Menu
▾
▴
[Jboss-svn-commits] JBL Code SVN: r1910 - in trunk/labs/jbossrules/drools-natural-language/src: main/org/drools/natural main/org/drools/natural/ast main/org/drools/natural/ruledoc main/org/drools/natural/ruledoc/html test/org/drools/natural/ruledoc test/org/drools/natural/ruledoc/html
|
From: <jbo...@li...> - 2005-12-21 07:32:33
|
Author: mic...@jb...
Date: 2005-12-21 02:32:06 -0500 (Wed, 21 Dec 2005)
New Revision: 1910
Added:
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/DictionaryHelper.java
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocument.java
trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/DictionaryHelperTest.java
trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/NaturalRulesSpec.html
trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/RuleDocumentListenerTest.java
trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/RuleDocumentTest.java
trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/HTMLDocParserTest.java
trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/sample.dictionary.properties
trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/simple-ruledoc.html
Removed:
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/ParseState.java
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocumentListenerImpl.java
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleSectionState.java
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/TableState.java
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/html/HTMLDocParserImpl.java
trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/HTMLDocParserImplTest.java
trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/NaturalRulesSpec.html
trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/test_rule_raw.txt
Modified:
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/NaturalLanguageException.java
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ast/AbstractSyntaxNode.java
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/Keywords.java
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocumentListener.java
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/html/HTMLDocParser.java
trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/keywords.properties
trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/simplest.html
Log:
some more work on the HTML parser
Modified: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/NaturalLanguageException.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/NaturalLanguageException.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/NaturalLanguageException.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -13,5 +13,9 @@
{
super(message);
}
+
+ public NaturalLanguageException(String message, Throwable cause) {
+ super(message, cause);
+ }
}
Modified: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ast/AbstractSyntaxNode.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ast/AbstractSyntaxNode.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ast/AbstractSyntaxNode.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -1,5 +1,14 @@
package org.drools.natural.ast;
+/**
+ * All natural language syntax nodes extend this class.
+ *
+ * Basically we start out with a list of tokens, each token is then classified into one of the syntax types,
+ * and then they build themselvs up into a list of little ASTs.
+ *
+ * @author <a href="mailto:mic...@gm..."> Michael Neale</a>
+ *
+ */
public abstract class AbstractSyntaxNode
{
Added: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/DictionaryHelper.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/DictionaryHelper.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/DictionaryHelper.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -0,0 +1,26 @@
+package org.drools.natural.ruledoc;
+
+import java.util.Properties;
+
+public class DictionaryHelper
+{
+
+ private Properties props;
+
+ public DictionaryHelper(Properties properties) {
+ props = properties;
+ }
+
+ public String getItem(String key) {
+ return props.getProperty(key);
+ }
+
+ public String getFunctions() {
+ return props.getProperty("functions");
+ }
+
+ public String getImports() {
+ if (props.containsKey("import")) return props.getProperty("import");
+ return props.getProperty("imports");
+ }
+}
Property changes on: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/DictionaryHelper.java
___________________________________________________________________
Name: svn:eol-style
+ native
Modified: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/Keywords.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/Keywords.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/Keywords.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -9,14 +9,14 @@
public class Keywords
{
- public static Keywords instance;
+ private static Keywords instance;
private Properties props;
private Keywords(Properties p) {
this.props = p;
}
- public static Keywords getInstance() {
+ private static Keywords getInstance() {
if (instance == null) {
Properties props = new Properties();
@@ -34,6 +34,10 @@
* Helper method to get a keyword
*/
public static String getKeyword(String key) {
+ Keywords keywords = getInstance();
+ if (!keywords.props.containsKey(key)) {
+ throw new IllegalArgumentException("The keyword [" + key + "] was not in the configuration.");
+ }
return getInstance().props.getProperty(key);
}
Deleted: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/ParseState.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/ParseState.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/ParseState.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -1,12 +0,0 @@
-package org.drools.natural.ruledoc;
-
-/**
- * This is the super class of all document parse states.
- * @author <a href="mailto:mic...@gm..."> Michael Neale</a>
- *
- */
-public abstract class ParseState
-{
- abstract void parseChunk(String text);
-
-}
Added: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocument.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocument.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocument.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -0,0 +1,31 @@
+package org.drools.natural.ruledoc;
+
+import java.io.IOException;
+import java.io.InputStream;
+import java.net.URL;
+import java.util.List;
+import java.util.Properties;
+
+import org.drools.natural.ruledoc.html.HTMLDocParser;
+
+/**
+ * This is the class that does it all for rule documents (HTML based).
+ * @author <a href="mailto:mic...@gm..."> Michael Neale</a>
+ */
+public class RuleDocument
+{
+
+ public RuleDocument() {
+ }
+
+ List buildRuleListFromDocument(URL document, Properties dictionary) {
+ HTMLDocParser parser = new HTMLDocParser();
+ RuleDocumentListener listener = new RuleDocumentListener();
+ parser.parseDocument(document, listener);
+
+ return listener.getRules();
+
+ }
+
+
+}
Property changes on: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocument.java
___________________________________________________________________
Name: svn:eol-style
+ native
Modified: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocumentListener.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocumentListener.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocumentListener.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -1,27 +1,91 @@
package org.drools.natural.ruledoc;
-public interface RuleDocumentListener
+import java.util.ArrayList;
+import java.util.List;
+
+/**
+ * This is a simple rule listener implementation. Doesn't do anything clever, just extracts the rules.
