitrary value can be associated with endnodes.
+ * However, it should not be allowed to increase node size, since most nodes in the tree are not endnodes.
+ * In this library we store the mapped value directly in end-node if it is <tt> <= \c sizeof(void*). </tt>
+ * Larger objects are allocated on the heap, and a pointer to the copy is stored in end-node
+ * (the copy is managed by the tree).
+ *
+ * <h4>Now for some optimization</h4>
+ * We use a \c vector<node> as pool allocator, and record eq-hi-lo links as vector index instead of pointers.
+ * The pool allocation essentially follows original C code insert2() principle.
+ * For us, it also simplifies reallocation, since pointers do not have to be rebound;
+ * the indices are always valid.
+ * This has the following consequences:
+ * - allow the option of 4-byte indices also on 64-bit systems (with obvious resulting tree size limit)
+ * - When a new key is inserted, the last part (unique to the key) is always allocated in a batch.
+ * This means that one node member, \c "eqkid", becomes redundant, as it is always the next index
+ * (except after terminal nodes of course).
+ * - in DDJ code the end-node value is stored in union with the eqkid. We note that the \c lokid node index
+ * is also unused by end-nodes (as no char should be lower than zero), so all endnode children
+ * are linked to the hi node.
+ *
+ * (In our binary-cognizant version where zero is a regular char value, this still holds,
+ * we just change the end-node test accordingly)
+ *
+ * In the final cut, our node struct data members appear roughly like this: \code
+ * struct node {
+ * CharType splitchar; // key letter, or 0 at end of key (to make sure lokid is never allowed)
+ ...
[truncated message content] |
|
From: <den...@us...> - 2009-07-12 16:27:06
|
Revision: 122
http://opentrep.svn.sourceforge.net/opentrep/?rev=122&view=rev
Author: denis_arnaud
Date: 2009-07-12 16:27:02 +0000 (Sun, 12 Jul 2009)
Log Message:
-----------
[Test] First basic working version of the travel search string.
Modified Paths:
--------------
trunk/opentrep/test/parsers/search_string_parser.cpp
Modified: trunk/opentrep/test/parsers/search_string_parser.cpp
===================================================================
--- trunk/opentrep/test/parsers/search_string_parser.cpp 2009-07-12 16:13:18 UTC (rev 121)
+++ trunk/opentrep/test/parsers/search_string_parser.cpp 2009-07-12 16:27:02 UTC (rev 122)
@@ -361,8 +361,10 @@
date =
lexeme_d[ limit_d(1u,12u)[uint2_p][assign_a(self._searchString._tmpDate._month)]
- >> '-' >> limit_d(1u,31u)[uint2_p][assign_a(self._searchString._tmpDate._day)]
- >> '-' >> limit_d(2000u,2099u)[uint4_p][assign_a(self._searchString._tmpDate._year)]]
+ >> boost::spirit::chset_p("/-")
+ >> limit_d(1u,31u)[uint2_p][assign_a(self._searchString._tmpDate._day)]
+ >> boost::spirit::chset_p("/-")
+ >> limit_d(2000u,2099u)[uint4_p][assign_a(self._searchString._tmpDate._year)]]
;
preferred_airlines =
@@ -383,11 +385,11 @@
;
passenger_number =
- lexeme_d[ limit_d(0u, 20u)[uint2_p][store_passenger_number(self._searchString)] ]
+ lexeme_d[ limit_d(1u, 9u)[uint1_p][store_passenger_number(self._searchString)] ]
;
passenger_type =
- lexeme_d[ (repeat_p(4,20)[chset_p("0-9A-Z")])[store_adult_passenger_type(self._searchString)] ]
+ lexeme_d[ (repeat_p(3,20)[chset_p("0-9a-z")])[store_adult_passenger_type(self._searchString)] ]
;
BOOST_SPIRIT_DEBUG_NODE (search_string);
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-12 16:13:23
|
Revision: 121
http://opentrep.svn.sourceforge.net/opentrep/?rev=121&view=rev
Author: denis_arnaud
Date: 2009-07-12 16:13:18 +0000 (Sun, 12 Jul 2009)
Log Message:
-----------
[Test] First compiling version of the travel search string.
Modified Paths:
--------------
trunk/opentrep/test/parsers/Makefile.am
Added Paths:
-----------
trunk/opentrep/test/parsers/search_string_parser.cpp
Property Changed:
----------------
trunk/opentrep/test/parsers/
Property changes on: trunk/opentrep/test/parsers
___________________________________________________________________
Modified: svn:ignore
- .libs
.deps
Makefile
Makefile.in
parameter_parser*
full_calculator*
schedule_parser*
levenshtein*
+ .libs
.deps
Makefile
Makefile.in
parameter_parser*
full_calculator*
schedule_parser*
search_string_parser*
levenshtein*
Modified: trunk/opentrep/test/parsers/Makefile.am
===================================================================
--- trunk/opentrep/test/parsers/Makefile.am 2009-07-12 13:36:35 UTC (rev 120)
+++ trunk/opentrep/test/parsers/Makefile.am 2009-07-12 16:13:18 UTC (rev 121)
@@ -3,7 +3,8 @@
MAINTAINERCLEANFILES = Makefile.in
-check_PROGRAMS = full_calculator parameter_parser schedule_parser levenshtein
+check_PROGRAMS = full_calculator parameter_parser schedule_parser \
+ search_string_parser levenshtein
full_calculator_SOURCES = full_calculator.cpp
full_calculator_CXXFLAGS = $(BOOST_CFLAGS)
@@ -17,6 +18,10 @@
schedule_parser_CXXFLAGS = $(BOOST_CFLAGS)
schedule_parser_LDADD = $(BOOST_LIBS) $(BOOST_DATE_TIME_LIB)
+search_string_parser_SOURCES = search_string_parser.cpp
+search_string_parser_CXXFLAGS = $(BOOST_CFLAGS)
+search_string_parser_LDADD = $(BOOST_LIBS) $(BOOST_DATE_TIME_LIB)
+
levenshtein_SOURCES = levenshtein.cpp
levenshtein_LDADD =
Added: trunk/opentrep/test/parsers/search_string_parser.cpp
===================================================================
--- trunk/opentrep/test/parsers/search_string_parser.cpp (rev 0)
+++ trunk/opentrep/test/parsers/search_string_parser.cpp 2009-07-12 16:13:18 UTC (rev 121)
@@ -0,0 +1,461 @@
+// C
+#include <assert.h>
+// STL
+#include <iostream>
+#include <sstream>
+#include <fstream>
+#include <string>
+#include <map>
+#include <set>
+#include <vector>
+// Boost (Extended STL)
+#include <boost/date_time/posix_time/posix_time.hpp>
+#include <boost/date_time/gregorian/gregorian.hpp>
+// Boost Spirit (Parsing)
+#define BOOST_SPIRIT_DEBUG
+#include <boost/spirit/core.hpp>
+#include <boost/spirit/attribute.hpp>
+#include <boost/spirit/utility/functor_parser.hpp>
+#include <boost/spirit/utility/loops.hpp>
+#include <boost/spirit/utility/chset.hpp>
+#include <boost/spirit/utility/confix.hpp>
+#include <boost/spirit/iterator/file_iterator.hpp>
+#include <boost/spirit/actor/push_back_actor.hpp>
+#include <boost/spirit/actor/assign_actor.hpp>
+
+// Type definitions
+typedef char char_t;
+typedef char const* iterator_t;
+//typedef boost::spirit::file_iterator<char_t> iterator_t;
+typedef boost::spirit::scanner<iterator_t> scanner_t;
+typedef boost::spirit::rule<scanner_t> rule_t;
+
+/** Place. */
+struct Place_T {
+ // Attributes
+ std::string _name;
+ std::string _code;
+ /** Constructor. */
+ Place_T () : _name (""), _code ("") {}
+ /* Display. */
+ void display() const {
+ std::cout << "Place: " << _name << " (" << _code << ")" << std::endl;
+ }
+};
+
+/** List of Place strucutres. */
+typedef std::vector<Place_T> PlaceList_T;
+
+/** Date. */
+struct Date_T {
+ // Attributes
+ boost::gregorian::date _date;
+ unsigned int _reldays;
+ unsigned int _day;
+ unsigned int _month;
+ unsigned int _year;
+ /** Constructor. */
+ Date_T () : _reldays (14), _day(1), _month(1), _year(1970) {}
+ /* Display. */
+ void display() const {
+ std::cout << "Date: " << _date << " (" << _day << "/" << _month
+ << "/" << _year << "), i.e. in " << _reldays << " days"
+ << std::endl;
+ }
+ /** Set the date from the staging details. */
+ boost::gregorian::date getDate() const {
+ return boost::gregorian::date (_year, _month, _day);
+ }
+};
+
+/** List of Date strucutres. */
+typedef std::vector<Date_T> DateList_T;
+
+/** Airline. */
+struct Airline_T {
+ // Attributes
+ bool _isPreferred;
+ std::string _name;
+ std::string _code;
+ /** Constructor. */
+ Airline_T () : _isPreferred (true), _name(""), _code("") {}
+ /* Display. */
+ void display() const {
+ std::cout << "Airline: " << _name << " (" << _code << ")" << std::endl;
+ }
+};
+
+/** List of Airline strucutres. */
+typedef std::vector<Airline_T> AirlineList_T;
+
+/** Passenger. */
+struct Passenger_T {
+ // Attributes
+ typedef enum { ADULT = 0, CHILD, PET, LAST_VALUE } PassengerType_T;
+ static const std::string _labels[LAST_VALUE];
+ PassengerType_T _type;
+ unsigned short _number;
+ /** Constructor. */
+ Passenger_T () : _type(ADULT), _number(1) {}
+ /* Display. */
+ void display() const {
+ std::cout << "Passenger: " << _number << " (" << _labels[_type] << ")"
+ << std::endl;
+ }
+};
+
+/** Passenger type labels. */
+const std::string Passenger_T::_labels[Passenger_T::LAST_VALUE] =
+ { "Adult", "Child", "Pet" };
+
+/** List of Passenger strucutres. */
+typedef std::vector<Passenger_T> PassengerList_T;
+
+/** Search string. */
+struct SearchString_T {
+ // Attributes
+ PlaceList_T _placeList;
+ DateList_T _dateList;
+ AirlineList_T _airlineList;
+ PassengerList_T _passengerList;
+
+ /** Constructor. */
+ SearchString_T () {}
+
+ /* Display. */
+ void display() const {
+ std::cout << std::endl;
+
+ for (PlaceList_T::const_iterator itPlace = _placeList.begin();
+ itPlace != _placeList.end(); ++itPlace) {
+ const Place_T& lPlace = *itPlace;
+ lPlace.display();
+ }
+
+ for (DateList_T::const_iterator itDate = _dateList.begin();
+ itDate != _dateList.end(); ++itDate) {
+ const Date_T& lDate = *itDate;
+ lDate.display();
+ }
+
+ for (AirlineList_T::const_iterator itAirline = _airlineList.begin();
+ itAirline != _airlineList.end(); ++itAirline) {
+ const Airline_T& lAirline = *itAirline;
+ lAirline.display();
+ }
+
+ for (PassengerList_T::const_iterator itPassenger = _passengerList.begin();
+ itPassenger != _passengerList.end(); ++itPassenger) {
+ const Passenger_T& lPassenger = *itPassenger;
+ lPassenger.display();
+ }
+
+ std::cout << "-- Staging --" << std::endl;
+ _tmpPlace.display();
+ }
+
+ // //// Staging ////
+ Place_T _tmpPlace;
+ Date_T _tmpDate;
+ Airline_T _tmpAirline;
+ Passenger_T _tmpPassenger;
+};
+
+///////////////////////////////////////////////////////////////////////////////
+//
+// Semantic actions
+//
+///////////////////////////////////////////////////////////////////////////////
+namespace {
+
+ /** Store the parsed place element. */
+ struct store_place_element {
+ store_place_element (SearchString_T& ioSearchString)
+ : _searchString (ioSearchString) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lPlace (iStr, iStrEnd);
+ // std::cout << "Place: " << lPlace << std::endl;
+
+ // Set the place
+ _searchString._tmpPlace._name += " " + lPlace;
+
+ // Add the parsed place to the list
+ // _searchString._placeList.push_back (_searchString._tmpPlace);
+ }
+
+ SearchString_T& _searchString;
+ };
+
+ /** Store a parsed date. */
+ struct store_date {
+ store_date (SearchString_T& ioSearchString)
+ : _searchString (ioSearchString) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ _searchString._tmpDate._date = _searchString._tmpDate.getDate();
+ // std::cout << "Board date: "
+ // << _searchString._date << std::endl;
+
+ // Add the parsed date to the list
+ _searchString._dateList.push_back (_searchString._tmpDate);
+ }
+
+ SearchString_T& _searchString;
+ };
+
+ /** Store the airline sign (+/-). */
+ struct store_airline_sign {
+ store_airline_sign (SearchString_T& ioSearchString)
+ : _searchString (ioSearchString) {}
+
+ void operator() (bool iAirlineSign) const {
+ _searchString._tmpAirline._isPreferred = iAirlineSign;
+ // std::cout << "Airline is preferred: " << iAirlineSign << std::endl;
+ }
+
+ SearchString_T& _searchString;
+ };
+
+ /** Store the parsed airline code. */
+ struct store_airline_code {
+ store_airline_code (SearchString_T& ioSearchString)
+ : _searchString (ioSearchString) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lAirlineCode (iStr, iStrEnd);
+ _searchString._tmpAirline._code = lAirlineCode;
+ // std::cout << "Airline code: " << lAirlineCode << std::endl;
+
+ // Add the parsed airline to the list
+ _searchString._airlineList.push_back (_searchString._tmpAirline);
+ }
+
+ SearchString_T& _searchString;
+ };
+
+ /** Store the parsed airline name. */
+ struct store_airline_name {
+ store_airline_name (SearchString_T& ioSearchString)
+ : _searchString (ioSearchString) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lAirlineName (iStr, iStrEnd);
+ _searchString._tmpAirline._name = lAirlineName;
+ // std::cout << "Airline: " << lAirlineName << std::endl;
+
+ // Add the parsed airline to the list
+ _searchString._airlineList.push_back (_searchString._tmpAirline);
+ }
+
+ SearchString_T& _searchString;
+ };
+
+ /** Store the parsed number of passengers. */
+ struct store_passenger_number {
+ store_passenger_number (SearchString_T& ioSearchString)
+ : _searchString (ioSearchString) {}
+
+ void operator() (unsigned int iNumber) const {
+ _searchString._tmpPassenger._number = iNumber;
+ // std::cout << "Number of passengers: " << iNumber << std::endl;
+ }
+
+ SearchString_T& _searchString;
+ };
+
+ /** Store the parsed passenger type. */
+ struct store_adult_passenger_type {
+ store_adult_passenger_type (SearchString_T& ioSearchString)
+ : _searchString (ioSearchString) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lPassengerType (iStr, iStrEnd);
+ _searchString._tmpPassenger._type = Passenger_T::ADULT;
+ // std::cout << "Passenger type: " << lPassengerType << std::endl;
+
+ // Add the parsed passenger to the list
+ _searchString._passengerList.push_back (_searchString._tmpPassenger);
+ }
+
+ SearchString_T& _searchString;
+ };
+
+}
+
+// /////////// Utilities /////////////
+/** 1-digit-integer parser */
+boost::spirit::int_parser<unsigned int, 10, 1, 1> int1_p;
+/** 1-digit-integer parser */
+boost::spirit::uint_parser<unsigned int, 10, 1, 1> uint1_p;
+/** 2-digit-integer parser */
+boost::spirit::uint_parser<int, 10, 2, 2> uint2_p;
+/** 4-digit-integer parser */
+boost::spirit::uint_parser<int, 10, 4, 4> uint4_p;
+/** Up-to-4-digit-integer parser */
+boost::spirit::uint_parser<int, 10, 1, 4> uint1_4_p;
+
+///////////////////////////////////////////////////////////////////////////////
+//
+// Our calculator grammar (using subrules)
+//
+///////////////////////////////////////////////////////////////////////////////
+ /**
+ Sample
+ guadeloupe rio de janeiro 07/22/2009 +aa -ua 2 adults 1 dog
+
+ Grammar:
+ search_string ::= places
+ [dates]
+ *(preferred_airlines)
+ *(passengers)
+ dates ::= board_date [off_date]
+ places ::= [board_place] off_place
+ board_place ::= place_elements
+ off_place ::= place_elements
+ place_elements ::= country | city | airport
+ country ::= country_code | country_name
+ city ::= city_code | city_name
+ airport ::= airport_code | airport_name
+ preferred_airlines ::= [+|-] airline_code | airline_name
+ passengers ::= adult_number adult_description
+ [child_number child_description]
+ [pet_number pet_description]
+ adult_description ::= 'adult' | 'adults' | 'pax' | 'passengers'
+ child_description ::= 'child' | 'children' | 'kid' | 'kids'
+ pet_description ::= 'dog' | 'dogs' | 'cat' | 'cats'
+ */
+
+using namespace boost::spirit;
+
+/** Grammar for the search string parser. */
+struct SearchStringParser :
+ public boost::spirit::grammar<SearchStringParser> {
+
+ SearchStringParser (SearchString_T& ioSearchString)
+ : _searchString (ioSearchString) {
+ }
+
+ template <typename ScannerT>
+ struct definition {
+ definition (SearchStringParser const& self) {
+
+ search_string = places
+ >> !( dates )
+ >> *( preferred_airlines )
+ >> *( passengers )
+ ;
+
+ places =
+ +( place_element )
+ ;
+
+ place_element =
+ lexeme_d[ (repeat_p(1,20)[chset_p("a-z")])[store_place_element(self._searchString)] ]
+ ;
+
+ dates =
+ date[store_date(self._searchString)]
+ >> !date[store_date(self._searchString)]
+ ;
+
+ date =
+ lexeme_d[ limit_d(1u,12u)[uint2_p][assign_a(self._searchString._tmpDate._month)]
+ >> '-' >> limit_d(1u,31u)[uint2_p][assign_a(self._searchString._tmpDate._day)]
+ >> '-' >> limit_d(2000u,2099u)[uint4_p][assign_a(self._searchString._tmpDate._year)]]
+ ;
+
+ preferred_airlines =
+ !(boost::spirit::sign_p)[store_airline_sign(self._searchString)]
+ >> airline_code | airline_name
+ ;
+
+ airline_code =
+ lexeme_d[ (repeat_p(2,3)[chset_p("0-9a-z")])[store_airline_code(self._searchString)] ]
+ ;
+
+ airline_name =
+ lexeme_d[ (repeat_p(4,20)[chset_p("0-9a-z")])[store_airline_name(self._searchString)] ]
+ ;
+
+ passengers =
+ passenger_number >> passenger_type
+ ;
+
+ passenger_number =
+ lexeme_d[ limit_d(0u, 20u)[uint2_p][store_passenger_number(self._searchString)] ]
+ ;
+
+ passenger_type =
+ lexeme_d[ (repeat_p(4,20)[chset_p("0-9A-Z")])[store_adult_passenger_type(self._searchString)] ]
+ ;
+
+ BOOST_SPIRIT_DEBUG_NODE (search_string);
+ BOOST_SPIRIT_DEBUG_NODE (places);
+ BOOST_SPIRIT_DEBUG_NODE (place_element);
+ BOOST_SPIRIT_DEBUG_NODE (dates);
+ BOOST_SPIRIT_DEBUG_NODE (date);
+ BOOST_SPIRIT_DEBUG_NODE (preferred_airlines);
+ BOOST_SPIRIT_DEBUG_NODE (airline_code);
+ BOOST_SPIRIT_DEBUG_NODE (airline_name);
+ BOOST_SPIRIT_DEBUG_NODE (passengers);
+ BOOST_SPIRIT_DEBUG_NODE (passenger_number);
+ BOOST_SPIRIT_DEBUG_NODE (passenger_type);
+ }
+
+ boost::spirit::rule<ScannerT> search_string, places, place_element,
+ dates, date, preferred_airlines, airline_code, airline_name,
+ passengers, passenger_number, passenger_type;
+
+ boost::spirit::rule<ScannerT> const& start() const { return search_string; }
+ };
+
+ SearchString_T& _searchString;
+};
+
+// /////////////// M A I N /////////////////
+int main (int argc, char* argv[]) {
+ try {
+
+ // File to be parsed
+ std::string lSearchString ("guadeloupe rio de janeiro 07/22/2009 +aa -ua 2 adults 1 dog");
+
+ // Read the command-line parameters
+ if (argc >= 1 && argv[1] != NULL) {
+ std::istringstream istr (argv[1]);
+ istr >> lSearchString;
+ }
+
+ // Read the search string
+ iterator_t lStringIterator = lSearchString.c_str();
+
+ // Instantiate the structure that will hold the result of the parsing.
+ SearchString_T lSearchStringStruct;
+ SearchStringParser lSearchStringParser (lSearchStringStruct);
+ boost::spirit::parse_info<iterator_t> info =
+ boost::spirit::parse (lStringIterator, lSearchStringParser,
+ boost::spirit::space_p);
+
+ // DEBUG
+ std::cout << "Search string:" << std::endl;
+ lSearchStringStruct.display();
+
+ std::cout << "-------------------------" << std::endl;
+ if (info.full) {
+ std::cout << "Parsing succeeded" << std::endl;
+
+ } else {
+ std::cout << "Parsing failed" << std::endl;
+ }
+ std::cout << "-------------------------" << std::endl;
+
+ } catch (const std::exception& stde) {
+ std::cerr << "Standard exception: " << stde.what() << std::endl;
+ return -1;
+
+ } catch (...) {
+ return -1;
+ }
+
+ return 0;
+}
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-12 13:36:46
|
Revision: 120
http://opentrep.svn.sourceforge.net/opentrep/?rev=120&view=rev
Author: denis_arnaud
Date: 2009-07-12 13:36:35 +0000 (Sun, 12 Jul 2009)
Log Message:
-----------
[Test] Added some tests on parsers.
Modified Paths:
--------------
trunk/opentrep/configure.ac
trunk/opentrep/test/Makefile.am
Added Paths:
-----------
trunk/opentrep/test/parsers/
trunk/opentrep/test/parsers/Makefile.am
trunk/opentrep/test/parsers/full_calculator.cpp
trunk/opentrep/test/parsers/levenshtein.cpp
trunk/opentrep/test/parsers/parameter_parser.cpp
trunk/opentrep/test/parsers/schedule_parser.cpp
trunk/opentrep/test/parsers/test_full_calculator.sh
trunk/opentrep/test/parsers/test_parameter_parser.sh
trunk/opentrep/test/parsers/test_schedule_parser.sh
trunk/opentrep/test/parsers/world_schedule.csv
Modified: trunk/opentrep/configure.ac
===================================================================
--- trunk/opentrep/configure.ac 2009-07-12 12:59:01 UTC (rev 119)
+++ trunk/opentrep/configure.ac 2009-07-12 13:36:35 UTC (rev 120)
@@ -233,6 +233,7 @@
doc/sourceforge/howto_release_opentrep.html
po/Makefile.in
test/com/Makefile
+ test/parsers/Makefile
test/Makefile
win32/Makefile)
AC_OUTPUT
Modified: trunk/opentrep/test/Makefile.am
===================================================================
--- trunk/opentrep/test/Makefile.am 2009-07-12 12:59:01 UTC (rev 119)
+++ trunk/opentrep/test/Makefile.am 2009-07-12 13:36:35 UTC (rev 120)
@@ -4,7 +4,7 @@
MAINTAINERCLEANFILES = Makefile.in
##
-SUBDIRS = com
+SUBDIRS = com parsers
##
check_PROGRAMS = IndexBuildingTestSuite
Property changes on: trunk/opentrep/test/parsers
___________________________________________________________________
Added: svn:ignore
+ .libs
.deps
Makefile
Makefile.in
parameter_parser*
full_calculator*
schedule_parser*
levenshtein*
Added: trunk/opentrep/test/parsers/Makefile.am
===================================================================
--- trunk/opentrep/test/parsers/Makefile.am (rev 0)
+++ trunk/opentrep/test/parsers/Makefile.am 2009-07-12 13:36:35 UTC (rev 120)
@@ -0,0 +1,24 @@
+## command sub-directory
+include $(top_srcdir)/Makefile.common
+
+MAINTAINERCLEANFILES = Makefile.in
+
+check_PROGRAMS = full_calculator parameter_parser schedule_parser levenshtein
+
+full_calculator_SOURCES = full_calculator.cpp
+full_calculator_CXXFLAGS = $(BOOST_CFLAGS)
+full_calculator_LDADD = $(BOOST_LIB)
+
+parameter_parser_SOURCES = parameter_parser.cpp
+parameter_parser_CXXFLAGS = $(BOOST_CFLAGS)
+parameter_parser_LDADD = $(BOOST_LIB)
+
+schedule_parser_SOURCES = schedule_parser.cpp
+schedule_parser_CXXFLAGS = $(BOOST_CFLAGS)
+schedule_parser_LDADD = $(BOOST_LIBS) $(BOOST_DATE_TIME_LIB)
+
+levenshtein_SOURCES = levenshtein.cpp
+levenshtein_LDADD =
+
+EXTRA_DIST = test_full_calculator.sh test_parameter_parser.sh \
+ test_schedule_parser.sh
Added: trunk/opentrep/test/parsers/full_calculator.cpp
===================================================================
--- trunk/opentrep/test/parsers/full_calculator.cpp (rev 0)
+++ trunk/opentrep/test/parsers/full_calculator.cpp 2009-07-12 13:36:35 UTC (rev 120)
@@ -0,0 +1,131 @@
+/*=============================================================================
+ Copyright (c) 2002-2003 Joel de Guzman
+ http://spirit.sourceforge.net/
+
+ Use, modification and distribution is subject to the Boost Software
+ License, Version 1.0. (See accompanying file LICENSE_1_0.txt or copy at
+ http://www.boost.org/LICENSE_1_0.txt)
+ =============================================================================*/
+////////////////////////////////////////////////////////////////////////////
+//
+// Full calculator example demonstrating Phoenix
+// This is discussed in the "Closures" chapter in the Spirit User's Guide.
+//
+// [ JDG 6/29/2002 ]
+//
+////////////////////////////////////////////////////////////////////////////
+#include <boost/spirit/core.hpp>
+#include <boost/spirit/attribute.hpp>
+#include <iostream>
+#include <string>
+
+////////////////////////////////////////////////////////////////////////////
+using namespace std;
+using namespace boost::spirit;
+using namespace phoenix;
+
+////////////////////////////////////////////////////////////////////////////
+//
+// Our calculator grammar using phoenix to do the semantics
+//
+// Note: The top rule propagates the expression result (value) upwards
+// to the calculator grammar self.val closure member which is
+// then visible outside the grammar (i.e. since self.val is the
+// member1 of the closure, it becomes the attribute passed by
+// the calculator to an attached semantic action. See the
+// driver code that uses the calculator below).
+//
+////////////////////////////////////////////////////////////////////////////
+struct calc_closure : boost::spirit::closure<calc_closure, double>
+{
+ member1 val;
+};
+
+struct calculator : public grammar<calculator, calc_closure::context_t>
+{
+ template <typename ScannerT>
+ struct definition
+ {
+ definition(calculator const& self)
+ {
+ top = expression[self.val = arg1];
+
+ expression
+ = term[expression.val = arg1]
+ >> *( ('+' >> term[expression.val += arg1])
+ | ('+' >> term[expression.val -= arg1])
+ | ('-' >> term[expression.val -= arg1])
+ )
+ ;
+
+ term
+ = factor[term.val = arg1]
+ >> *( ('*' >> factor[term.val *= arg1])
+ | ('/' >> factor[term.val /= arg1])
+ )
+ ;
+
+ factor
+ = ureal_p[factor.val = arg1]
+ | '(' >> expression[factor.val = arg1] >> ')'
+ | ('-' >> factor[factor.val = -arg1])
+ | ('+' >> factor[factor.val = arg1])
+ ;
+ }
+
+ typedef rule<ScannerT, calc_closure::context_t> rule_t;
+ rule_t expression, term, factor;
+ rule<ScannerT> top;
+
+ rule<ScannerT> const&
+ start() const { return top; }
+ };
+};
+
+////////////////////////////////////////////////////////////////////////////
+//
+// Main program
+//
+////////////////////////////////////////////////////////////////////////////
+int
+main()
+{
+ cout << "/////////////////////////////////////////////////////////\n\n";
+ cout << "\t\tExpression parser using Phoenix...\n\n";
+ cout << "/////////////////////////////////////////////////////////\n\n";
+ cout << "Type an expression...or [q or Q] to quit\n\n";
+
+ calculator calc; // Our parser
+
+ string str;
+ while (getline(cin, str))
+ {
+ if (str.empty() || str[0] == 'q' || str[0] == 'Q')
+ break;
+
+ double n = 0;
+ parse_info<> info = parse(str.c_str(), calc[var(n) = arg1], space_p);
+
+ // calc[var(n) = arg1] invokes the calculator and extracts
+ // the result of the computation. See calculator grammar
+ // note above.
+
+ if (info.full)
+ {
+ cout << "-------------------------\n";
+ cout << "Parsing succeeded\n";
+ cout << "result = " << n << endl;
+ cout << "-------------------------\n";
+ }
+ else
+ {
+ cout << "-------------------------\n";
+ cout << "Parsing failed\n";
+ cout << "stopped at: \": " << info.stop << "\"\n";
+ cout << "-------------------------\n";
+ }
+ }
+
+ cout << "Bye... :-) \n\n";
+ return 0;
+}
Added: trunk/opentrep/test/parsers/levenshtein.cpp
===================================================================
--- trunk/opentrep/test/parsers/levenshtein.cpp (rev 0)
+++ trunk/opentrep/test/parsers/levenshtein.cpp 2009-07-12 13:36:35 UTC (rev 120)
@@ -0,0 +1,131 @@
+// Levenshtein Distance Algorithm: C++ Implementation by Anders Sewerin Johansen
+// STL
+#include <iostream>
+#include <string>
+#include <vector>
+
+// //////////////////////////////////////////////////////////////////
+int getLevenshteinDistance (const std::string& source,
+ const std::string& target) {
+
+ // Step 1
+
+ const int n = source.length();
+ const int m = target.length();
+ if (n == 0) {
+ return m;
+ }
+ if (m == 0) {
+ return n;
+ }
+
+ // Definition of Matrix Type
+ typedef std::vector< std::vector<int> > Matrix_T;
+
+ Matrix_T matrix (n+1);
+
+ // Size the vectors in the 2.nd dimension. Unfortunately C++ doesn't
+ // allow for allocation on declaration of 2.nd dimension of vec of vec
+
+ for (int i = 0; i <= n; i++) {
+ matrix[i].resize(m+1);
+ }
+
+ // Step 2
+
+ for (int i = 0; i <= n; i++) {
+ matrix[i][0]=i;
+ }
+
+ for (int j = 0; j <= m; j++) {
+ matrix[0][j]=j;
+ }
+
+ // Step 3
+
+ for (int i = 1; i <= n; i++) {
+
+ const char s_i = source[i-1];
+
+ // Step 4
+
+ for (int j = 1; j <= m; j++) {
+
+ const char t_j = target[j-1];
+
+ // Step 5
+
+ int cost;
+ if (s_i == t_j) {
+ cost = 0;
+ }
+ else {
+ cost = 1;
+ }
+
+ // Step 6
+
+ const int above = matrix[i-1][j];
+ const int left = matrix[i][j-1];
+ const int diag = matrix[i-1][j-1];
+ int cell = std::min ( above + 1, std::min (left + 1, diag + cost));
+
+ // Step 6A: Cover transposition, in addition to deletion,
+ // insertion and substitution. This step is taken from:
+ // Berghel, Hal ; Roach, David : "An Extension of Ukkonen's
+ // Enhanced Dynamic Programming ASM Algorithm"
+ // (http://www.acm.org/~hlb/publications/asm/asm.html)
+
+ if (i>2 && j>2) {
+ int trans = matrix[i-2][j-2] + 1;
+ if (source[i-2] != t_j) {
+ trans++;
+ }
+ if (s_i != target[j-2]) {
+ trans++;
+ }
+ if (cell > trans) {
+ cell = trans;
+ }
+ }
+
+ matrix[i][j] = cell;
+ }
+ }
+
+ // Step 7
+
+ return matrix[n][m];
+}
+
+
+// /////////// M A I N ////////////////
+int main (int argc, char* argv[]) {
+
+ const std::string lLax1Str = "los angeles";
+ const std::string lLax2Str = "lso angeles";
+ const std::string lRio1Str = "rio de janeiro";
+ const std::string lRio2Str = "rio de janero";
+ const std::string lRek1Str = "reikjavik";
+ const std::string lRek2Str = "rekyavik";
+ const std::string lSfoRio1Str = "san francisco rio de janeiro";
+ const std::string lSfoRio2Str = "san francicso rio de janero";
+
+ std::cout << "Distance between '" << lLax1Str
+ << "' and '" << lLax2Str << "' is: "
+ << getLevenshteinDistance (lLax1Str, lLax2Str) << std::endl;
+
+ std::cout << "Distance between '" << lRio1Str
+ << "' and '" << lRio2Str << "' is: "
+ << getLevenshteinDistance (lRio1Str, lRio2Str) << std::endl;
+
+ std::cout << "Distance between '" << lRek1Str
+ << "' and '" << lRek2Str << "' is: "
+ << getLevenshteinDistance (lRek1Str, lRek2Str) << std::endl;
+
+ std::cout << "Distance between '" << lSfoRio1Str
+ << "' and '" << lSfoRio2Str << "' is: "
+ << getLevenshteinDistance (lSfoRio1Str, lSfoRio2Str) << std::endl;
+
+ return 0;
+}
Added: trunk/opentrep/test/parsers/parameter_parser.cpp
===================================================================
--- trunk/opentrep/test/parsers/parameter_parser.cpp (rev 0)
+++ trunk/opentrep/test/parsers/parameter_parser.cpp 2009-07-12 13:36:35 UTC (rev 120)
@@ -0,0 +1,216 @@
+/*=============================================================================
+ Copyright (c) 2001-2003 Hartmut Kaiser
+ http://spirit.sourceforge.net/
+
+ Use, modification and distribution is subject to the Boost Software
+ License, Version 1.0. (See accompanying file LICENSE_1_0.txt or copy at
+ http://www.boost.org/LICENSE_1_0.txt)
+ =============================================================================*/
+///////////////////////////////////////////////////////////////////////////////
+// This sample show the usage of parser parameters.