+ * Leaves everything else alone.
+ * No special treatment of tables.
+ * @author <a href="mailto:mic...@gm..."> Michael Neale</a>
+ */
+public class RuleDocumentListener
{
- /**
- * Process a line of text.
- */
- public abstract void handleText(String text);
+ private boolean inComment;
+ private List rules;
+ private StringBuffer ruleBuffer;
+ private boolean inRule;
+
+ public RuleDocumentListener() {
+ rules = new ArrayList();
+ }
- public abstract void startTable();
+ public void handleText(String text)
+ {
+ if (text.trim().startsWith(Keywords.getKeyword("rule.start"))) {
+ startNewRule(text);
+ } else if (text.trim().endsWith(Keywords.getKeyword("rule.end"))) {
+ finishCurrentRule(text);
+ } else if (inComment) {
+ return;
+ } else if (inRule) {
+ ruleBuffer.append(text);
+ }
+
+ }
- public abstract void startColumn();
+ private void finishCurrentRule(String text)
+ {
+ ruleBuffer.append(text);
+ rules.add(ruleBuffer.toString());
+ inRule = false;
+ }
- public abstract void startRow();
+ private void startNewRule(String text)
+ {
+ ruleBuffer = new StringBuffer();
+ ruleBuffer.append(text);
+ inRule = true;
+ }
- public abstract void endTable();
+ public void startTable()
+ {
+ // TODO Auto-generated method stub
+
+ }
- public abstract void endColumn();
+ public void startColumn()
+ {
+ // TODO Auto-generated method stub
+
+ }
- public abstract void endRow();
+ public void startRow()
+ {
+ // TODO Auto-generated method stub
+
+ }
- public abstract void startComment();
+ public void endTable()
+ {
+ // TODO Auto-generated method stub
+
+ }
- public abstract void endComment();
+ public void startComment()
+ {
+ this.inComment = true;
+ }
-}
\ No newline at end of file
+ public void endComment()
+ {
+ this.inComment = false;
+ }
+
+ public List getRules() {
+ return rules;
+ }
+
+
+}
Deleted: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocumentListenerImpl.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocumentListenerImpl.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleDocumentListenerImpl.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -1,83 +0,0 @@
-package org.drools.natural.ruledoc;
-
-/**
- * This builds up the rules from the document events.
- * HTML is obviously supported, but this interface is not bound to HTML.
- *
- * This is the class context for the parsing state, like GoF state pattern.
- *
- * @author <a href="mailto:mic...@gm..."> Michael Neale</a>
- *
- */
-public class RuleDocumentListenerImpl implements RuleDocumentListener
-{
-
-
- /* (non-Javadoc)
- * @see org.drools.natural.ruledoc.RuleDocumentListener#handleText(java.lang.String)
- */
- public void handleText(String text) {
-
- }
-
- /* (non-Javadoc)
- * @see org.drools.natural.ruledoc.RuleDocumentListener#startTable()
- */
- public void startTable() {
-
- }
-
- /* (non-Javadoc)
- * @see org.drools.natural.ruledoc.RuleDocumentListener#startColumn()
- */
- public void startColumn() {
-
- }
-
- /* (non-Javadoc)
- * @see org.drools.natural.ruledoc.RuleDocumentListener#startRow()
- */
- public void startRow() {
-
- }
-
- /* (non-Javadoc)
- * @see org.drools.natural.ruledoc.RuleDocumentListener#endTable()
- */
- public void endTable() {
-
- }
-
- /* (non-Javadoc)
- * @see org.drools.natural.ruledoc.RuleDocumentListener#endColumn()
- */
- public void endColumn() {
-
- }
-
- /* (non-Javadoc)
- * @see org.drools.natural.ruledoc.RuleDocumentListener#endRow()
- */
- public void endRow() {
-
- }
-
- /* (non-Javadoc)
- * @see org.drools.natural.ruledoc.RuleDocumentListener#startComment()
- */
- public void startComment() {
-
- }
-
- /* (non-Javadoc)
- * @see org.drools.natural.ruledoc.RuleDocumentListener#endComment()
- */
- public void endComment() {
-
- }
-
-
-
-
-
-}
Deleted: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleSectionState.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleSectionState.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/RuleSectionState.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -1,31 +0,0 @@
-package org.drools.natural.ruledoc;
-
-import org.apache.commons.lang.StringUtils;
-
-public class RuleSectionState extends ParseState
-{
-
- private StringBuffer buf = new StringBuffer();
- private String name;
-
-
- public RuleSectionState(String name) {
- this.buf.append(name + " ");
- }
-
- void parseChunk(String text)
- {
- buf.append(text);
- }
-
- public static boolean isStart(String text)
- {
- return StringUtils.contains(text, Keywords.getKeyword("rule.start"));
- }
-
- public static boolean isEnd(String text)
- {
- return StringUtils.contains(text, Keywords.getKeyword("rule.end"));
- }
-
-}
Deleted: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/TableState.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/TableState.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/TableState.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -1,28 +0,0 @@
-package org.drools.natural.ruledoc;
-
-import java.util.Properties;
-
-public class TableState extends ParseState
-{
-
- private String name;
- private Properties data;
- private boolean inKey;
-
-
- public TableState(String name) {
- this.name = name;
- data = new Properties();
- }
-
-
-
- void parseChunk(String text)
- {
-
-
-
-
- }
-
-}
Modified: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/html/HTMLDocParser.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/html/HTMLDocParser.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/html/HTMLDocParser.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -1,30 +1,154 @@
package org.drools.natural.ruledoc.html;
import java.io.IOException;
-import java.io.InputStream;
+import java.net.URL;
+import java.net.URLConnection;
+import org.drools.natural.NaturalLanguageException;
import org.drools.natural.ruledoc.RuleDocumentListener;
+import org.htmlparser.Node;
+import org.htmlparser.Parser;
+import org.htmlparser.nodes.RemarkNode;
+import org.htmlparser.nodes.TagNode;
+import org.htmlparser.nodes.TextNode;
+import org.htmlparser.util.NodeIterator;
+import org.htmlparser.util.NodeList;
+import org.htmlparser.util.ParserException;
-/**
+/**
* The HTML document parser will treat italics as comments.
* Must be tolerant of dodgy HTML, and handle quirky HTML like the stuff
- * that Microsoft word spits out.
- *
- * Any html parsers must implement this.
- * My suggested ones are:
- * 1) javax.swing
- * 2) HotSax
- * 3) that one from SiteMesh
- * 4) roll your own.
+ * that Microsoft word spits out.
* @author <a href="mailto:mic...@gm..."> Michael Neale</a>
- *
*/
-public interface HTMLDocParser
+public class HTMLDocParser
{
- /**
- * @param input The input to the HTML.
- * @param listener The listener to raise document events against.
- * @throws IOException If there is something wrong with the stream.
- */
- void parseDocument(InputStream input, RuleDocumentListener listener) throws IOException;
+
+ private RuleDocumentListener listener;
+
+ public void parseDocument(URL url, RuleDocumentListener listener) {
+ try {
+ parseDocument(url.openConnection(), listener);
+ } catch (IOException e) {
+ throw new NaturalLanguageException("Unable to open URL to rule document", e);
+ }
+ }
+
+ public void parseDocument(URLConnection input,
+ RuleDocumentListener listener)
+ {
+ this.listener = listener;
+ try
+ {
+ Parser parser = new Parser(input);
+ for (NodeIterator i = parser.elements (); i.hasMoreNodes(); ) {
+ processNodes (i.nextNode ());
+ }
+ }
+ catch ( ParserException e )
+ {
+ throw new NaturalLanguageException("Error in the HTML parser.", e);
+ }
+ }
+
+ private void processNodes (Node node) throws ParserException
+ {
+ if (node instanceof TextNode)
+ {
+ // downcast to TextNode
+ TextNode text = (TextNode)node;
+ // do whatever processing you want with the text
+ handleText(text.getText());
+ }
+ if (node instanceof RemarkNode)
+ {
+ // downcast to RemarkNode
+ //RemarkNode remark = (RemarkNode)node;
+ }
+ else if (node instanceof TagNode)
+ {
+ // downcast to TagNode
+ TagNode tag = (TagNode)node;
+ // do whatever processing you want with the tag itself
+ handleTag(tag);
+ // process recursively (nodes within nodes) via getChildren()
+ NodeList nl = tag.getChildren ();
+ if (null != nl)
+ for (NodeIterator i = nl.elements (); i.hasMoreNodes(); ) {
+ processNodes (i.nextNode ());
+ }
+
+ }
+ }
+
+
+
+ private void handleText(String text)
+ {
+ listener.handleText(unescapeSmartQuotes(unescapeEntities(text)));
+
+ }
+
+ private void handleTag(TagNode tag)
+ {
+
+ String tagName = tag.getTagName();
+ boolean isEnding = tag.isEndTag();
+
+ if (tagName.equals("TABLE") ) {
+ if (!isEnding) {
+ listener.startTable();
+ } else {
+ listener.startTable();
+ }
+ } else if (tagName.equals("TH")) {
+ if (!isEnding) {
+ listener.startRow();
+ }
+ } else if (tagName.equalsIgnoreCase("TR")) {
+ if (!isEnding) {
+ listener.startRow();
+ }
+ } else if (tagName.equals("TD")) {
+ if (!isEnding) {
+ listener.startColumn();
+ }
+ } else if (tagName.equals("I")) {
+ if (!isEnding) {
+ listener.startComment();
+ } else {
+ listener.endComment();
+ }
+ } else if (tagName.equals("P")) {
+ handleText(" ");
+ } else if (tagName.equals("BR")) {
+ handleText(" ");
+ }
+
+
+ }
+
+
+
+ private static String unescapeSmartQuotes(String s) {
+ s = s.replace('\u201c', '"');
+ s = s.replace('\u201d', '"');
+ s = s.replace('\u2018', '\'');
+ s = s.replace('\u2019', '\'');
+ return s;
+ }
+
+ private static String unescapeEntities(String s) {
+ s = s.replaceAll("<", "<");
+ s = s.replaceAll(">", ">");
+ s = s.replaceAll(" ", " ");
+ s = s.replaceAll(""", "\"");
+ s = s.replaceAll("&", "&");
+ s = s.replaceAll("&rdquo", "\"");
+ s = s.replaceAll("“", "\"");
+ return s;
+ }
+
+
+
}
Deleted: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/html/HTMLDocParserImpl.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/html/HTMLDocParserImpl.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/html/HTMLDocParserImpl.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -1,122 +0,0 @@
-package org.drools.natural.ruledoc.html;
-
-import java.io.IOException;
-import java.io.InputStream;
-import java.io.InputStreamReader;
-import java.io.Reader;
-
-import javax.swing.text.MutableAttributeSet;
-import javax.swing.text.html.HTMLEditorKit;
-import javax.swing.text.html.HTML.Tag;
-import javax.swing.text.html.parser.ParserDelegator;
-
-
-import org.drools.natural.ruledoc.RuleDocumentListener;
-
-/**
- * This implementation of a doc parser uses the javax.swing package.