+//
+// Parser parameters are used to pass some values from the outer parsing scope
+// to the next inner scope. They can be imagined as the opposite to the return
+// value paradigm, which returns some value from the inner to the next outer
+// scope. See the "Closures" chapter in the User's Guide.
+
+#include <string>
+#include <iostream>
+#include <cassert>
+
+#if defined(_MSC_VER) /*&& !defined(__COMO__)*/
+#pragma warning(disable: 4244)
+#pragma warning(disable: 4355)
+#endif // defined(_MSC_VER) && !defined(__COMO__)
+
+#include <boost/spirit/core.hpp>
+#include <boost/spirit/symbols/symbols.hpp>
+
+#include <boost/spirit/phoenix/tuples.hpp>
+#include <boost/spirit/phoenix/tuple_helpers.hpp>
+#include <boost/spirit/phoenix/primitives.hpp>
+#include <boost/spirit/attribute/closure.hpp>
+
+///////////////////////////////////////////////////////////////////////////////
+// used namespaces
+using namespace boost::spirit;
+using namespace phoenix;
+using namespace std;
+
+///////////////////////////////////////////////////////////////////////////////
+// Helper class for encapsulation of the type for the parsed variable names
+class declaration_type
+{
+public:
+ enum vartype {
+ vartype_unknown = 0, // unknown variable type
+ vartype_int = 1, // 'int'
+ vartype_real = 2 // 'real'
+ };
+
+ declaration_type() : type(vartype_unknown)
+ {
+ }
+ template <typename ItT>
+ declaration_type(ItT const &first, ItT const &last)
+ {
+ init(string(first, last-first-1));
+ }
+ declaration_type(declaration_type const &type_) : type(type_.type)
+ {
+ }
+ declaration_type(string const &type_) : type(vartype_unknown)
+ {
+ init(type_);
+ }
+
+ // access to the variable type
+ operator vartype const &() const { return type; }
+ operator string ()
+ {
+ switch(type) {
+ default:
+ case vartype_unknown: break;
+ case vartype_int: return string("int");
+ case vartype_real: return string("real");
+ }
+ return string ("unknown");
+ }
+
+ void swap(declaration_type &s) { std::swap(type, s.type); }
+
+protected:
+ void init (string const &type_)
+ {
+ if (type_ == "int")
+ type = vartype_int;
+ else if (type_ == "real")
+ type = vartype_real;
+ else
+ type = vartype_unknown;
+ }
+
+private:
+ vartype type;
+};
+
+///////////////////////////////////////////////////////////////////////////////
+//
+// used closure type
+//
+///////////////////////////////////////////////////////////////////////////////
+struct var_decl_closure : boost::spirit::closure<var_decl_closure, declaration_type>
+{
+ member1 val;
+};
+
+///////////////////////////////////////////////////////////////////////////////
+//
+// symbols_with_data
+//
+// Helper class for inserting an item with data into a symbol table
+//
+///////////////////////////////////////////////////////////////////////////////
+template <typename T, typename InitT>
+class symbols_with_data
+{
+public:
+ typedef
+ symbol_inserter<T, boost::spirit::impl::tst<T, char> >
+ symbol_inserter_t;
+
+ symbols_with_data(symbol_inserter_t const &add_, InitT const &data_) :
+ add(add_), data(as_actor<InitT>::convert(data_))
+ {
+ }
+
+ template <typename IteratorT>
+ symbol_inserter_t const &
+ operator()(IteratorT const &first_, IteratorT const &last) const
+ {
+ IteratorT first = first_;
+ return add(first, last, data());
+ }
+
+private:
+ symbol_inserter_t const &add;
+ typename as_actor<InitT>::type data;
+};
+
+template <typename T, typename CharT, typename InitT>
+inline
+symbols_with_data<T, InitT>
+symbols_gen(symbol_inserter<T, boost::spirit::impl::tst<T, CharT> > const &add_,
+ InitT const &data_)
+{
+ return symbols_with_data<T, InitT>(add_, data_);
+}
+
+///////////////////////////////////////////////////////////////////////////////
+// The var_decl_list grammar parses variable declaration list
+
+struct var_decl_list :
+ public grammar<var_decl_list, var_decl_closure::context_t>
+{
+ template <typename ScannerT>
+ struct definition
+ {
+ definition(var_decl_list const &self)
+ {
+ // pass variable type returned from 'type' to list closure member 0
+ decl = type[self.val = arg1] >> +space_p >> list(self.val);
+
+ // m0 to access arg 0 of list --> passing variable type down to ident
+ list = ident(list.val) >> *(',' >> ident(list.val));
+
+ // store identifier and type into the symbol table
+ ident = (*alnum_p)[symbols_gen(symtab.add, ident.val)];
+
+ // the type of the decl is returned in type's closure member 0
+ type =
+ str_p("int")[type.val = construct_<string>(arg1, arg2)]
+ | str_p("real")[type.val = construct_<string>(arg1, arg2)]
+ ;
+
+ BOOST_SPIRIT_DEBUG_RULE(decl);
+ BOOST_SPIRIT_DEBUG_RULE(list);
+ BOOST_SPIRIT_DEBUG_RULE(ident);
+ BOOST_SPIRIT_DEBUG_RULE(type);
+ }
+
+ rule<ScannerT> const&
+ start() const { return decl; }
+
+ private:
+ typedef rule<ScannerT, var_decl_closure::context_t> rule_t;
+ rule_t type;
+ rule_t list;
+ rule_t ident;
+ symbols<declaration_type> symtab;
+
+ rule<ScannerT> decl; // start rule
+ };
+};
+
+///////////////////////////////////////////////////////////////////////////////
+// main entry point
+int main()
+{
+ var_decl_list decl;
+ declaration_type type;
+ char const *pbegin = "int var1";
+
+ if (parse (pbegin, decl[assign(type)]).full) {
+ cout << endl
+ << "Parsed variable declarations successfully!" << endl
+ << "Detected type: " << declaration_type::vartype(type)
+ << " (" << string(type) << ")"
+ << endl;
+ } else {
+ cout << endl
+ << "Parsing the input stream failed!"
+ << endl;
+ }
+ return 0;
+}
+
Added: trunk/opentrep/test/parsers/schedule_parser.cpp
===================================================================
--- trunk/opentrep/test/parsers/schedule_parser.cpp (rev 0)
+++ trunk/opentrep/test/parsers/schedule_parser.cpp 2009-07-12 13:36:35 UTC (rev 120)
@@ -0,0 +1,911 @@
+// C
+#include <assert.h>
+// STL
+#include <iostream>
+#include <sstream>
+#include <fstream>
+#include <string>
+#include <map>
+#include <set>
+#include <vector>
+// Boost (Extended STL)
+#include <boost/date_time/posix_time/posix_time.hpp>
+#include <boost/date_time/gregorian/gregorian.hpp>
+// Boost Spirit (Parsing)
+//#define BOOST_SPIRIT_DEBUG
+#include <boost/spirit/core.hpp>
+#include <boost/spirit/attribute.hpp>
+#include <boost/spirit/utility/functor_parser.hpp>
+#include <boost/spirit/utility/loops.hpp>
+#include <boost/spirit/utility/chset.hpp>
+#include <boost/spirit/utility/confix.hpp>
+#include <boost/spirit/iterator/file_iterator.hpp>
+#include <boost/spirit/actor/push_back_actor.hpp>
+#include <boost/spirit/actor/assign_actor.hpp>
+
+// Type definitions
+typedef char char_t;
+//typedef char const* iterator_t;
+typedef boost::spirit::file_iterator<char_t> iterator_t;
+typedef boost::spirit::scanner<iterator_t> scanner_t;
+typedef boost::spirit::rule<scanner_t> rule_t;
+
+/** LegCabin-Details. */
+struct Cabin_T {
+ // Attributes
+ std::string _cabinCode;
+ double _capacity;
+
+ void display() const {
+ std::cout << " " << _cabinCode << " " << _capacity << ", ";
+ }
+};
+
+/** List of Cabin-Detail strucutres. */
+typedef std::vector<Cabin_T> CabinList_T;
+
+/** Leg. */
+struct Leg_T {
+ // Attributes
+ std::string _boardPoint;
+ boost::posix_time::time_duration _boardTime;
+ boost::gregorian::date_duration _boardDateOffSet;
+ std::string _offPoint;
+ boost::posix_time::time_duration _offTime;
+ boost::gregorian::date_duration _offDateOffSet;
+ boost::posix_time::time_duration _elapsed;
+ CabinList_T _cabinList;
+
+ /** Constructor. */
+ Leg_T () : _boardDateOffSet (0), _offDateOffSet (0) {}
+
+ void display() const {
+ std::cout << " " << _boardPoint << " / "
+ << boost::posix_time::to_simple_string (_boardTime)
+ << " -- " << _offPoint << " / "
+ << boost::posix_time::to_simple_string (_offTime)
+ << " --> " << boost::posix_time::to_simple_string (_elapsed)
+ << std::endl;
+ for (CabinList_T::const_iterator itCabin = _cabinList.begin();
+ itCabin != _cabinList.end(); itCabin++) {
+ const Cabin_T& lCabin = *itCabin;
+ lCabin.display();
+ }
+ std::cout << std::endl;
+ }
+};
+
+/** List of Leg strucutres. */
+typedef std::vector<Leg_T> LegList_T;
+
+/** SegmentCabin-Details. */
+struct SegmentCabin_T {
+ // Attributes
+ std::string _cabinCode;
+ std::string _classes;
+
+ void display() const {
+ std::cout << " " << _cabinCode << " " << _classes << ", ";
+ }
+};
+
+/** List of SegmentCabin-Detail strucutres. */
+typedef std::vector<SegmentCabin_T> SegmentCabinList_T;
+
+/** Segment. */
+struct Segment_T {
+ // Attributes
+ std::string _boardPoint;
+ boost::posix_time::time_duration _boardTime;
+ boost::gregorian::date_duration _boardDateOffSet;
+ std::string _offPoint;
+ boost::posix_time::time_duration _offTime;
+ boost::gregorian::date_duration _offDateOffSet;
+ boost::posix_time::time_duration _elapsed;
+ SegmentCabinList_T _cabinList;
+
+ /** Constructor. */
+ Segment_T () : _boardDateOffSet (0), _offDateOffSet (0) {}
+
+ void display() const {
+ std::cout << " " << _boardPoint << " / "
+ << boost::posix_time::to_simple_string (_boardTime)
+ << " -- " << _offPoint << " / "
+ << boost::posix_time::to_simple_string (_offTime)
+ << " --> " << boost::posix_time::to_simple_string (_elapsed)
+ << std::endl;
+ for (SegmentCabinList_T::const_iterator itCabin = _cabinList.begin();
+ itCabin != _cabinList.end(); itCabin++) {
+ const SegmentCabin_T& lCabin = *itCabin;
+ lCabin.display();
+ }
+ std::cout << std::endl;
+ }
+};
+
+/** List of Segment strucutres. */
+typedef std::vector<Segment_T> SegmentList_T;
+
+/** Flight-Period. */
+struct FlightPeriod_T {
+ // Attributes
+ std::string _airlineCode;
+ unsigned int _flightNumber;
+ boost::gregorian::date _dateRangeStart;
+ boost::gregorian::date _dateRangeEnd;
+ std::string _dow;
+ LegList_T _legList;
+ SegmentList_T _segmentList;
+
+ /** Constructor. */
+ FlightPeriod_T () : _legAlreadyDefined (false), _itSeconds (0) {}
+
+ /** Set the date from the staging details. */
+ boost::gregorian::date getDate() const {
+ return boost::gregorian::date (_itYear, _itMonth, _itDay);
+ }
+
+ /** Set the time from the staging details. */
+ boost::posix_time::time_duration getTime() const {
+ return boost::posix_time::hours (_itHours)
+ + boost::posix_time::minutes (_itMinutes)
+ + boost::posix_time::seconds (_itSeconds);
+ }
+
+ void display() const {
+ std::cout << _airlineCode << _flightNumber
+ << ", " << boost::gregorian::to_simple_string (_dateRangeStart)
+ << " - " << boost::gregorian::to_simple_string (_dateRangeEnd)
+ << " - " << _dow
+ << std::endl;
+
+ for (LegList_T::const_iterator itLeg = _legList.begin();
+ itLeg != _legList.end(); itLeg++) {
+ const Leg_T& lLeg = *itLeg;
+ lLeg.display();
+ }
+
+ for (SegmentList_T::const_iterator itSegment = _segmentList.begin();
+ itSegment != _segmentList.end(); itSegment++) {
+ const Segment_T& lSegment = *itSegment;
+ lSegment.display();
+ }
+
+ //std::cout << "[Debug] - Staging Leg: ";
+ //_itLeg.display();
+ //std::cout << "[Debug] - Staging Cabin: ";
+ //_itCabin.display();
+ //std::cout << "[Debug] - Staging Segment: ";
+ //_itSegment.display();
+ }
+
+ /** Add the given airport to the internal lists (if not already existing). */
+ void addAirport (const std::string& iAirport) {
+ std::set<std::string>::const_iterator itAirport =
+ _airportList.find (iAirport);
+ if (itAirport == _airportList.end()) {
+ // Add the airport code to the airport set
+ const bool insertSuccessful = _airportList.insert (iAirport).second;
+
+ if (insertSuccessful == false) {
+ // TODO: throw an exception
+ }
+ // Add the airport code to the airport vector
+ _airportOrderedList.push_back (iAirport);
+ }
+ }
+
+ /** Build the routing (segments). */
+ void buildSegments () {
+ // The list of airports encompasses all the airports on which
+ // the flight takes off or lands. Moreover, that list is
+ // time-ordered: the first airport is the initial departure of
+ // the flight, and the last airport is the eventual point of
+ // rest of the flight.
+ // Be l the size of the ordered list of airports.
+ // We want to generate all the segment combinations from the legs
+ // and, hence, from all the possible (time-ordered) airport pairs.
+ // Thus, we both iterator on i=0...l-1 and j=i+1...l
+ assert (_airportOrderedList.size() >= 2);
+
+ _segmentList.clear();
+ for (std::vector<std::string>::const_iterator itAirport_i =
+ _airportOrderedList.begin();
+ itAirport_i != _airportOrderedList.end()-1; ++itAirport_i) {
+ for (std::vector<std::string>::const_iterator itAirport_j =
+ itAirport_i + 1;
+ itAirport_j != _airportOrderedList.end(); ++itAirport_j) {
+ Segment_T lSegment;
+ lSegment._boardPoint = *itAirport_i;
+ lSegment._offPoint = *itAirport_j;
+
+ _segmentList.push_back (lSegment);
+ }
+ }
+
+ // Clear the lists of airports, so that it is ready for the next flight
+ _airportList.clear();
+ _airportOrderedList.clear();
+ }
+
+ /** Add, to the Segment whose key corresponds to the
+ given (board point, off point) pair, the specific segment cabin
+ details (mainly, the list of the class codes).