- * @author <a href="mailto:mic...@gm..."> Michael Neale</a>
- */
-public class HTMLDocParserImpl extends HTMLEditorKit.ParserCallback implements HTMLDocParser
-{
-
- private RuleDocumentListener listener;
-
- public void handleEndTag(Tag tag,
- int arg1)
- {
- System.out.println("Breaksflow: " + tag.breaksFlow());
-
- if ( tag == Tag.I )
- {
- listener.endComment();
- }
- else if ( tag == Tag.TABLE )
- {
- listener.endTable();
- }
- else if ( tag == Tag.TH )
- {
- listener.endRow();
- }
- else if ( tag == Tag.TR )
- {
- listener.endRow();
- }
- else if ( tag == Tag.TD )
- {
- listener.endColumn();
- } else if (tag == Tag.BR) {
- System.out.println("BREAK");
- } else if (tag == Tag.P) {
- System.out.println("PARA");
- }
-
- }
-
- public void handleStartTag(Tag tag,
- MutableAttributeSet arg1,
- int arg2)
- {
-
- // System.out.println("Start TAG: " + name);
- if ( tag == Tag.I )
- {
- listener.startComment();
- }
- else if ( tag == Tag.TABLE )
- {
- listener.startTable();
- }
- else if ( tag == Tag.TH )
- {
- listener.startRow();
- }
- else if ( tag == Tag.TR )
- {
- listener.startRow();
- }
- else if ( tag == Tag.TD )
- {
- listener.startColumn();
- }
-
- }
-
- public void handleText(char[] chars,
- int arg1)
- {
- String s = new String(chars);
-
- listener.handleText(unescapeSmartQuotes(unescapeEntities(s)));
- }
-
-
-
- public void parseDocument(InputStream input,
- RuleDocumentListener listener) throws IOException
- {
- this.listener = listener;
- ParserDelegator del = new ParserDelegator();
- Reader reader = new InputStreamReader(input);
- del.parse(reader, this, true);
- }
-
- private static String unescapeSmartQuotes(String s) {
- s = s.replace('\u201c', '"');
- s = s.replace('\u201d', '"');
- s = s.replace('\u2018', '\'');
- s = s.replace('\u2019', '\'');
- return s;
- }
-
- private static String unescapeEntities(String s) {
- s = s.replaceAll("<", "<");
- s = s.replaceAll(">", ">");
- s = s.replaceAll(" ", " ");
- s = s.replaceAll(""", "\"");
- s = s.replaceAll("&", "&");
- return s;
- }
-
-}
Modified: trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/keywords.properties
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/keywords.properties 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/main/org/drools/natural/ruledoc/keywords.properties 2005-12-21 07:32:06 UTC (rev 1910)
@@ -1,16 +1,9 @@
#here we configure the syntax for rules
-rule.start=Rule-start:
-rule.end=Rule-end
+rule.start=Start-rule
+rule.end=End-rule
rule.if=IF
rule.then=THEN
rule.condition.and=AND
rule.consequence.and=AND
-#and here is the config for what tables to look at. All tables have only 2 cols.
-rule.config.table.header=Rules
-dictionary.table.header=Dictionary
-ruleset.table.header=Ruleset
-#not sure if I will integrate tests or not...
-test.table.header=Test
-
Added: trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/DictionaryHelperTest.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/DictionaryHelperTest.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/DictionaryHelperTest.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -0,0 +1,17 @@
+package org.drools.natural.ruledoc;
+
+import java.io.IOException;
+import java.util.Properties;
+
+import junit.framework.TestCase;
+
+public class DictionaryHelperTest extends TestCase
+{
+ public void testLoad() throws IOException {
+
+ Properties props = new Properties();
+ props.load(getClass().getResourceAsStream("sample.dictionary.properties"));
+ DictionaryHelper dic = new DictionaryHelper(props);
+ assertEquals("convertToDate(\"${right}\")", dic.getItem("date of"));
+ }
+}
Property changes on: trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/DictionaryHelperTest.java
___________________________________________________________________
Name: svn:eol-style
+ native
Added: trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/NaturalRulesSpec.html
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/NaturalRulesSpec.html 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/NaturalRulesSpec.html 2005-12-21 07:32:06 UTC (rev 1910)
@@ -0,0 +1,315 @@
+<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
+<HTML>
+<HEAD>
+ <META HTTP-EQUIV="CONTENT-TYPE" CONTENT="text/html; charset=windows-1252">
+ <TITLE></TITLE>
+ <META NAME="GENERATOR" CONTENT="OpenOffice.org 2.0 (Win32)">
+ <META NAME="CREATED" CONTENT="20051214;11592400">
+ <META NAME="CHANGEDBY" CONTENT="Michael Neale">
+ <META NAME="CHANGED" CONTENT="20051221;17212142">
+ <STYLE>
+ <!--
+ @page { size: 21cm 29.7cm; margin: 2cm }
+ P { margin-bottom: 0.21cm }
+ H1 { margin-bottom: 0.21cm }
+ H1.western { font-family: "Arial", sans-serif; font-size: 16pt }
+ H1.cjk { font-family: "Lucida Sans Unicode"; font-size: 16pt }
+ H1.ctl { font-family: "Tahoma"; font-size: 16pt }
+ H3 { margin-bottom: 0.21cm }
+ H3.western { font-family: "Arial", sans-serif }
+ TD P { margin-bottom: 0cm }
+ TH P { margin-bottom: 0cm; font-style: italic }
+ -->
+ </STYLE>
+</HEAD>
+<BODY LANG="" DIR="LTR">
+<H1 CLASS="western"><FONT FACE="Verdana, sans-serif">Drools natural
+rule language</FONT><IMG SRC="NaturalRulesSpec_html_m48cc1843.jpg" NAME="graphics1" ALIGN=LEFT HSPACE=23 VSPACE=23 WIDTH=185 HEIGHT=246 BORDER=0><BR CLEAR=LEFT></H1>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=2 STYLE="font-size: 9pt">Natural
+rules, in textual documents specification<BR>(this is an executable
+document)</FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=2 STYLE="font-size: 9pt"><I>TODO:
+do we prefer the term “Natural rules” or “Literate
+rules”?</I></FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=2 STYLE="font-size: 9pt"><B>Outline</B></FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=2 STYLE="font-size: 9pt">Natural
+language rules work off any textual document that can be transformed
+to HTML (this is similar in concept to Ward Cunningham’s
+fit.c2.com framework) – including this one. Some simple rules
+are followed to make the natural language very easy to lexically
+analyse (which is normally the hardest bit of parsing). Combine this
+with some conventions and object introspection, and we have an
+interesting idea.</FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=2 STYLE="font-size: 9pt"><B>This
+is essentially an instance of Literate programming
+(www.literateprogramming.com),</B> which was coined by Donald Knuth
+(why isn’t he “Sir” yet, damn it if Elton John can
+get a knighthood, so can Don). A core premise of literate programming
+is that “comments” are first class, code is woven in.</FONT></FONT></FONT></P>
+<H3 CLASS="western"><FONT FACE="Verdana, sans-serif">Features</FONT></H3>
+<OL>
+ <LI><P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+ <FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Must
+ be able to expose the full power of DRL features if needed (a lot of
+ them won’t be, but generally at least one person will ask for
+ each feature). </FONT></FONT></FONT>
+ </P>
+ <LI><P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+ <FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Should
+ be able to produce readable rules (readable by a suitably trained
+ non technical person) – having a non programmer edit rules in
+ entirety is a lofty goal ;) But it may be a poison chalice ;) </FONT></FONT></FONT>
+ </P>
+ <OL>
+ <LI><P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+ <FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Can
+ use it to create rule templates, where the templates are not
+ editable, but the rule parameters are.</FONT></FONT></FONT></P>
+ </OL>
+ <LI><P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+ <FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Should
+ allow editing in normal drools snippets format for power users when
+ natural language is not appropriate.</FONT></FONT></FONT></P>
+ <LI><P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+ <FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Allow
+ rapid building of Domain Specific Languages (using the built in
+ dictionary)</FONT></FONT></FONT></P>
+ <LI><P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+ <FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">As
+ it works off HTML (or any appropriate markup language) can be driven
+ off a wiki. DHTML may be used to provide “intellisense”
+ in the future, or swing, but at the end of the day it is JUST TEXT
+ so it can easily be diff’ed (this is important).</FONT></FONT></FONT></P>
+ <LI><P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+ <FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Should
+ produce a HTML “output” file showing how the document
+ was parsed (this is hard... but worth it), and FIT test results.</FONT></FONT></FONT></P>
+ <LI><P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+ <FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">As
+ it HTML, source could be edited by an enhanced WYSIWYG DHTML editor
+ (which could restrict what is edited, provide “intellisense”
+ etc), providing a web interface for rules (an alternative to
+ document based rules).</FONT></FONT></FONT></P>
+ <LI><P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+ <FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Tests
+ can be embedded in the document by utilizing FIT (and fit-for-rules
+ from Michael), providing what if testing, and regression testing of
+ rules in isolation.</FONT></FONT></FONT></P>
+</OL>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<IMG SRC="NaturalRulesSpec_html_m48cc1843.jpg" NAME="graphics2" ALIGN=LEFT HSPACE=23 VSPACE=23 WIDTH=185 HEIGHT=246 BORDER=0><BR CLEAR=LEFT><BR><BR>
+</P>
+<H1 CLASS="western"><FONT FACE="Verdana, sans-serif">OK Here we go. </FONT>
+</H1>
+<P><BR><BR>
+</P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">The
+term “Rule-start” indicates the start of a rule section,
+which continues until “Rule-end”. This is from literate
+programming. So just about everything outside is a “comment”.