+ <br>Note that the Segment structure is retrieved from the internal
+ list, already filled by a previous step (the buildSegments()
+ method). */
+ void addSegmentCabin (const Segment_T& iSegment,
+ const SegmentCabin_T& iCabin) {
+ // Retrieve the Segment structure corresponding to the (board, off) point
+ // pair.
+ SegmentList_T::iterator itSegment = _segmentList.begin();
+ for ( ; itSegment != _segmentList.end(); ++itSegment) {
+ const Segment_T& lSegment = *itSegment;
+
+ const std::string& lBoardPoint = iSegment._boardPoint;
+ const std::string& lOffPoint = iSegment._offPoint;
+ if (lSegment._boardPoint == lBoardPoint
+ && lSegment._offPoint == lOffPoint) {
+ break;
+ }
+ }
+
+ // If the segment key (airport pair) given in the schedule input file
+ // does not correspond to the leg (board, off) points, throw an exception
+ // so that the user knows the schedule input file is corrupted.
+ if (itSegment == _segmentList.end()) {
+ std::cerr << "Within the schedule input file, there is a flight for which the airports of segments and those of the legs do not correspond.";
+ throw std::exception();
+ }
+
+ // Add the Cabin structure to the Segment Cabin structure.
+ assert (itSegment != _segmentList.end());
+ Segment_T& lSegment = *itSegment;
+ lSegment._cabinList.push_back (iCabin);
+ }
+
+ /** Add, to all the Segment, the general segment cabin details
+ (mainly, the list of the class codes).
+ <br>Note that the Segment structures are stored within the internal
+ list, already filled by a previous step (the buildSegments()
+ method). */
+ void addSegmentCabin (const SegmentCabin_T& iCabin) {
+ // Iterate on all the Segment (as they get the same cabin definitions)
+ for (SegmentList_T::iterator itSegment = _segmentList.begin();
+ itSegment != _segmentList.end(); ++itSegment) {
+ Segment_T& lSegment = *itSegment;
+ lSegment._cabinList.push_back (iCabin);
+ }
+ }
+
+ /** Staging Leg (resp. Cabin) structure, gathering the result of the iteration
+ on one leg (resp. cabin). */
+ bool _legAlreadyDefined;
+ Leg_T _itLeg;
+ Cabin_T _itCabin;
+
+ /** Staging Date. */
+ unsigned int _itYear;
+ unsigned int _itMonth;
+ unsigned int _itDay;
+
+ /** Staging Time. */
+ long _itHours;
+ long _itMinutes;
+ long _itSeconds;
+ int _dateOffSet;
+
+ /** Staging Airport List (helper to derive the list of Segment
+ structures). */
+ std::set<std::string> _airportList;
+ std::vector<std::string> _airportOrderedList;
+
+ /** Staging Segment-related attributes. */
+ bool _areSegmentDefinitionsSpecific;
+ Segment_T _itSegment;
+ SegmentCabin_T _itSegmentCabin;
+};
+
+///////////////////////////////////////////////////////////////////////////////
+//
+// Semantic actions
+//
+///////////////////////////////////////////////////////////////////////////////
+namespace {
+
+ /** Store the parsed airline code. */
+ struct store_airline_code {
+ store_airline_code (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lAirlineCode (iStr, iStrEnd);
+ _flightPeriod._airlineCode = lAirlineCode;
+ // std::cout << "Airline code: " << lAirlineCode << std::endl;
+
+ // As that's the beginning of a new flight, the list of legs must be reset
+ _flightPeriod._legList.clear();
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the parsed flight number. */
+ struct store_flight_number {
+ store_flight_number (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ void operator() (unsigned int iNumber) const {
+ _flightPeriod._flightNumber = iNumber;
+ // std::cout << "Flight number: " << iNumber << std::endl;
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the start of the date range. */
+ struct store_date_range_start {
+ store_date_range_start (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ _flightPeriod._dateRangeStart = _flightPeriod.getDate();
+ // std::cout << "Date Range Start: "
+ // << _flightPeriod._dateRangeStart << std::endl;
+
+ // Reset the number of seconds
+ _flightPeriod._itSeconds = 0;
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the end of the date range. */
+ struct store_date_range_end {
+ store_date_range_end (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ _flightPeriod._dateRangeEnd = _flightPeriod.getDate();
+ // std::cout << "Date Range End: "
+ // << _flightPeriod._dateRangeEnd << std::endl;
+
+ // Reset the number of seconds
+ _flightPeriod._itSeconds = 0;
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the DOW (day of the Week). */
+ struct store_dow {
+ store_dow (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lDow (iStr, iStrEnd);
+ _flightPeriod._dow = lDow;
+ // std::cout << "DOW: " << lDow << std::endl;
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the parsed board point. */
+ struct store_board_point {
+ store_board_point (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lBoardPoint (iStr, iStrEnd);
+ // std::cout << "Board point: " << lBoardPoint << std::endl;
+
+ // If a leg has already been parsed, add it to the FlightPeriod
+ if (_flightPeriod._legAlreadyDefined == true) {
+ _flightPeriod._legList.push_back (_flightPeriod._itLeg);
+ } else {
+ _flightPeriod._legAlreadyDefined = true;
+ }
+
+ // Set the (new) board point
+ _flightPeriod._itLeg._boardPoint = lBoardPoint;
+
+ // As that's the beginning of a new leg, the list of cabins must be reset
+ _flightPeriod._itLeg._cabinList.clear();
+
+ // Add the airport code if it is not already stored in the airport lists
+ _flightPeriod.addAirport (lBoardPoint);
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the parsed off point. */
+ struct store_off_point {
+ store_off_point (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lOffPoint (iStr, iStrEnd);
+ _flightPeriod._itLeg._offPoint = lOffPoint;
+ // std::cout << "Off point: " << lOffPoint << std::endl;
+
+ // Add the airport code if it is not already stored in the airport lists
+ _flightPeriod.addAirport (lOffPoint);
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the board time. */
+ struct store_board_time {
+ store_board_time (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ _flightPeriod._itLeg._boardTime = _flightPeriod.getTime();
+
+ // Reset the number of seconds
+ _flightPeriod._itSeconds = 0;
+
+ // Reset the date off-set
+ _flightPeriod._dateOffSet = 0;
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the off time. */
+ struct store_off_time {
+ store_off_time (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ _flightPeriod._itLeg._offTime = _flightPeriod.getTime();
+
+ // Reset the number of seconds
+ _flightPeriod._itSeconds = 0;
+
+ // As the board date off set is optional, it can be set only afterwards,
+ // based on the staging date off-set value (_flightPeriod._dateOffSet).
+ const boost::gregorian::date_duration lDateOffSet (_flightPeriod._dateOffSet);
+ _flightPeriod._itLeg._boardDateOffSet = lDateOffSet;
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the elapsed time. */
+ struct store_elapsed_time {
+ store_elapsed_time (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ _flightPeriod._itLeg._elapsed = _flightPeriod.getTime();
+
+ // Reset the number of seconds
+ _flightPeriod._itSeconds = 0;
+
+ // As the board date off set is optional, it can be set only afterwards,
+ // based on the staging date off-set value (_flightPeriod._dateOffSet).
+ const boost::gregorian::date_duration lDateOffSet (_flightPeriod._dateOffSet);
+ _flightPeriod._itLeg._offDateOffSet = lDateOffSet;
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the parsed cabin code. */
+ struct store_cabin_code {
+ store_cabin_code (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ void operator() (char iChar) const {
+ _flightPeriod._itCabin._cabinCode = iChar;
+ // std::cout << "Cabin code: " << iChar << std::endl;
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the parsed capacity. */
+ struct store_capacity {
+ store_capacity (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ void operator() (double iReal) const {
+ _flightPeriod._itCabin._capacity = iReal;
+ // std::cout << "Capacity: " << iReal << std::endl;
+
+ // The capacity is the last (according to arrival order) detail
+ // of the cabin. Hence, when a capacity is parsed, it means that
+ // the full cabin details have already been parsed as well: the
+ // cabin can thus be added to the leg.
+ _flightPeriod._itLeg._cabinList.push_back (_flightPeriod._itCabin);
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store whether or not all the segments are the same. */
+ struct store_segment_specificity {
+ store_segment_specificity (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {
+ }
+
+ void operator() (char iChar) const {
+ if (iChar == '0') {
+ _flightPeriod._areSegmentDefinitionsSpecific = false;
+ } else {
+ _flightPeriod._areSegmentDefinitionsSpecific = true;
+ }
+
+ // Do a few sanity checks: the two lists should get exactly the same
+ // content (in terms of airport codes). The only difference is that one
+ // is a STL set, and the other a STL vector.
+ assert (_flightPeriod._airportList.size()
+ == _flightPeriod._airportOrderedList.size());
+ assert (_flightPeriod._airportList.size() >= 2);
+
+ // Since all the legs have now been parsed, we get all the airports
+ // and the segments may be built.
+ _flightPeriod.buildSegments();
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the board point of the segment. */
+ struct store_segment_board_point {
+ store_segment_board_point (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {
+ }
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lBoardPoint (iStr, iStrEnd);
+ _flightPeriod._itSegment._boardPoint = lBoardPoint;
+ // std::cout << "Board point: " << lBoardPoint << std::endl;
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the off point of the segment. */
+ struct store_segment_off_point {
+ store_segment_off_point (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {
+ }
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lOffPoint (iStr, iStrEnd);
+ _flightPeriod._itSegment._offPoint = lOffPoint;
+ // std::cout << "Off point: " << lOffPoint << std::endl;
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the off point of the segment. */
+ struct store_segment_cabin_code {
+ store_segment_cabin_code (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {
+ }
+
+ void operator() (char iChar) const {
+ _flightPeriod._itSegmentCabin._cabinCode = iChar;
+ // std::cout << "Cabin code: " << iChar << std::endl;
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Store the classes of the segment-cabin. */
+ struct store_classes {
+ store_classes (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {
+ }
+
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ std::string lClasses (iStr, iStrEnd);
+ _flightPeriod._itSegmentCabin._classes = lClasses;
+ // std::cout << "Classes: " << lClasses << std::endl;
+
+ // The list of classes is the last (according to the arrival order
+ // within the schedule input file) detail of the segment cabin. Hence,
+ // when a list of classes is parsed, it means that the full segment
+ // cabin details have already been parsed as well: the segment cabin
+ // can thus be added to the segment.
+ if (_flightPeriod._areSegmentDefinitionsSpecific == true) {
+ _flightPeriod.addSegmentCabin (_flightPeriod._itSegment,
+ _flightPeriod._itSegmentCabin);
+ } else {
+ _flightPeriod.addSegmentCabin (_flightPeriod._itSegmentCabin);
+ }
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+
+ /** Mark the end of the flight-period parsing. */
+ struct do_end_flight {
+ do_end_flight (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {}
+
+ // void operator() (char iChar) const {
+ void operator() (iterator_t iStr, iterator_t iStrEnd) const {
+ // std::cout << "End of Flight-Period " << std::endl;
+
+ assert (_flightPeriod._legAlreadyDefined == true);
+ _flightPeriod._legList.push_back (_flightPeriod._itLeg);
+
+ // The lists of legs and cabins must be reset
+ _flightPeriod._legAlreadyDefined = false;
+ _flightPeriod._itLeg._cabinList.clear();
+
+ // Display the result
+ _flightPeriod.display();
+ }
+
+ FlightPeriod_T& _flightPeriod;
+ };
+}
+
+// /////////// Utilities /////////////
+/** 1-digit-integer parser */
+boost::spirit::int_parser<unsigned int, 10, 1, 1> int1_p;
+/** 2-digit-integer parser */
+boost::spirit::uint_parser<int, 10, 2, 2> uint2_p;
+/** 4-digit-integer parser */
+boost::spirit::uint_parser<int, 10, 4, 4> uint4_p;
+/** Up-to-4-digit-integer parser */
+boost::spirit::uint_parser<int, 10, 1, 4> uint1_4_p;
+
+///////////////////////////////////////////////////////////////////////////////
+//
+// Our calculator grammar (using subrules)
+//
+///////////////////////////////////////////////////////////////////////////////
+ /**
+ AirlineCode; FlightNumber; DateRangeStart; DateRangeEnd; DOW;
+ (list) BoardPoint; OffPoint; BoardTime; DateOffSet; OffTime;
+ ElapsedTime;
+ (list) CabinCode; Capacity;
+ SegmentSpecificty (0 or 1);
+ (list) (optional BoardPoint; OffPoint); CabinCode; Classes
+
+ BA; 9; 2007-04-20; 2007-06-30; 0000011;
+ LHR; BKK; 22:00; 15:15 / +1; 11:15; F; 5; J; 12; W; 20; Y; 300;
+ BKK; SYD; 18:10 / +1; 06:05 / +2; 08:55; F; 5; J; 12; W; 20; Y; 300;
+ 0; F; FA; J; JCDI; W; WT; Y; YBHKMLSQ;
+ BA; 9; 2007-04-20; 2007-06-30; 1111100;
+ LHR; BKK; 22:00; 15:15 / +1; 11:15; F; 5; J; 12; W; 20; Y; 300;
+ BKK; SYD; 18:10 / +1; 06:05 / +2; 08:55; F; 5; J; 12; W; 20; Y; 300;
+ 1; LHR; BKK; F; FA; J; JCDI; W; WT; Y; YBHKMLSQ;
+ BKK; SYD; F; FA; J; JCDI; W; WT; Y; YBHKMLSQ;
+ LHR; SYD; F; FA; J; JCDI; W; WT; Y; YBHKMLSQ;
+
+ Grammar:
+ DOW ::= int
+ FlightKey ::= AirlineCode ';' FlightNumber
+ ';' DateRangeStart ';' DateRangeEnd ';' DOW
+ LegKey ::= BoardPoint ';' OffPoint
+ LegDetails ::= BoardTime ['/' BoardDateOffSet]