+This is the opposite of normal source.</FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt"><SPAN STYLE="background: #e6e6e6">Start-rule
+[In-the-beginning]<BR>Priority[1]<BR>Group[group1]<BR>Input[String
+message]</SPAN></FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt"><B>IF<BR> </B><SPAN STYLE="font-weight: medium">Today
+<I>is</I> before date of 10-Jul-2005 or Today is date of
+10-Jul-2006<BR></SPAN><B>THEN<BR> </B><SPAN STYLE="font-weight: medium">Log
+“I am a parameter to the log function”</SPAN></FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt"> [System.out.println(“boo”)]</FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt"><SPAN STYLE="font-weight: medium"><I><BR></I><SPAN STYLE="background: #e6e6e6">End-rule</SPAN></SPAN></FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<BR><BR>
+</P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<BR><BR>
+</P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<BR><BR>
+</P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Note
+that the configuration of the rule, such as parameters, salience etc.
+is done in a tabular fashion <I>somewhere else</I> in the document. </FONT></FONT></FONT>
+</P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">The
+expressions in the rule are compiled with the aid of a dictionary.
+You can help the lexer out by putting “[ and ]” around
+multi word expressions. Alternatively, you can put whole expressions
+in square brackets, and they will be passed through the natural
+language compiler.</FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt"><I>TODO:
+should some attributes like parameters be in the rule definition?</I></FONT></FONT></FONT></P>
+<H3 CLASS="western"><FONT FACE="Verdana, sans-serif">The Dictionary
+(how it works for starters)</FONT></H3>
+<P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">The
+dictionary is a little embedded grammar. I chose the word dictionary
+as I think it is more intuitive, and it is a very primitive grammar.
+This ideally can be imported from another file if needs be. It would
+be possibly to map between languages this way, within reason.</FONT></FONT></P>
+<P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">This
+should be in an appendix hidden and protected. You can also think of
+things on the right hand side as templates (although it is more
+complex then just string interpolation – such as order of
+operations, and nesting).</FONT></FONT></P>
+<P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">This
+dictionary provides a grammar that is used to help process stuff in
+the Rule sections of the document.</FONT></FONT></P>
+<P><BR><BR>
+</P>
+<DL>
+ <DD>
+ <TABLE WIDTH=652 BORDER=0 CELLPADDING=0 CELLSPACING=0>
+ <COL WIDTH=178>
+ <COL WIDTH=474>
+ <THEAD>
+ <TR>
+ <TH COLSPAN=2 WIDTH=652 VALIGN=TOP BGCOLOR="#e6e6e6">
+ <P ALIGN=LEFT STYLE="font-style: normal; font-weight: medium"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Dictionary</FONT></FONT></P>
+ </TH>
+ </TR>
+ </THEAD>
+ <TBODY>
+ <TR VALIGN=TOP>
+ <TD WIDTH=178>
+ <P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Today</FONT></FONT></P>
+ </TD>
+ <TD WIDTH=474>
+ <P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">new
+ java.util.Date()</FONT></FONT></P>
+ </TD>
+ </TR>
+ <TR VALIGN=TOP>
+ <TD WIDTH=178>
+ <P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">before</FONT></FONT></P>
+ </TD>
+ <TD WIDTH=474>
+ <P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">${left}.before(${right})</FONT></FONT></P>
+ </TD>
+ </TR>
+ <TR VALIGN=TOP>
+ <TD WIDTH=178>
+ <P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">date
+ of</FONT></FONT></P>
+ </TD>
+ <TD WIDTH=474>
+ <P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">convertToDate(“${right}”)
+ //use of a drools function</FONT></FONT></P>
+ </TD>
+ </TR>
+ <TR VALIGN=TOP>
+ <TD WIDTH=178>
+ <P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Log</FONT></FONT></P>
+ </TD>
+ <TD WIDTH=474>
+ <P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">System.out.println(${right})</FONT></FONT></P>
+ </TD>
+ </TR>
+ </TBODY>
+ </TABLE>
+</DL>
+<P><BR><BR>
+</P>
+<P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Ok
+lets explain that. The first cell of the table tells drools that it
+is a dictionary. The left column are the keywords (watch for case
+sensitivity, can be a pain). On the right, there are the little
+expressions. Notice that there is no prescribed language yet. In this
+case it is a mix of java and drools functions. ${right} means the
+token to the right, ${left} means the token to the left. You can also
+use ${2} and ${-3} as well to indicate positions. </FONT></FONT>
+</P>
+<P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Pretty
+simple, it gets a little more complex, later (ie order of operations
+and nesting, some special keywords). </FONT></FONT>
+</P>
+<P><BR><BR>
+</P>
+<H3 CLASS="western"><FONT FACE="Verdana, sans-serif">Tables for
+configuration</FONT></H3>
+<P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">This
+is “somewhere else” in the document. Ideally this sort of
+stuff can be included from external resources if needed (to share
+common stuff). Where this appears in the document is not important.