+ ';' OffTime ['/' BoardDateOffSet]
+ ';' Elapsed
+ LegCabinDetails ::= CabinCode ';' Capacity
+ Leg ::= LegKey ';' LegDetails (';' CabinDetails)+
+ SegmentKey ::= BoardPoint ';' OffPoint
+ SegmentCabinDetails ::= CabinCode ';' Classes
+ FullSegmentCabinDetails::= (';' SegmentCabinDetails)+
+ GeneralSegments ::= '0' (';' SegmentCabinDetails)+
+ SpecificSegments ::= '1' (';' SegmentKey
+ ';' FullSegmentCabinDetails)+
+ Segment ::= GeneralSegment | SpecificSegment
+ FlightPeriod ::= FlightKey (';' Leg)+
+ (';' Segment)+ ';' EndOfFlight
+ EndOfFlight ::= ';'
+ */
+
+using namespace boost::spirit;
+
+/** Grammar for the Flight-Period parser. */
+struct FlightPeriodParser :
+ public boost::spirit::grammar<FlightPeriodParser> {
+
+ FlightPeriodParser (FlightPeriod_T& ioFlightPeriod)
+ : _flightPeriod (ioFlightPeriod) {
+ }
+
+ template <typename ScannerT>
+ struct definition {
+ definition (FlightPeriodParser const& self) {
+
+ flight_period_list = *( boost::spirit::comment_p("//")
+ | boost::spirit::comment_p("/*", "*/")
+ | flight_period )
+ ;
+
+ flight_period = flight_key
+ >> +( ';' >> leg )
+ >> +( ';' >> segment )
+ >> flight_period_end[do_end_flight(self._flightPeriod)]
+ ;
+
+ flight_period_end =
+ boost::spirit::ch_p(';')
+ ;
+
+ flight_key = airline_code
+ >> ';' >> flight_number
+ >> ';' >> date[store_date_range_start(self._flightPeriod)]
+ >> ';' >> date[store_date_range_end(self._flightPeriod)]
+ >> ';' >> dow[store_dow(self._flightPeriod)]
+ ;
+
+ airline_code =
+ lexeme_d[ (repeat_p(2,3)[chset_p("0-9A-Z")])[store_airline_code(self._flightPeriod)] ]
+ ;
+
+ flight_number =
+ lexeme_d[ limit_d(0u, 9999u)[uint1_4_p][store_flight_number(self._flightPeriod)] ]
+ ;
+
+ date =
+ lexeme_d[ limit_d(2000u,2099u)[uint4_p][assign_a(self._flightPeriod._itYear)]
+ >> '-' >> limit_d(1u,12u)[uint2_p][assign_a(self._flightPeriod._itMonth)]
+ >> '-' >> limit_d(1u,31u)[uint2_p][assign_a(self._flightPeriod._itDay)] ]
+ ;
+
+ dow =
+ lexeme_d[ repeat_p(7)[chset_p("0-1")] ]
+ ;
+
+ leg = leg_key >> ';' >> leg_details >> +( ';' >> cabin_details )
+ ;
+
+ leg_key =
+ (repeat_p(3)[chset_p("0-9A-Z")])[store_board_point(self._flightPeriod)]
+ >> ';'
+ >> (repeat_p(3)[chset_p("0-9A-Z")])[store_off_point(self._flightPeriod)]
+ ;
+
+ leg_details =
+ time[store_board_time(self._flightPeriod)]
+ >> !(date_offset)
+ >> ';'
+ >> time[store_off_time(self._flightPeriod)]
+ >> !(date_offset)
+ >> ';'
+ >> time[store_elapsed_time(self._flightPeriod)]
+ ;
+
+ time = lexeme_d[ limit_d(0u,23u)[uint2_p][assign_a(self._flightPeriod._itHours)]
+ >> ':' >> limit_d(0u,59u)[uint2_p][assign_a(self._flightPeriod._itMinutes)]
+ >> !(':' >> limit_d(0u,59u)[uint2_p][assign_a(self._flightPeriod._itSeconds)]) ]
+ ;
+
+ date_offset =
+ boost::spirit::ch_p('/')
+ >> (int1_p)[boost::spirit::assign_a(self._flightPeriod._dateOffSet)]
+ ;
+
+ cabin_details = (chset_p("A-Z"))[store_cabin_code(self._flightPeriod)]
+ >> ';' >> (boost::spirit::ureal_p)[store_capacity(self._flightPeriod)]
+ ;
+
+ segment_key =
+ (repeat_p(3)[chset_p("0-9A-Z")])[store_segment_board_point(self._flightPeriod)]
+ >> ';'
+ >> (repeat_p(3)[chset_p("0-9A-Z")])[store_segment_off_point(self._flightPeriod)]
+ ;
+
+ segment =
+ general_segments | specific_segments
+ ;
+
+ general_segments =
+ boost::spirit::ch_p('0')[store_segment_specificity(self._flightPeriod)]
+ >> +(';' >> segment_cabin_details)
+ ;
+
+ specific_segments =
+ boost::spirit::ch_p('1')[store_segment_specificity(self._flightPeriod)]
+ >> +(';' >> segment_key >> full_segment_cabin_details)
+ ;
+
+ full_segment_cabin_details =
+ +(';' >> segment_cabin_details)
+ ;
+
+ segment_cabin_details =
+ (chset_p("A-Z"))[store_segment_cabin_code(self._flightPeriod)]
+ >> ';' >> (repeat_p(1,26)[chset_p("A-Z")])[store_classes(self._flightPeriod)]
+ ;
+
+ BOOST_SPIRIT_DEBUG_NODE (flight_period_list);
+ BOOST_SPIRIT_DEBUG_NODE (flight_period);
+ BOOST_SPIRIT_DEBUG_NODE (flight_period_end);
+ BOOST_SPIRIT_DEBUG_NODE (flight_key);
+ BOOST_SPIRIT_DEBUG_NODE (airline_code);
+ BOOST_SPIRIT_DEBUG_NODE (flight_number);
+ BOOST_SPIRIT_DEBUG_NODE (date);
+ BOOST_SPIRIT_DEBUG_NODE (dow);
+ BOOST_SPIRIT_DEBUG_NODE (leg);
+ BOOST_SPIRIT_DEBUG_NODE (leg_key);
+ BOOST_SPIRIT_DEBUG_NODE (leg_details);
+ BOOST_SPIRIT_DEBUG_NODE (time);
+ BOOST_SPIRIT_DEBUG_NODE (date_offset);
+ BOOST_SPIRIT_DEBUG_NODE (cabin_details);
+ BOOST_SPIRIT_DEBUG_NODE (segment);
+ BOOST_SPIRIT_DEBUG_NODE (segment_key);
+ BOOST_SPIRIT_DEBUG_NODE (general_segments);
+ BOOST_SPIRIT_DEBUG_NODE (specific_segments);
+ BOOST_SPIRIT_DEBUG_NODE (full_segment_cabin_details);
+ BOOST_SPIRIT_DEBUG_NODE (segment_cabin_details);
+ }
+
+ boost::spirit::rule<ScannerT> flight_period_list, flight_period,
+ flight_period_end, flight_key, airline_code, flight_number,
+ date, dow, leg, leg_key, leg_details, time, date_offset, cabin_details,
+ segment, segment_key, general_segments, specific_segments,
+ full_segment_cabin_details, segment_cabin_details;
+
+ boost::spirit::rule<ScannerT> const& start() const { return flight_period_list; }
+ };
+
+ FlightPeriod_T& _flightPeriod;
+};
+
+// /////////////// M A I N /////////////////
+int main (int argc, char* argv[]) {
+ try {
+
+ // File to be parsed
+ std::string lFilename ("world_schedule.csv");
+
+ // Read the command-line parameters
+ if (argc >= 1 && argv[1] != NULL) {
+ std::istringstream istr (argv[1]);
+ istr >> lFilename;
+ }
+
+ // Open the file
+ iterator_t lFileIterator (lFilename);
+ if (!lFileIterator) {
+ std::cerr << "The file " << lFilename << " can not be open." << std::endl;
+ }
+
+ // Create an EOF iterator
+ iterator_t lFileIteratorEnd = lFileIterator.make_end();
+
+ // Instantiate the structure that will hold the result of the parsing.
+ FlightPeriod_T lFlightPeriod;
+ FlightPeriodParser lFlightPeriodParser (lFlightPeriod);
+ boost::spirit::parse_info<iterator_t> info =
+ boost::spirit::parse (lFileIterator, lFileIteratorEnd,
+ lFlightPeriodParser,
+ boost::spirit::space_p);
+
+ // DEBUG
+ std::cout << "Flight Period:" << std::endl;
+ lFlightPeriod.display();
+
+ std::cout << "-------------------------" << std::endl;
+ if (info.full) {
+ std::cout << "Parsing succeeded" << std::endl;
+
+ } else {
+ std::cout << "Parsing failed" << std::endl;
+ }
+ std::cout << "-------------------------" << std::endl;
+
+ } catch (const std::exception& stde) {
+ std::cerr << "Standard exception: " << stde.what() << std::endl;
+ return -1;
+
+ } catch (...) {
+ return -1;
+ }
+
+ return 0;
+}
Added: trunk/opentrep/test/parsers/test_full_calculator.sh
===================================================================
--- trunk/opentrep/test/parsers/test_full_calculator.sh (rev 0)
+++ trunk/opentrep/test/parsers/test_full_calculator.sh 2009-07-12 13:36:35 UTC (rev 120)
@@ -0,0 +1,25 @@
+#!/bin/sh
+
+INSTALL_DIR=`grep "^prefix =" ../Makefile | cut -d"=" -d" " -f3`
+TST_PROG=./full_calculator
+LATUS_API_VERSION=`grep "^LATUS_API_VERSION =" ../Makefile | cut -d"=" -d" " -f3`
+LATUS_LIBRARY_NAME=`grep "^LATUS_LIBRARY_NAME =" ../Makefile | cut -d"=" -d" " -f3`
+LATUS_LIB=lib${LATUS_LIBRARY_NAME}-${LATUS_API_VERSION}.so
+
+if [ ! -x ${TST_PROG} ];
+then
+ echo "The sample program does not seem to have been compiled. Try 'make check' first."
+ exit -1
+fi
+
+if [ "$1" = "-h" -o "$1" = "-H" -o "$1" = "--h" -o "$1" = "--help" ];
+then
+ echo "Usage: $0 [<String to be parsed>]"
+ echo " The list to be parsed should contain floating point numbers"
+ echo " separated by commas, and should not contain spaces."
+ echo " Example: 10.2,5.4"
+ echo "The program parses a line and fills a flight-period structure."
+ exit 0
+fi
+
+${TST_PROG} $1
Property changes on: trunk/opentrep/test/parsers/test_full_calculator.sh
___________________________________________________________________
Added: svn:executable
+ *
Added: trunk/opentrep/test/parsers/test_parameter_parser.sh
===================================================================
--- trunk/opentrep/test/parsers/test_parameter_parser.sh (rev 0)
+++ trunk/opentrep/test/parsers/test_parameter_parser.sh 2009-07-12 13:36:35 UTC (rev 120)
@@ -0,0 +1,25 @@
+#!/bin/sh
+
+INSTALL_DIR=`grep "^prefix =" ../Makefile | cut -d"=" -d" " -f3`
+TST_PROG=./parameter_parser
+LATUS_API_VERSION=`grep "^LATUS_API_VERSION =" ../Makefile | cut -d"=" -d" " -f3`
+LATUS_LIBRARY_NAME=`grep "^LATUS_LIBRARY_NAME =" ../Makefile | cut -d"=" -d" " -f3`
+LATUS_LIB=lib${LATUS_LIBRARY_NAME}-${LATUS_API_VERSION}.so
+
+if [ ! -x ${TST_PROG} ];
+then
+ echo "The sample program does not seem to have been compiled. Try 'make check' first."
+ exit -1
+fi
+
+if [ "$1" = "-h" -o "$1" = "-H" -o "$1" = "--h" -o "$1" = "--help" ];
+then
+ echo "Usage: $0 [<String to be parsed>]"
+ echo " The list to be parsed should contain floating point numbers"
+ echo " separated by commas, and should not contain spaces."
+ echo " Example: 10.2,5.4"
+ echo "The program parses a line and fills a flight-period structure."
+ exit 0
+fi
+
+${TST_PROG} $1
Property changes on: trunk/opentrep/test/parsers/test_parameter_parser.sh
___________________________________________________________________
Added: svn:executable
+ *
Added: trunk/opentrep/test/parsers/test_schedule_parser.sh
===================================================================
--- trunk/opentrep/test/parsers/test_schedule_parser.sh (rev 0)
+++ trunk/opentrep/test/parsers/test_schedule_parser.sh 2009-07-12 13:36:35 UTC (rev 120)
@@ -0,0 +1,25 @@
+#!/bin/sh
+
+INSTALL_DIR=`grep "^prefix =" ../Makefile | cut -d"=" -d" " -f3`
+TST_PROG=./schedule_parser
+LATUS_API_VERSION=`grep "^LATUS_API_VERSION =" ../Makefile | cut -d"=" -d" " -f3`
+LATUS_LIBRARY_NAME=`grep "^LATUS_LIBRARY_NAME =" ../Makefile | cut -d"=" -d" " -f3`
+LATUS_LIB=lib${LATUS_LIBRARY_NAME}-${LATUS_API_VERSION}.so
+
+if [ ! -x ${TST_PROG} ];
+then
+ echo "The sample program does not seem to have been compiled. Try 'make check' first."
+ exit -1
+fi
+
+if [ "$1" = "-h" -o "$1" = "-H" -o "$1" = "--h" -o "$1" = "--help" ];
+then
+ echo "Usage: $0 [<String to be parsed>]"
+ echo " The list to be parsed should contain floating point numbers"
+ echo " separated by commas, and should not contain spaces."
+ echo " Example: 10.2,5.4"
+ echo "The program parses a line and fills a flight-period structure."
+ exit 0
+fi
+
+${TST_PROG} $1
Property changes on: trunk/opentrep/test/parsers/test_schedule_parser.sh
___________________________________________________________________
Added: svn:executable
+ *
Added: trunk/opentrep/test/parsers/world_schedule.csv
===================================================================
--- trunk/opentrep/test/parsers/world_schedule.csv (rev 0)
+++ trunk/opentrep/test/parsers/world_schedule.csv 2009-07-12 13:36:35 UTC (rev 120)
@@ -0,0 +1,21 @@
+// Flights: AirlineCode; FlightNumber; Date-Range; ; DOW; Legs; Segments;
+// Legs: BoardPoint; OffPoint; BoardTime; ArrivalDateOffSet; ArrivalTime;
+// ElapsedTime; LegCabins;
+// LegCabins: CabinCode; Capacity;
+// Segments: Specific;
+BA; 9; 2007-04-20; 2007-06-30; 0000011; LHR; BKK; 22:00; 15:15 / +1; 11:15; F; 5; J; 12; W; 20; Y; 300; BKK; SYD; 18:10 / +1; 06:05 / +2; 08:55; F; 5; J; 12; W; 20; Y; 300; 0; F; FA; J; JCDI; W; WT; Y; YBHKMLSQ;
+BA; 9; 2007-04-20; 2007-06-30; 1111100; LHR; BKK; 22:00; 15:15 / +1; 11:15; F; 5; J; 12; W; 20; Y; 300; BKK; SYD; 18:10 / +1; 06:05 / +2; 08:55; F; 5; J; 12; W; 20; Y; 300; 1; LHR; BKK; F; FA; J; JCDI; W; WT; Y; YBHKMLSQ; BKK; SYD; F; FA; J; JCDI; W; WT; Y; YBHKMLSQ; LHR; SYD; F; FA; J; JCDI; W; WT; Y; YBHKMLSQ;
+BA; 117; 2007-04-20; 2007-06-30; 1111111; LHR; JFK; 08:20; 11:00; 07:40; F; 5; J; 12; W; 20; Y; 300; 0; F; FA; J; JCDI; W; WT; Y; YBHKM;
+BA; 175; 2007-04-20; 2007-06-30; 1111111; LHR; JFK; 10:55; 13:35; 07:40; F; 5; J; 12; W; 20; Y; 300; 0; F; FA; J; JCDI; W; WT; Y; YBHKMRL;
+BA; 179; 2007-04-20; 2007-06-30; 1111111; LHR; JFK; 18:05; 20:45; 07:40; F; 5; J; 12; W; 20; Y; 300; 0; F; FA; J; JCDI; W; WT; Y; YBHKMRVNELSQO;
+BA; 207; 2007-04-20; 2007-06-30; 1111111; LHR; MIA; 09:40; 14:25; 09:45; F; 5; J; 12; W; 20; Y; 300; 0; F; FA; J; JCDI; W; WT; Y; YBHKMRVNELSQO;
+BA; 279; 2007-04-20; 2007-06-30; 1111111; LHR; LAX; 10:05; 13:10; 11:05; F; 5; J; 12; W; 20; Y; 300; 0; F; FA; J; JCDI; W; WT; Y; YBHKMRVNELSQO;
+BA; 295; 2007-04-20; 2007-06-30; 1111111; LHR; ORD; 11:35; 14:00; 08:25; F; 5; J; 12; W; 20; Y; 300; 0; F; FA; J; JCDI; W; WT; Y; YBHKMRVNELSQO;
+BA; 341; 2007-04-20; 2007-06-30; 1111111; NCE; LHR; 08:55; 10:05; 02:10; J; 12; Y; 300; 0; J; JCDI; Y; YBHKMRVNEQLSO;
+BA; 343; 2007-04-20; 2007-06-30; 1111111; NCE; LHR; 11:00; 12:15; 02:15; J; 12; Y; 300; 0; J; JCDI; Y; YBHKMRVNEQLSO;
+BA; 345; 2007-04-20; 2007-06-30; 1111111; NCE; LHR; 16:20; 17:25; 02:05; J; 12; Y; 300; 0; J; JCDI; Y; YBHKMRVNEQLSO;
+BA; 347; 2007-04-20; 2007-06-30; 1111111; NCE; LHR; 13:55; 15:00; 02:05; J; 12; Y; 300; 0; J; JCDI; Y; YBHKMRVNEQLSO;
+AA; 101; 2007-04-20; 2007-06-30; 1111111; LHR; JFK; 09:55; 12:50; 07:55; G; 300; 0; G; GHQKLMVSOWN;
+AA; 117; 2007-04-20; 2007-06-30; 1111111; JFK; LAX; 14:20; 17:25; 06:05; F; 12; J; 20; Y; 300; 0; F; FA; J; JDI; Y; YBGHQKLMVSOWN;
+AA; 181; 2007-04-20; 2007-06-30; 1111111; JFK; LAX; 17:00; 20:00; 06:00; F; 12; J; 20; Y; 300; 0; F; FA; J; JDI; Y; YBHKMLWVGSNOQ;
+AA; 585; 2007-04-20; 2007-06-30; 1111111; JFK; MIA; 15:40; 18:50; 03:10; F; 12; Y; 300; 0; F; FAP; Y; YBHKMLWVGSONQ;
\ No newline at end of file
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <den...@us...> - 2009-07-12 12:59:06
|
Revision: 119
http://opentrep.svn.sourceforge.net/opentrep/?rev=119&view=rev
Author: denis_arnaud
Date: 2009-07-12 12:59:01 +0000 (Sun, 12 Jul 2009)
Log Message:
-----------
[Structure] Re-organised a little bit the hierarchy of directories, so as to add core, config and batches.
Modified Paths:
--------------
trunk/opentrep/config/soci.m4
trunk/opentrep/configure.ac
trunk/opentrep/opentrep/Makefile.am
trunk/opentrep/opentrep/dbadaptor/Makefile.am
trunk/opentrep/opentrep/sources.mk
trunk/opentrep/po/POTFILES.in
trunk/opentrep/test/testIndexer.sh
trunk/opentrep/test/testSearcher.sh
Added Paths:
-----------
trunk/opentrep/opentrep/batches/
trunk/opentrep/opentrep/batches/Makefile.am
trunk/opentrep/opentrep/batches/indexer.cpp
trunk/opentrep/opentrep/batches/searcher.cpp
trunk/opentrep/opentrep/batches/sources.mk
trunk/opentrep/opentrep/config/
trunk/opentrep/opentrep/config/Makefile.am
trunk/opentrep/opentrep/core/
trunk/opentrep/opentrep/core/Makefile.am
trunk/opentrep/opentrep/core/sources.mk
Removed Paths:
-------------
trunk/opentrep/opentrep/indexer.cpp
trunk/opentrep/opentrep/searcher.cpp
Property Changed:
----------------
trunk/opentrep/config/
trunk/opentrep/opentrep/
Property changes on: trunk/opentrep/config
___________________________________________________________________
Modified: svn:ignore
- install-sh
missing
depcomp
mdate-sh
texinfo.tex
ltmain.sh
lib-ld.m4
lib-link.m4
lib-prefix.m4
config.rpath
config.sub
config.guess
mkinstalldirs
printf-posix.m4
uintmax_t.m4
signed.m4
iconv.m4
longlong.m4
inttypes.m4
glibc21.m4
codeset.m4
inttypes_h.m4
longdouble.m4
nls.m4
po.m4
intmax.m4
xsize.m4
lcmessage.m4
wint_t.m4
ulonglong.m4
progtest.m4
inttypes-pri.m4
stdint_h.m4
intdiv0.m4
isc-posix.m4
size_max.m4
gettext.m4
wchar_t.m4
+ install-sh
missing
depcomp
mdate-sh
texinfo.tex
ltmain.sh
libtool.m4
lt~obsolete.m4
ltsugar.m4
ltversion.m4
ltoptions.m4
lib-ld.m4
lib-link.m4
lib-prefix.m4
config.rpath
config.sub
config.guess
mkinstalldirs
printf-posix.m4
uintmax_t.m4
signed.m4
iconv.m4
longlong.m4
inttypes.m4
glibc21.m4
codeset.m4
inttypes_h.m4
longdouble.m4
nls.m4
po.m4
intmax.m4
xsize.m4
lcmessage.m4
wint_t.m4
ulonglong.m4
progtest.m4
inttypes-pri.m4
stdint_h.m4
intdiv0.m4
isc-posix.m4
size_max.m4
gettext.m4
wchar_t.m4
Modified: trunk/opentrep/config/soci.m4
===================================================================
--- trunk/opentrep/config/soci.m4 2009-05-30 16:41:07 UTC (rev 118)
+++ trunk/opentrep/config/soci.m4 2009-07-12 12:59:01 UTC (rev 119)
@@ -4,115 +4,137 @@
dnl
dnl We define the following configure script flags:
dnl
-dnl --with-soci: Give prefix for both library and headers, and try
-dnl to guess subdirectory names for each. (e.g. add /lib and
-dnl /include onto given dir names, and other common schemes.)
+dnl --with-soci: Give prefix for both library and headers, and try
+dnl to guess subdirectory names for each. (e.g. Tack /lib and
+dnl /include onto given dir name, and other common schemes.)
+dnl --with-soci-lib: Similar to --with-soci, but for library only.
+dnl --with-soci-include: Similar to --with-soci, but for headers
+dnl only.
dnl
-dnl @version 1.3, 2009/05/02
-dnl @author Denis Arnaud <den...@us...>
-dnl
dnl @version 1.2, 2007/02/20
dnl @author Warren Young <so...@et...>
AC_DEFUN([AX_SOCI],
[
-#
-# Set up configure script macros
-#
-AC_ARG_WITH(soci,
- [--with-soci=<path> root directory path of Soci installation],
- [SOCI_lib_check="$with_soci/lib64/soci $with_soci/lib/soci $with_soci/lib64 $with_soci/lib"
- SOCI_inc_check="$with_soci/include $with_soci/include/soci"],
- [SOCI_lib_check="/usr/lib64 /usr/lib /usr/lib64/soci /usr/lib/soci /usr/local/lib64 /usr/local/lib /opt/soci/lib64 /opt/soci/lib /usr/local/lib64/soci /usr/local/lib/soci /usr/local/soci/lib64 /usr/local/soci/lib /opt/soci/lib64/soci /opt/soci/lib/soci"
- SOCI_inc_check="/usr/include /usr/include/soci /usr/local/include /opt/soci/include /usr/local/include/soci /usr/local/soci/include /usr/local/soci/include/soci /opt/soci/include/soci"])
+ #
+ # Set up configure script macros
+ #
+ AC_ARG_WITH(soci,
+ [ --with-soci=<path> root directory path of Soci installation],
+ [SOCI_lib_check="$with_soci/lib64/soci $with_soci/lib/soci $with_soci/lib64 $with_soci/lib"
+ SOCI_inc_check="$with_soci/include $with_soci/include/soci"],
+ [SOCI_lib_check="/usr/lib64 /usr/lib /usr/lib64/soci /usr/lib/soci /usr/local/lib64 /usr/local/lib /usr/local/lib/soci /usr/local/soci/lib /usr/local/soci/lib/soci /opt/soci/lib /opt/soci/lib/soci"
+ SOCI_inc_check="/usr/include /usr/include/soci /usr/local/include/soci /usr/local/soci/include /usr/local/soci/include/soci /opt/soci/include /opt/soci/include/soci"])
-# SOCI library
-SOCI_CORE_LIB=soci_core
-SOCI_MYSQL_LIB=soci_mysql
+ AC_ARG_WITH(soci-lib,
+ [ --with-soci-lib=<path> directory path of Soci library installation],
+ [SOCI_lib_check="$with_soci_lib $with_soci_lib/lib64 $with_soci_lib/lib $with_soci_lib/lib64/soci $with_soci_lib/lib/soci"])
-#
-# Look for Soci Core API library
-#
-AC_MSG_CHECKING([for Soci library directory])
-SOCI_libdir=
-for m in $SOCI_lib_check
-do
- if test -d "$m" && \
- (test -f "$m/lib$SOCI_CORE_LIB.so" || test -f "$m/lib$SOCI_CORE_LIB.a")
+ AC_ARG_WITH(soci-include,
+ [ --with-soci-include=<path> directory path of Soci header installation],
+ [SOCI_inc_check="$with_soci_include $with_soci_include/include $with_soci_include/include/soci"])
+
+ # SOCI library
+ SOCI_CORE_LIB=soci_core
+ SOCI_MYSQL_LIB=soci_mysql
+ SOCI_LIB_SUFFIX=gcc-3_0
+
+ #
+ # Look for Soci Core API library
+ #
+ AC_MSG_CHECKING([for Soci library directory])
+ SOCI_libdir=
+ for m in $SOCI_lib_check
+ do
+ if test -d "$m"
+ then
+ for socilib in "$SOCI_CORE_LIB $SOCI_CORE_LIB-${SOCI_LIB_SUFFIX}"
+ do
+ if (test -f "$m/lib$SOCI_CORE_LIB.so" || test -f "$m/lib$SOCI_CORE_LIB.a")
+ then
+ SOCI_libdir=$m
+ fi
+ if (test -f "$m/lib${SOCI_CORE_LIB}-${SOCI_LIB_SUFFIX}.so" \
+ || test -f "$m/lib${SOCI_CORE_LIB}-${SOCI_LIB_SUFFIX}.a")
+ then
+ SOCI_CORE_LIB=${SOCI_CORE_LIB}-${SOCI_LIB_SUFFIX}
+ SOCI_MYSQL_LIB=${SOCI_MYSQL_LIB}-${SOCI_LIB_SUFFIX}
+ SOCI_libdir=$m
+ fi
+ done
+ break
+ fi
+ done
+
+ if test -z "$SOCI_libdir"
then
- SOCI_libdir=$m
- break
+ AC_MSG_ERROR([Didn't find $SOCI_CORE_LIB library in '$SOCI_lib_check'])
fi
-done
-if test -z "$SOCI_libdir"
-then
- AC_MSG_ERROR([Didn't find $SOCI_CORE_LIB library in '$SOCI_lib_check'])
-fi
+ case "$SOCI_libdir" in
+ /* ) ;;
+ * ) AC_MSG_ERROR([The Soci library directory ($SOCI_libdir) must be an absolute path.]) ;;
+ esac
-case "$SOCI_libdir" in
- /* ) ;;
- * ) AC_MSG_ERROR([The Soci library directory ($SOCI_libdir) must be an absolute path.]) ;;
-esac
+ AC_MSG_RESULT([$SOCI_libdir])
-AC_MSG_RESULT([$SOCI_libdir])
+ case "$SOCI_libdir" in
+ /usr/lib) ;;
+ *) LDFLAGS="$LDFLAGS -L${SOCI_libdir}" ;;
+ esac
-case "$SOCI_libdir" in
- /usr/lib64) ;;
- /usr/lib) ;;
- *) SOCI_LIBS="-L${SOCI_libdir}" ;;
-esac
-LDFLAGS="$LDFLAGS ${SOCI_LIBS}"
+ #
+ # Look for Soci Core API headers
+ #
+ AC_MSG_CHECKING([for Soci include directory])
+ SOCI_incdir=
+ for m in $SOCI_inc_check
+ do
+ if test -d "$m" && (test -f "$m/soci/core/soci.h" || test -f "$m/soci/soci.h")
+ then
+ SOCI_incdir=$m
+ break
+ fi
+ done
-#
-# Look for Soci Core API headers
-#
-AC_MSG_CHECKING([for Soci include directory])
-SOCI_incdir=
-for m in $SOCI_inc_check
-do
- if test -d "$m" && test -f "$m/soci/core/soci.h"
+ if test -z "$SOCI_incdir"
then
- SOCI_incdir=$m
- break
+ AC_MSG_ERROR([Didn't find the Soci include dir in '$SOCI_inc_check'])
fi
-done
-if test -z "$SOCI_incdir"
-then
- AC_MSG_ERROR([Didn't find the Soci include dir in '$SOCI_inc_check'])
-fi
+ case "$SOCI_incdir" in
+ /* ) ;;
+ * ) AC_MSG_ERROR([The Soci include directory ($SOCI_incdir) must be an absolute path.]) ;;
+ esac
-case "$SOCI_incdir" in
- /* ) ;;
- * ) AC_MSG_ERROR([The Soci include directory ($SOCI_incdir) must be an absolute path.]) ;;
-esac
+ AC_MSG_RESULT([$SOCI_incdir])
-AC_MSG_RESULT([$SOCI_incdir])
+ if test "$SOCI_incdir" != "/usr/include"
+ then
+ SOCI_CFLAGS="-I${SOCI_incdir}"
+ fi
+ if test "$SOCI_libdir" != "/usr/lib" -a "$SOCI_libdir" != "/usr/lib64"
+ then
+ SOCI_LIBS="-L${SOCI_libdir}"
+ fi
+ SOCI_CFLAGS="-DSOCI_HEADERS_BURIED -DSOCI_MYSQL_HEADERS_BURIED $SOCI_CFLAGS"
+ SOCI_LIBS="$SOCI_LIBS -l${SOCI_CORE_LIB} -l${SOCI_MYSQL_LIB} -ldl"
+ AC_SUBST(SOCI_CFLAGS)
+ AC_SUBST(SOCI_LIBS)
-case "$SOCI_incdir" in
- /usr/include) ;;
- *) SOCI_CFLAGS="-I${SOCI_incdir}" ;;
-esac
-
-SOCI_LIBS="${SOCI_LIBS} -l${SOCI_CORE_LIB} -l${SOCI_MYSQL_LIB} -ldl"
-
-AC_SUBST(SOCI_CFLAGS)
-AC_SUBST(SOCI_LIBS)
-
# Test linking with soci (note that it needs MySQL client to have been defined
# before)
-save_LIBS="$LIBS"
-if test -z "$MYSQL_LIBS"
-then
- MYSQL_LIBS="-L/usr/lib64/mysql -L/usr/lib/mysql -lmysqlclient"
-fi
-LIBS="$LIBS $MYSQL_LIBS $SOCI_LIBS"
-AC_CHECK_LIB($SOCI_CORE_LIB, soci_begin,
- [],
- [AC_MSG_ERROR([Could not find working Soci client library!])]
- )
-LIBS="$save_LIBS"
-AC_SUBST(SOCI_CORE_LIB)
+ save_LIBS="$LIBS"
+ if test -z "$MYSQL_LIBS"
+ then
+ MYSQL_LIBS="-L/usr/lib64/mysql -L/usr/lib/mysql -lmysqlclient"
+ fi
+ LIBS="$LIBS $MYSQL_LIBS $SOCI_LIBS"
+ AC_CHECK_LIB($SOCI_CORE_LIB, soci_begin,
+ [],
+ [AC_MSG_ERROR([Could not find working Soci client library!])]