+Appendix probably makes most sense.</FONT></FONT></P>
+<P><BR><BR>
+</P>
+<P><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">The
+"Rule configuration" table sets up technical attributes and
+parameters for the rule.</FONT></FONT></P>
+<DL>
+ <DD>
+ <TABLE WIDTH=560 BORDER=0 CELLPADDING=0 CELLSPACING=0>
+ <COL WIDTH=140>
+ <COL WIDTH=420>
+ <THEAD>
+ <TR>
+ <TH COLSPAN=2 WIDTH=560 VALIGN=TOP BGCOLOR="#e6e6e6">
+ <P ALIGN=LEFT STYLE="font-style: normal; font-weight: medium"><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Rule-configuration</FONT></FONT></P>
+ </TH>
+ </TR>
+ </THEAD>
+ <TBODY>
+ <TR VALIGN=TOP>
+ <TD WIDTH=140>
+ <P ALIGN=LEFT><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">In-the-beginning</FONT></FONT></P>
+ </TD>
+ <TD WIDTH=420>
+ <P ALIGN=LEFT><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">Salience=1</FONT></FONT></P>
+ <P ALIGN=LEFT><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">xor-group=group1
+ </FONT></FONT>
+ </P>
+ <P ALIGN=LEFT><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">no-loop=false</FONT></FONT></P>
+ <P ALIGN=LEFT><FONT FACE="Verdana, sans-serif"><FONT SIZE=1 STYLE="font-size: 8pt">parameters=(String
+ message, Integer number)</FONT></FONT></P>
+ </TD>
+ </TR>
+ <TR VALIGN=TOP>
+ <TD WIDTH=140>
+ <P ALIGN=LEFT><BR>
+ </P>
+ </TD>
+ <TD WIDTH=420>
+ <P ALIGN=LEFT><BR>
+ </P>
+ </TD>
+ </TR>
+ </TBODY>
+ </TABLE>
+</DL>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<BR><BR>
+</P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=2 STYLE="font-size: 9pt"><I>Thats
+it for now... go have a coffee. Spend some quality time with your
+family. Remember them?</I></FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<BR><BR>
+</P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=2 STYLE="font-size: 9pt"><I>TODO:
+Macros for content sharing. eg (Include something here) and (Import
+something here). Just like literate programming.</I></FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-weight: medium; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<FONT COLOR="#000000"><FONT FACE="Verdana, sans-serif"><FONT SIZE=2 STYLE="font-size: 9pt"><I>TODO:
+Output as HTML the bits that are parsed out.</I></FONT></FONT></FONT></P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; font-style: normal; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<BR><BR>
+</P>
+<P LANG="en-US" ALIGN=LEFT STYLE="margin-bottom: 0.32cm; background: transparent; line-height: 130%; page-break-inside: auto; widows: 2; orphans: 2; page-break-before: auto; page-break-after: auto">
+<BR><BR>
+</P>
+</BODY>
+</HTML>
\ No newline at end of file
Property changes on: trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/NaturalRulesSpec.html
___________________________________________________________________
Name: svn:eol-style
+ native
Added: trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/RuleDocumentListenerTest.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/RuleDocumentListenerTest.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/RuleDocumentListenerTest.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -0,0 +1,32 @@
+package org.drools.natural.ruledoc;
+
+import java.util.List;
+
+import junit.framework.TestCase;
+
+public class RuleDocumentListenerTest extends TestCase
+{
+
+ public void testListBuilder() {
+ RuleDocumentListener listener = new RuleDocumentListener();
+ listener.handleText("ignore me\t \n");
+ listener.handleText("Start-rule");
+ listener.handleText("in a rule");
+ listener.startComment();
+ listener.handleText("a comment");
+ listener.endComment();
+ listener.handleText("End-rule");
+
+ listener.handleText("Start-rule something here");
+ listener.handleText("in a rule");
+ listener.handleText("End-rule");
+
+
+ List list = listener.getRules();
+ assertEquals(2, list.size());
+ String rule = (String) list.get(0);
+ assertEquals("Start-rulein a ruleEnd-rule", rule);
+
+ }
+
+}
Property changes on: trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/RuleDocumentListenerTest.java
___________________________________________________________________
Name: svn:eol-style
+ native
Added: trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/RuleDocumentTest.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/RuleDocumentTest.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/RuleDocumentTest.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -0,0 +1,29 @@
+package org.drools.natural.ruledoc;
+
+import java.util.List;
+import java.util.Properties;
+import java.io.InputStream;
+import java.net.URL;
+
+import junit.framework.TestCase;
+
+public class RuleDocumentTest extends TestCase
+{
+ public void testSimpleParser() {
+ URL url = this.getClass().getResource("simple-ruledoc.html");
+ RuleDocument doc = new RuleDocument();
+ List rules = doc.buildRuleListFromDocument(url, new Properties());
+ assertEquals(1, rules.size());
+ System.out.println(rules.get(0));
+ }
+
+
+ public void testTheSpec() {
+ URL url = this.getClass().getResource("NaturalRulesSpec.html");
+ RuleDocument doc = new RuleDocument();
+ List rules = doc.buildRuleListFromDocument(url, new Properties());
+ assertEquals(1, rules.size());