+ )
+ LIBS="$save_LIBS"
+ AC_SUBST(SOCI_CORE_LIB)
]) dnl AX_SOCI
Modified: trunk/opentrep/configure.ac
===================================================================
--- trunk/opentrep/configure.ac 2009-05-30 16:41:07 UTC (rev 118)
+++ trunk/opentrep/configure.ac 2009-07-12 12:59:01 UTC (rev 119)
@@ -218,6 +218,9 @@
opentrep/dbadaptor/Makefile
opentrep/command/Makefile
opentrep/service/Makefile
+ opentrep/config/Makefile
+ opentrep/core/Makefile
+ opentrep/batches/Makefile
man/Makefile
info/Makefile
doc/Makefile
Property changes on: trunk/opentrep/opentrep
___________________________________________________________________
Modified: svn:ignore
- .libs
.deps
stamp-h1
config.h
config.h.in
Makefile
Makefile.in
opentrep-paths.hpp
opentrep_indexer
opentrep_searcher
+ .libs
.deps
stamp-h1
config.h
config.h.in
Makefile
Makefile.in
Modified: trunk/opentrep/opentrep/Makefile.am
===================================================================
--- trunk/opentrep/opentrep/Makefile.am 2009-05-30 16:41:07 UTC (rev 118)
+++ trunk/opentrep/opentrep/Makefile.am 2009-07-12 12:59:01 UTC (rev 119)
@@ -7,7 +7,7 @@
MAINTAINERCLEANFILES = Makefile.in
-SUBDIRS = basic bom factory dbadaptor command service
+SUBDIRS = basic bom factory dbadaptor command service core config batches
EXTRA_DIST = config_msvc.h
@@ -24,33 +24,6 @@
$(top_builddir)/@PACKAGE@/service/libsvc.la
lib@PACKAGE@_la_LDFLAGS = -version-info $(GENERIC_LIBRARY_VERSION)
-# Binaries
-bin_PROGRAMS = opentrep_indexer opentrep_searcher
-
-opentrep_indexer_SOURCES = $(bin_idx_h_sources) $(bin_idx_cc_sources)
-opentrep_indexer_CXXFLAGS = $(BOOST_CFLAGS)
-opentrep_indexer_LDFLAGS = $(BOOST_PROGRAM_OPTIONS_LIB) $(SOCI_LIBS)
-opentrep_indexer_LDADD = lib@PACKAGE@.la
-
-opentrep_searcher_SOURCES = $(bin_srh_h_sources) $(bin_srh_cc_sources)
-opentrep_searcher_CXXFLAGS = $(BOOST_CFLAGS)
-opentrep_searcher_LDFLAGS = $(BOOST_PROGRAM_OPTIONS_LIB) $(SOCI_LIBS)
-opentrep_searcher_LDADD = lib@PACKAGE@.la
-
# Header files
nobase_pkginclude_HEADERS = $(service_h_sources)
nobase_nodist_pkginclude_HEADERS = $(top_builddir)/@PACKAGE@/config.h
-
-
-# Targets
-all-local: @PACKAGE@-paths.hpp
-
-@PACKAGE@-paths.hpp: Makefile
- @echo '#ifndef __OPENTREP_PATHS_HPP' > $@
- @echo '#define __OPENTREP_PATHS_HPP' >> $@
- @echo '#define PREFIXDIR "$(prefix)"' >> $@
- @echo '#define BINDIR "$(bindir)"' >> $@
- @echo '#define LIBEXECDIR "$(libexecdir)"' >> $@
- @echo '#define DATADIR "$(datadir)"' >> $@
- @echo '#define DOCDIR "$(docdir)"' >> $@
- @echo '#endif // __OPENTREP_PATHS_HPP' >> $@
Property changes on: trunk/opentrep/opentrep/batches
___________________________________________________________________
Added: svn:ignore
+ .deps
.libs
Makefile
Makefile.in
opentrep_indexer
opentrep_searcher
Added: trunk/opentrep/opentrep/batches/Makefile.am
===================================================================
--- trunk/opentrep/opentrep/batches/Makefile.am (rev 0)
+++ trunk/opentrep/opentrep/batches/Makefile.am 2009-07-12 12:59:01 UTC (rev 119)
@@ -0,0 +1,23 @@
+include $(top_srcdir)/Makefile.common
+include $(srcdir)/sources.mk
+
+## Source directory
+
+MAINTAINERCLEANFILES = Makefile.in
+
+
+# Binaries (batches)
+bin_PROGRAMS = opentrep_indexer opentrep_searcher
+
+opentrep_indexer_SOURCES = $(batches_idx_h_sources) $(batches_idx_cc_sources)
+opentrep_indexer_CXXFLAGS = $(BOOST_CFLAGS)
+#opentrep_indexer_LDADD =
+opentrep_indexer_LDFLAGS = $(BOOST_PROGRAM_OPTIONS_LIB) $(SOCI_LIBS) \
+ $(top_builddir)/@PACKAGE@/core/lib@PACKAGE@.la
+
+
+opentrep_searcher_SOURCES = $(batches_srh_h_sources) $(batches_srh_cc_sources)
+opentrep_searcher_CXXFLAGS = $(BOOST_CFLAGS)
+#opentrep_searcher_LDADD =
+opentrep_searcher_LDFLAGS = $(BOOST_PROGRAM_OPTIONS_LIB) $(SOCI_LIBS) \
+ $(top_builddir)/@PACKAGE@/core/lib@PACKAGE@.la
Copied: trunk/opentrep/opentrep/batches/indexer.cpp (from rev 114, trunk/opentrep/opentrep/indexer.cpp)
===================================================================
--- trunk/opentrep/opentrep/batches/indexer.cpp (rev 0)
+++ trunk/opentrep/opentrep/batches/indexer.cpp 2009-07-12 12:59:01 UTC (rev 119)
@@ -0,0 +1,153 @@
+// C
+#include <assert.h>
+// STL
+#include <iostream>
+#include <sstream>
+#include <fstream>
+#include <map>
+#include <vector>
+// Boost (Extended STL)
+#include <boost/date_time/posix_time/posix_time.hpp>
+#include <boost/date_time/gregorian/gregorian.hpp>
+#include <boost/program_options.hpp>
+// OPENTREP
+#include <opentrep/OPENTREP_Service.hpp>
+
+// ///////// Parsing of Options & Configuration /////////
+// A helper function to simplify the main part.
+template<class T> std::ostream& operator<< (std::ostream& os,
+ const std::vector<T>& v) {
+ std::copy (v.begin(), v.end(), std::ostream_iterator<T> (std::cout, " "));
+ return os;
+}
+
+int readConfiguration (int argc, char* argv[]) {
+ int opt;
+
+ // Declare a group of options that will be
+ // allowed only on command line
+ boost::program_options::options_description generic("Generic options");
+ generic.add_options()
+ ("version,v", "print version string")
+ ("help,h", "produce help message");
+
+ // Declare a group of options that will be allowed both on command line and in
+ // config file
+ boost::program_options::options_description config("Configuration");
+ config.add_options()
+ ("optimization",
+ boost::program_options::value<int>(&opt)->default_value(10),
+ "optimization level")
+ ("include-path,I",
+ boost::program_options::value< std::vector<std::string> >()->composing(),
+ "include path");
+
+ // Hidden options, will be allowed both on command line and
+ // in config file, but will not be shown to the user.
+ boost::program_options::options_description hidden("Hidden options");
+ hidden.add_options()
+ ("input-file",
+ boost::program_options::value< std::vector<std::string> >(),
+ "input file");
+
+ boost::program_options::options_description cmdline_options;
+ cmdline_options.add(generic).add(config).add(hidden);
+
+ boost::program_options::options_description config_file_options;
+ config_file_options.add(config).add(hidden);
+
+ boost::program_options::options_description visible("Allowed options");
+ visible.add(generic).add(config);
+
+ boost::program_options::positional_options_description p;
+ p.add("input-file", -1);
+
+ boost::program_options::variables_map vm;
+ boost::program_options::
+ store (boost::program_options::command_line_parser(argc, argv).
+ options (cmdline_options).positional(p).run(), vm);
+
+ std::ifstream ifs ("request_parser.cfg");
+ boost::program_options::store (parse_config_file (ifs, config_file_options),
+ vm);
+ boost::program_options::notify (vm);
+
+ if (vm.count ("help")) {
+ std::cout << visible << std::endl;
+ return 0;
+ }
+
+ if (vm.count ("version")) {
+ std::cout << "Open Travel Request Parser, version 1.0" << std::endl;
+ return 0;
+ }
+
+ if (vm.count ("include-path")) {
+ std::cout << "Include paths are: "
+ << vm["include-path"].as< std::vector<std::string> >()
+ << std::endl;
+ }
+
+ if (vm.count ("input-file")) {
+ std::cout << "Input files are: "
+ << vm["input-file"].as< std::vector<std::string> >()
+ << std::endl;
+ }
+
+ std::cout << "Optimization level is " << opt << std::endl;
+
+ return 0;
+}
+
+
+// /////////////// M A I N /////////////////
+int main (int argc, char* argv[]) {
+ try {
+
+ // Output log File
+ std::string lLogFilename ("indexer.log");
+
+ // Xapian database name (directory of the index)
+ OPENTREP::TravelDatabaseName_T lXapianDatabaseName ("traveldb");
+
+ if (argc >= 1 && argv[1] != NULL) {
+ std::istringstream istr (argv[1]);
+ istr >> lLogFilename;
+ }
+
+ if (argc >= 2 && argv[2] != NULL) {
+ std::istringstream istr (argv[2]);
+ istr >> lXapianDatabaseName;
+ }
+
+ // Set the log parameters
+ std::ofstream logOutputFile;
+ // open and clean the log outputfile
+ logOutputFile.open (lLogFilename.c_str());
+ logOutputFile.clear();
+
+ // Initialise the context
+ OPENTREP::OPENTREP_Service opentrepService;
+ opentrepService.init (logOutputFile, lXapianDatabaseName);
+
+ // Launch the indexation
+ opentrepService.buildSearchIndex();
+
+ // Close the Log outputFile
+ logOutputFile.close();
+
+
+ } catch (const OPENTREP::RootException& otexp) {
+ std::cerr << "Standard exception: " << otexp.what() << std::endl;
+ return -1;
+
+ } catch (const std::exception& stde) {
+ std::cerr << "Standard exception: " << stde.what() << std::endl;
+ return -1;
+
+ } catch (...) {
+ return -1;
+ }
+
+ return 0;
+}
Copied: trunk/opentrep/opentrep/batches/searcher.cpp (from rev 114, trunk/opentrep/opentrep/searcher.cpp)
===================================================================
--- trunk/opentrep/opentrep/batches/searcher.cpp (rev 0)
+++ trunk/opentrep/opentrep/batches/searcher.cpp 2009-07-12 12:59:01 UTC (rev 119)
@@ -0,0 +1,160 @@
+// C
+#include <assert.h>
+// STL
+#include <iostream>
+#include <sstream>
+#include <fstream>
+#include <map>
+#include <vector>
+// Boost (Extended STL)
+#include <boost/date_time/posix_time/posix_time.hpp>
+#include <boost/date_time/gregorian/gregorian.hpp>
+#include <boost/program_options.hpp>
+// OPENTREP
+#include <opentrep/OPENTREP_Service.hpp>
+
+// ///////// Parsing of Options & Configuration /////////
+// A helper function to simplify the main part.
+template<class T> std::ostream& operator<< (std::ostream& os,
+ const std::vector<T>& v) {
+ std::copy (v.begin(), v.end(), std::ostream_iterator<T> (std::cout, " "));
+ return os;
+}
+
+int readConfiguration (int argc, char* argv[]) {
+ int opt;
+
+ // Declare a group of options that will be
+ // allowed only on command line
+ boost::program_options::options_description generic("Generic options");
+ generic.add_options()
+ ("version,v", "print version string")
+ ("help,h", "produce help message");
+
+ // Declare a group of options that will be allowed both on command line and in
+ // config file
+ boost::program_options::options_description config("Configuration");
+ config.add_options()
+ ("optimization",
+ boost::program_options::value<int>(&opt)->default_value(10),
+ "optimization level")
+ ("include-path,I",
+ boost::program_options::value< std::vector<std::string> >()->composing(),
+ "include path");
+
+ // Hidden options, will be allowed both on command line and
+ // in config file, but will not be shown to the user.
+ boost::program_options::options_description hidden("Hidden options");
+ hidden.add_options()
+ ("input-file",
+ boost::program_options::value< std::vector<std::string> >(),
+ "input file");
+
+ boost::program_options::options_description cmdline_options;
+ cmdline_options.add(generic).add(config).add(hidden);
+
+ boost::program_options::options_description config_file_options;
+ config_file_options.add(config).add(hidden);
+
+ boost::program_options::options_description visible("Allowed options");
+ visible.add(generic).add(config);
+
+ boost::program_options::positional_options_description p;
+ p.add("input-file", -1);
+
+ boost::program_options::variables_map vm;
+ boost::program_options::
+ store (boost::program_options::command_line_parser(argc, argv).
+ options (cmdline_options).positional(p).run(), vm);
+
+ std::ifstream ifs ("request_parser.cfg");
+ boost::program_options::store (parse_config_file (ifs, config_file_options),
+ vm);
+ boost::program_options::notify (vm);
+
+ if (vm.count ("help")) {
+ std::cout << visible << std::endl;
+ return 0;
+ }
+
+ if (vm.count ("version")) {
+ std::cout << "Open Travel Request Parser, version 1.0" << std::endl;
+ return 0;
+ }
+
+ if (vm.count ("include-path")) {
+ std::cout << "Include paths are: "
+ << vm["include-path"].as< std::vector<std::string> >()
+ << std::endl;
+ }
+
+ if (vm.count ("input-file")) {
+ std::cout << "Input files are: "
+ << vm["input-file"].as< std::vector<std::string> >()
+ << std::endl;
+ }
+
+ std::cout << "Optimization level is " << opt << std::endl;
+
+ return 0;
+}
+
+
+// /////////////// M A I N /////////////////
+int main (int argc, char* argv[]) {
+ try {
+
+ // Travel query
+ OPENTREP::TravelQuery_T lTravelQuery ("cdg");
+
+ // Output log File
+ std::string lLogFilename ("searcher.log");
+
+ // Xapian database name (directory of the index)
+ OPENTREP::TravelDatabaseName_T lXapianDatabaseName ("traveldb");
+
+ if (argc >= 1 && argv[1] != NULL) {
+ std::istringstream istr (argv[1]);
+ istr >> lTravelQuery;
+ }
+
+ if (argc >= 2 && argv[2] != NULL) {
+ std::istringstream istr (argv[2]);
+ istr >> lLogFilename;
+ }
+
+ if (argc >= 3 && argv[3] != NULL) {
+ std::istringstream istr (argv[3]);
+ istr >> lXapianDatabaseName;
+ }
+
+ // Set the log parameters
+ std::ofstream logOutputFile;
+ // open and clean the log outputfile
+ logOutputFile.open (lLogFilename.c_str());
+ logOutputFile.clear();
+
+ // Initialise the context
+ OPENTREP::OPENTREP_Service opentrepService;
+ opentrepService.init (logOutputFile, lXapianDatabaseName);
+
+ // Query the Xapian database (index)
+ opentrepService.interpretTravelRequest (lTravelQuery);
+
+ // Close the Log outputFile
+ logOutputFile.close();
+
+ } catch (const OPENTREP::RootException& otexp) {
+ std::cerr << "Standard exception: " << otexp.what() << std::endl;
+ return -1;
+
+ } catch (const std::exception& stde) {
+ std::cerr << "Standard exception: " << stde.what() << std::endl;
+ return -1;
+
+ } catch (...) {
+ return -1;
+ }
+
+ return 0;
+}
Added: trunk/opentrep/opentrep/batches/sources.mk
===================================================================
--- trunk/opentrep/opentrep/batches/sources.mk (rev 0)
+++ trunk/opentrep/opentrep/batches/sources.mk 2009-07-12 12:59:01 UTC (rev 119)
@@ -0,0 +1,4 @@
+batches_idx_h_sources =
+batches_idx_cc_sources = $(top_srcdir)/opentrep/batches/indexer.cpp
+batches_srh_h_sources =
+batches_srh_cc_sources = $(top_srcdir)/opentrep/batches/searcher.cpp
Property changes on: trunk/opentrep/opentrep/config
___________________________________________________________________
Added: svn:ignore
+ .deps
.libs
Makefile
Makefile.in
opentrep-paths.hpp
Added: trunk/opentrep/opentrep/config/Makefile.am
===================================================================
--- trunk/opentrep/opentrep/config/Makefile.am (rev 0)
+++ trunk/opentrep/opentrep/config/Makefile.am 2009-07-12 12:59:01 UTC (rev 119)
@@ -0,0 +1,25 @@
+include $(top_srcdir)/Makefile.common
+
+## Source directory
+
+DISTCLEANFILES = @PACKAGE@-paths.hpp
+
+MAINTAINERCLEANFILES = Makefile.in
+
+EXTRA_DIST = @PACKAGE@-paths.hpp
+
+# Targets
+all-local: @PACKAGE@-paths.hpp
+
+@PACKAGE@-paths.hpp: Makefile
+ @echo '#ifndef __@PACKAGE_NAME@_PATHS_HPP' > $@
+ @echo '#define __@PACKAGE_NAME@_PATHS_HPP' >> $@
+ @echo '#define PACKAGE "@PACKAGE@"' >> $@
+ @echo '#define PACKAGE_NAME "@PACKAGE_NAME@"' >> $@
+ @echo '#define PACKAGE_VERSION "@VERSION@"' >> $@
+ @echo '#define PREFIXDIR "$(prefix)"' >> $@
+ @echo '#define BINDIR "$(bindir)"' >> $@
+ @echo '#define LIBEXECDIR "$(libexecdir)"' >> $@
+ @echo '#define DATADIR "$(datadir)"' >> $@
+ @echo '#define DOCDIR "$(docdir)"' >> $@
+ @echo '#endif // __@PACKAGE_NAME@_PATHS_HPP' >> $@
Property changes on: trunk/opentrep/opentrep/core
___________________________________________________________________
Added: svn:ignore
+ .deps
.libs
Makefile
Makefile.in
Added: trunk/opentrep/opentrep/core/Makefile.am
===================================================================
--- trunk/opentrep/opentrep/core/Makefile.am (rev 0)
+++ trunk/opentrep/opentrep/core/Makefile.am 2009-07-12 12:59:01 UTC (rev 119)
@@ -0,0 +1,24 @@
+include $(top_srcdir)/Makefile.common
+include $(srcdir)/sources.mk
+
+## Source directory
+
+MAINTAINERCLEANFILES = Makefile.in
+
+SUBDIRS =
+
+
+# Library
+lib_LTLIBRARIES = lib@PACKAGE@.la
+
+lib@PACKAGE@_la_SOURCES = $(service_h_sources) $(service_cc_sources)
+lib@PACKAGE@_la_LIBADD = \
+ $(top_builddir)/@PACKAGE@/basic/libbas.la \
+ $(top_builddir)/@PACKAGE@/bom/libbom.la \
+ $(top_builddir)/@PACKAGE@/factory/libfac.la \
+ $(top_builddir)/@PACKAGE@/dbadaptor/libdba.la \
+ $(top_builddir)/@PACKAGE@/command/libcmd.la \
+ $(top_builddir)/@PACKAGE@/service/libsvc.la
+lib@PACKAGE@_la_LDFLAGS = \
+ $(BOOST_DATE_TIME_LIB) $(BOOST_PROGRAM_OPTIONS_LIB) \
+ $(SOCI_LIBS) -version-info $(GENERIC_LIBRARY_VERSION)
Added: trunk/opentrep/opentrep/core/sources.mk
===================================================================
--- trunk/opentrep/opentrep/core/sources.mk (rev 0)
+++ trunk/opentrep/opentrep/core/sources.mk 2009-07-12 12:59:01 UTC (rev 119)
@@ -0,0 +1,3 @@
+service_h_sources = $(top_srcdir)/opentrep/OPENTREP_Types.hpp \
+ $(top_srcdir)/opentrep/OPENTREP_Service.hpp
+service_cc_sources =
Modified: trunk/opentrep/opentrep/dbadaptor/Makefile.am
===================================================================
--- trunk/opentrep/opentrep/dbadaptor/Makefile.am 2009-05-30 16:41:07 UTC (rev 118)
+++ trunk/opentrep/opentrep/dbadaptor/Makefile.am 2009-07-12 12:59:01 UTC (rev 119)
@@ -3,16 +3,10 @@
include $(srcdir)/sources.mk
noinst_LTLIBRARIES= libdba.la
-if ENABLE_DEBUG
-noinst_LTLIBRARIES += libdba_debug.la
-endif
+
libdba_la_SOURCES= $(dba_h_sources) $(dba_cc_sources)
libdba_la_CXXFLAGS = $(CXXFLAGS_OPT) $(SOCI_CFLAGS)
libdba_la_LIBADD = $(SOCI_LIBS)
-libdba_debug_la_SOURCES = $(dba_h_sources) $(dba_cc_sources)
-libdba_debug_la_CXXFLAGS = $(CXXFLAGS_DEBUG) $(SOCI_CFLAGS)
-libdba_debug_la_LIBADD = $(SOCI_LIBS)
-
#pkgincludedir = $(includedir)/@PACKAGE@/dba
#pkginclude_HEADERS = $(dba_h_sources)
Deleted: trunk/opentrep/opentrep/indexer.cpp
===================================================================
--- trunk/opentrep/opentrep/indexer.cpp 2009-05-30 16:41:07 UTC (rev 118)
+++ trunk/opentrep/opentrep/indexer.cpp 2009-07-12 12:59:01 UTC (rev 119)
@@ -1,153 +0,0 @@
-// C
-#include <assert.h>
-// STL
-#include <iostream>
-#include <sstream>
-#include <fstream>
-#include <map>
-#include <vector>
-// Boost (Extended STL)
-#include <boost/date_time/posix_time/posix_time.hpp>
-#include <boost/date_time/gregorian/gregorian.hpp>
-#include <boost/program_options.hpp>
-// OPENTREP
-#include <opentrep/OPENTREP_Service.hpp>
-
-// ///////// Parsing of Options & Configuration /////////
-// A helper function to simplify the main part.