+ System.out.println(rules.get(0));
+ }
+
+}
Property changes on: trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/RuleDocumentTest.java
___________________________________________________________________
Name: svn:eol-style
+ native
Deleted: trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/HTMLDocParserImplTest.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/HTMLDocParserImplTest.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/HTMLDocParserImplTest.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -1,73 +0,0 @@
-package org.drools.natural.ruledoc.html;
-
-import java.io.InputStream;
-
-import org.drools.natural.ruledoc.RuleDocumentListener;
-
-import junit.framework.TestCase;
-
-public class HTMLDocParserImplTest extends TestCase
-{
- public void testDocParsing() throws Exception {
-
- MockRuleDocumentListenter listener = new MockRuleDocumentListenter();
- InputStream stream = this.getClass().getResourceAsStream("simplest.html");
- HTMLDocParserImpl parser = new HTMLDocParserImpl();
- parser.parseDocument(stream, listener);
- assertTrue(listener.buf.toString().indexOf("line B") > 0);
- assertFalse(listener.comment.toString().indexOf("comment") > 0);
- System.out.println(listener.buf.toString());
- assertEquals("comment", listener.comment.toString());
- }
-
- static class MockRuleDocumentListenter implements RuleDocumentListener {
-
- public StringBuffer comment = new StringBuffer();
- public StringBuffer buf = new StringBuffer();
- private boolean inComment = false;
-
- public void handleText(String text)
- {
- if (!inComment) {
- buf.append(text + "|");
- } else {
- comment.append(text);
- }
- }
-
- public void startTable()
- {
- }
-
- public void startColumn()
- {
- }
-
- public void startRow()
- {
- }
-
- public void endTable()
- {
- }
-
- public void endColumn()
- {
- }
-
- public void endRow()
- {
- }
-
- public void startComment()
- {
- inComment = true;
- }
-
- public void endComment()
- {
- inComment = false;
- }
-
- }
-}
Added: trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/HTMLDocParserTest.java
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/HTMLDocParserTest.java 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/HTMLDocParserTest.java 2005-12-21 07:32:06 UTC (rev 1910)
@@ -0,0 +1,87 @@
+package org.drools.natural.ruledoc.html;
+
+import java.net.URL;
+import java.util.Properties;
+
+import junit.framework.TestCase;
+
+import org.drools.natural.ruledoc.RuleDocumentListener;
+
+public class HTMLDocParserTest extends TestCase
+{
+ public void testDocParsing() throws Exception {
+
+ MockRuleDocumentListenter listener = new MockRuleDocumentListenter();
+ URL url = this.getClass().getResource("simplest.html");
+ HTMLDocParser parser = new HTMLDocParser();
+ parser.parseDocument(url.openConnection(), listener);
+
+ assertTrue(listener.buf.toString().indexOf("A line") > 0);
+ assertFalse(listener.comment.toString().indexOf("comment") > 0);
+
+ assertEquals("comment", listener.comment.toString());
+
+ assertTrue(listener.table.toString().indexOf("left1|right1") > 0);
+ }
+
+ static class MockRuleDocumentListenter extends RuleDocumentListener {
+
+ public StringBuffer comment = new StringBuffer();
+ public StringBuffer buf = new StringBuffer();
+ private boolean inComment = false;
+ public StringBuffer table = new StringBuffer();
+ private boolean inTable = false;
+
+ public void handleText(String text)
+ {
+ if (inComment) {
+ comment.append(text);
+ } else if (inTable) {
+ table.append(text);
+ }
+ else {
+ buf.append(text);
+
+ }
+ }
+
+ public void startTable()
+ {
+ inTable = true;
+ }
+
+ public void startColumn()
+ {
+ this.table.append("|");
+ }
+
+ public void startRow()
+ {
+ }
+
+ public void endTable()
+ {
+ this.table.append("||");
+ inTable = false;
+ }
+
+ public void endColumn()
+ {
+ }
+
+ public void endRow()
+ {
+ }
+
+ public void startComment()
+ {
+ inComment = true;
+ }
+
+ public void endComment()
+ {
+ inComment = false;
+ }
+
+ }
+}
Property changes on: trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/HTMLDocParserTest.java
___________________________________________________________________
Name: svn:eol-style
+ native
Deleted: trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/NaturalRulesSpec.html
===================================================================
--- trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/NaturalRulesSpec.html 2005-12-21 02:40:16 UTC (rev 1909)
+++ trunk/labs/jbossrules/drools-natural-language/src/test/org/drools/natural/ruledoc/html/NaturalRulesSpec.html 2005-12-21 07:32:06 UTC (rev 1910)
@@ -1,303 +0,0 @@
-<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
-<HTML>
-<HEAD>
- <META HTTP-EQUIV="CONTENT-TYPE" CONTENT="text/html; charset=windows-1252">
- <TITLE></TITLE>
- <META NAME="GENERATOR" CONTENT="OpenOffice.org 2.0 (Win32)">
- <META NAME="CREATED" CONTENT="20051214;11592400">
- <META NAME="CHANGEDBY" CONTENT="Michael Neale">
- <META NAME="CHANGED" CONTENT="20051215;16490707">
- <STYLE>
- <!--
- @page { size: 21cm 29.7cm; margin: 2cm }
- P { margin-bottom: 0.21cm }
- H1 { margin-bottom: 0.21cm }
- H1.western { font-family: "Arial", sans-serif; font-...
[truncated message content] |