-template<class T> std::ostream& operator<< (std::ostream& os,
- const std::vector<T>& v) {
- std::copy (v.begin(), v.end(), std::ostream_iterator<T> (std::cout, " "));
- return os;
-}
-
-int readConfiguration (int argc, char* argv[]) {
- int opt;
-
- // Declare a group of options that will be
- // allowed only on command line
- boost::program_options::options_description generic("Generic options");
- generic.add_options()
- ("version,v", "print version string")
- ("help,h", "produce help message");
-
- // Declare a group of options that will be allowed both on command line and in
- // config file
- boost::program_options::options_description config("Configuration");
- config.add_options()
- ("optimization",
- boost::program_options::value<int>(&opt)->default_value(10),
- "optimization level")
- ("include-path,I",
- boost::program_options::value< std::vector<std::string> >()->composing(),
- "include path");
-
- // Hidden options, will be allowed both on command line and
- // in config file, but will not be shown to the user.
- boost::program_options::options_description hidden("Hidden options");
- hidden.add_options()
- ("input-file",
- boost::program_options::value< std::vector<std::string> >(),
- "input file");
-
- boost::program_options::options_description cmdline_options;
- cmdline_options.add(generic).add(config).add(hidden);
-
- boost::program_options::options_description config_file_options;
- config_file_options.add(config).add(hidden);
-
- boost::program_options::options_description visible("Allowed options");
- visible.add(generic).add(config);
-
- boost::program_options::positional_options_description p;
- p.add("input-file", -1);
-
- boost::program_options::variables_map vm;
- boost::program_options::
- store (boost::program_options::command_line_parser(argc, argv).
- options (cmdline_options).positional(p).run(), vm);
-
- std::ifstream ifs ("request_parser.cfg");
- boost::program_options::store (parse_config_file (ifs, config_file_options),
- vm);
- boost::program_options::notify (vm);
-
- if (vm.count ("help")) {
- std::cout << visible << std::endl;
- return 0;
- }
-
- if (vm.count ("version")) {
- std::cout << "Open Travel Request Parser, version 1.0" << std::endl;
- return 0;
- }
-
- if (vm.count ("include-path")) {
- std::cout << "Include paths are: "
- << vm["include-path"].as< std::vector<std::string> >()
- << std::endl;
- }
-
- if (vm.count ("input-file")) {
- std::cout << "Input files are: "
- << vm["input-file"].as< std::vector<std::string> >()
- << std::endl;
- }
-
- std::cout << "Optimization level is " << opt << std::endl;
-
- return 0;
-}
-
-
-// /////////////// M A I N /////////////////
-int main (int argc, char* argv[]) {
- try {
-
- // Output log File
- std::string lLogFilename ("indexer.log");
-
- // Xapian database name (directory of the index)
- OPENTREP::TravelDatabaseName_T lXapianDatabaseName ("traveldb");
-
- if (argc >= 1 && argv[1] != NULL) {
- std::istringstream istr (argv[1]);
- istr >> lLogFilename;
- }
-
- if (argc >= 2 && argv[2] != NULL) {
- std::istringstream istr (argv[2]);
- istr >> lXapianDatabaseName;
- }
-
- // Set the log parameters
- std::ofstream logOutputFile;
- // open and clean the log outputfile
- logOutputFile.open (lLogFilename.c_str());
- logOutputFile.clear();
-
- // Initialise the context
- OPENTREP::OPENTREP_Service opentrepService;
- opentrepService.init (logOutputFile, lXapianDatabaseName);
-
- // Launch the indexation
- opentrepService.buildSearchIndex();
-
- // Close the Log outputFile
- logOutputFile.close();
-
-
- } catch (const OPENTREP::RootException& otexp) {
- std::cerr << "Standard exception: " << otexp.what() << std::endl;
- return -1;
-
- } catch (const std::exception& stde) {
- std::cerr << "Standard exception: " << stde.what() << std::endl;
- return -1;
-
- } catch (...) {
- return -1;
- }
-
- return 0;
-}
Deleted: trunk/opentrep/opentrep/searcher.cpp
===================================================================
--- trunk/opentrep/opentrep/searcher.cpp 2009-05-30 16:41:07 UTC (rev 118)
+++ trunk/opentrep/opentrep/searcher.cpp 2009-07-12 12:59:01 UTC (rev 119)
@@ -1,160 +0,0 @@
-// C
-#include <assert.h>
-// STL
-#include <iostream>
-#include <sstream>
-#include <fstream>
-#include <map>
-#include <vector>
-// Boost (Extended STL)
-#include <boost/date_time/posix_time/posix_time.hpp>
-#include <boost/date_time/gregorian/gregorian.hpp>
-#include <boost/program_options.hpp>
-// OPENTREP
-#include <opentrep/OPENTREP_Service.hpp>
-
-// ///////// Parsing of Options & Configuration /////////
-// A helper function to simplify the main part.
-template<class T> std::ostream& operator<< (std::ostream& os,
- const std::vector<T>& v) {
- std::copy (v.begin(), v.end(), std::ostream_iterator<T> (std::cout, " "));
- return os;
-}
-
-int readConfiguration (int argc, char* argv[]) {
- int opt;
-
- // Declare a group of options that will be
- // allowed only on command line
- boost::program_options::options_description generic("Generic options");
- generic.add_options()
- ("version,v", "print version string")
- ("help,h", "produce help message");
-
- // Declare a group of options that will be allowed both on command line and in
- // config file
- boost::program_options::options_description config("Configuration");
- config.add_options()
- ("optimization",
- boost::program_options::value<int>(&opt)->default_value(10),
- "optimization level")
- ("include-path,I",
- boost::program_options::value< std::vector<std::string> >()->composing(),
- "include path");
-
- // Hidden options, will be allowed both on command line and
- // in config file, but will not be shown to the user.
- boost::program_options::options_description hidden("Hidden options");
- hidden.add_options()
- ("input-file",
- boost::program_options::value< std::vector<std::string> >(),
- "input file");
-
- boost::program_options::options_description cmdline_options;
- cmdline_options.add(generic).add(config).add(hidden);
-
- boost::program_options::options_description config_file_options;
- config_file_options.add(config).add(hidden);
-
- boost::program_options::options_description visible("Allowed options");
- visible.add(generic).add(config);
-
- boost::program_options::positional_options_description p;
- p.add("input-file", -1);
-
- boost::program_options::variables_map vm;
- boost::program_options::
- store (boost::program_options::command_line_parser(argc, argv).
- options (cmdline_options).positional(p).run(), vm);
-
- std::ifstream ifs ("request_parser.cfg");
- boost::program_options::store (parse_config_file (ifs, config_file_options),
- vm);
- boost::program_options::notify (vm);
-
- if (vm.count ("help")) {
- std::cout << visible << std::endl;
- return 0;
- }
-
- if (vm.count ("version")) {
- std::cout << "Open Travel Request Parser, version 1.0" << std::endl;
- return 0;
- }
-
- if (vm.count ("include-path")) {
- std::cout << "Include paths are: "
- << vm["include-path"].as< std::vector<std::string> >()
- << std::endl;
- }
-
- if (vm.count ("input-file")) {
- std::cout << "Input files are: "
- << vm["input-file"].as< std::vector<std::string> >()
- << std::endl;
- }
-
- std::cout << "Optimization level is " << opt << std::endl;
-
- return 0;
-}
-
-
-// /////////////// M A I N /////////////////
-int main (int argc, char* argv[]) {
- try {
-
- // Travel query
- OPENTREP::TravelQuery_T lTravelQuery ("cdg");
-
- // Output log File
- std::string lLogFilename ("searcher.log");
-
- // Xapian database name (directory of the index)
- OPENTREP::TravelDatabaseName_T lXapianDatabaseName ("traveldb");
-
- if (argc >= 1 && argv[1] != NULL) {
- std::istringstream istr (argv[1]);
- istr >> lTravelQuery;
- }
-
- if (argc >= 2 && argv[2] != NULL) {
- std::istringstream istr (argv[2]);
- istr >> lLogFilename;
- }
-
- if (argc >= 3 && argv[3] != NULL) {
- std::istringstream istr (argv[3]);
- istr >> lXapianDatabaseName;
- }
-
- // Set the log parameters
- std::ofstream logOutputFile;
- // open and clean the log outputfile
- logOutputFile.open (lLogFilename.c_str());
- logOutputFile.clear();
-
- // Initialise the context
- OPENTREP::OPENTREP_Service opentrepService;
- opentrepService.init (logOutputFile, lXapianDatabaseName);
-
- // Query the Xapian database (index)
- opentrepService.interpretTravelRequest (lTravelQuery);
-
- // Close the Log outputFile
- logOutputFile.close();
-
- } catch (const OPENTREP::RootException& otexp) {
- std::cerr << "Standard exception: " << otexp.what() << std::endl;
- return -1;
-
- } catch (const std::exception& stde) {
- std::cerr << "Standard exception: " << stde.what() << std::endl;
- return -1;
-
- } catch (...) {
- return -1;
- }
-
- return 0;
-}
Modified: trunk/opentrep/opentrep/sources.mk
===================================================================
--- trunk/opentrep/opentrep/sources.mk 2009-05-30 16:41:07 UTC (rev 118)
+++ trunk/opentrep/opentrep/sources.mk 2009-07-12 12:59:01 UTC (rev 119)
@@ -1,7 +1,4 @@
-service_h_sources = $(top_srcdir)/opentrep/OPENTREP_Types.hpp \
+service_h_sources = \
+ $(top_srcdir)/opentrep/OPENTREP_Types.hpp \
$(top_srcdir)/opentrep/OPENTREP_Service.hpp
service_cc_sources =
-bin_idx_h_sources =
-bin_idx_cc_sources = $(top_srcdir)/opentrep/indexer.cpp
-bin_srh_h_sources =
-bin_srh_cc_sources = $(top_srcdir)/opentrep/searcher.cpp
Modified: trunk/opentrep/po/POTFILES.in
===================================================================
--- trunk/opentrep/po/POTFILES.in 2009-05-30 16:41:07 UTC (rev 118)
+++ trunk/opentrep/po/POTFILES.in 2009-07-12 12:59:01 UTC (rev 119)
@@ -7,8 +7,8 @@
opentrep/service/Logger.hpp
opentrep/service/OPENTREP_ServiceContext.cpp
opentrep/service/OPENTREP_ServiceContext.hpp
-opentrep/indexer.cpp
-opentrep/searcher.cpp
+opentrep/batches/indexer.cpp
+opentrep/batches/searcher.cpp
opentrep/dbadaptor/DbaPlace.hpp
opentrep/dbadaptor/DbaAbstract.cpp
opentrep/dbadaptor/DbaPlace.cpp
Modified: trunk/opentrep/test/testIndexer.sh
===================================================================
--- trunk/opentrep/test/testIndexer.sh 2009-05-30 16:41:07 UTC (rev 118)
+++ trunk/opentrep/test/testIndexer.sh 2009-07-12 12:59:01 UTC (rev 119)
@@ -1,7 +1,7 @@
#!/bin/sh
INSTALL_DIR=`grep "^prefix =" ../Makefile | cut -d"=" -d" " -f3`
-TST_PROG=../opentrep/opentrep_indexer
+TST_PROG=../opentrep/batches/opentrep_indexer
OPENTREP=`grep "^PACKAGE_VERSION =" ../Makefile | cut -d"=" -d" " -f3`
OPENTREP_LIBRARY_NAME=`grep "^PACKAGE =" ../Makefile | cut -d"=" -d" " -f3`
OPENTREP_LIB=lib${OPENTREP_LIBRARY_NAME}-${OPENTREP_API_VERSION}.so
Modified: trunk/opentrep/test/testSearcher.sh
===================================================================
--- trunk/opentrep/test/testSearcher.sh 2009-05-30 16:41:07 UTC (rev 118)
+++ trunk/opentrep/test/testSearcher.sh 2009-07-12 12:59:01 UTC (rev 119)
@@ -1,7 +1,7 @@
#!/bin/sh
INSTALL_DIR=`grep "^prefix =" ../Makefile | cut -d"=" -d" " -f3`
-TST_PROG=../opentrep/opentrep_searcher
+TST_PROG=../opentrep/batches/opentrep_searcher
OPENTREP=`grep "^PACKAGE_VERSION =" ../Makefile | cut -d"=" -d" " -f3`
OPENTREP_LIBRARY_NAME=`grep "^PACKAGE =" ../Makefile | cut -d"=" -d" " -f3`
OPENTREP_LIB=lib${OPENTREP_LIBRARY_NAME}-${OPENTREP_API_VERSION}.so
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
