#250 cannot search for words "mortgage" and "mis"

resolved
closed-works-for-me
htfuzzy (8)
5
2005-07-12
2004-12-17
Mammai
No

we get the following error when searching the
word "mortgages"

HTTP/1.1 500 Internal Server Error Date: Fri, 17 Dec
2004 14:57:08 GMT Server: Apache/1.3.27 (Unix)
Connection: close Content-Type: text/html; charset=iso-
8859-1
Internal Server Error
The server encountered an internal error or
misconfiguration and was unable to complete your
request.
Please contact the server administrator,
can_webadmin@sunlife.com and inform them of the time
the error occurred, and anything you might have done
that may have caused the error.

More information about this error may be available in the
server error log.

We are using htdig version 3.1.6

Please contact me at mvenu@sunlife.com

Discussion

  • Nobody/Anonymous

    Logged In: NO

    what is "Bus Error" error

     
  • Mammai

    Mammai - 2004-12-17
    • assigned_to: nobody --> grdetil
    • priority: 5 --> 9
     
  • Mammai

    Mammai - 2004-12-17

    Logged In: YES
    user_id=1180246

    See below, when running the htdig directly on the server

    /www/v5/live_internet/docs/htdig
    ./htsearch.cgi -vvvvvvv -
    c /apps/htdig/conf/canada.conf
    Enter value for words: mis
    tempWords: 'mis:0 '
    Boolean: 'mis:0 '
    DB2 problem...: /apps/htdig/canada/db.accents.db: No such
    file or directory
    DB2 problem...: /apps/htdig/canada/db.soundex.db: No such
    file or directory
    initial: ''
    Fuzzy on: mis
    exact mis
    synonyms
    accents
    Bus Error

    The Bus Error is the problem we need to look into and it
    seems to come from htfuzzy.

     
  • Gilles Detillieux

    • priority: 9 --> 5
     
  • Gilles Detillieux

    • assigned_to: grdetil --> nobody
     
  • Gilles Detillieux

    Logged In: YES
    user_id=149687

    See http://www.htdig.org/FAQ.html#q5.7 and all the related
    questions to which it refers. If none of these offer any
    help, and there are no meaningful/helpful error messages in
    the web server's logs to guide you to a solution, then post
    a message to the htdig-general mailing list (see
    http://www.htdig.org/FAQ.html#q1.8 and
    http://www.htdig.org/FAQ.html#q1.15\), giving all relevant
    details (including anything in the server logs at the time
    of the error). This is likely a configuration problem
    rather than a bug.

    A Bus Error usually occurs when a program tries to
    access misaligned data on a CPU where alignment
    matters. What processor are you running htsearch on?
    On some systems, you also get a bus error when
    dereferencing a null pointer.

     
  • Gilles Detillieux

    Logged In: YES
    user_id=149687

    Yeah, the bus error does seem like it might be a bug in
    htfuzzy, but try running "htfuzzy accents" to rebuild your
    accents database and see if that makes the error go away.

     
  • Nobody/Anonymous

    Logged In: NO

    I rebuilt the htfuzzy, but did not resolve the issue??

     
  • Mammai

    Mammai - 2004-12-17
    • priority: 5 --> 8
    • assigned_to: nobody --> grdetil
     
  • Mammai

    Mammai - 2004-12-17

    Logged In: YES
    user_id=1180246

    Do i need to re-index the search after rebuilding the accents
    db??

     
  • Gilles Detillieux

    • priority: 8 --> 5
     
  • Gilles Detillieux

    Logged In: YES
    user_id=149687

    To answer your off-list followup questions:

    > How do i run "htfuzzy accents" to rebuild accents database??

    Normally you just type in those two words at the command
    prompt, or put them in the shell script you use to reindex
    your database.

    > Do i have to re-index the search after running the htfuzzy?

    No, it's the other way around. See the "Building the
    databases" section of "Running ht://Dig" at
    http://www.htdig.org/running.html for details. It states: "A
    few of the other fuzzy match algorithms use databases that
    are derived from the word database built by htdig/htmerge,
    so if you use these algorithms you should rebuild their
    databases with htfuzzy every time you update your index."
    The accents database is one of these, so you have to rerun
    "htfuzzy accents" after htdig and htmerge have rebuilt the
    words database. The same applies to the "soundex" database,
    which I also noticed htsearch is complaining about.

    Note that htdig, htmerge, htfuzzy and htsearch must all use
    the same configuration file, or at least files that define
    the same databases via the database_dir attribute. The error
    messages from htsearch below suggest that either you didn't
    build the accents and soundex databases, or if you did
    they're not in the same place as htsearch expects to find
    them. I.e., to follow your example below, you should use "-c
    /apps/htdig/conf/canada.conf" as a command line option for
    all ht://Dig programs, so "htfuzzy -c
    /apps/htdig/conf/canada.conf accents soundex".

     
  • Mammai

    Mammai - 2004-12-17

    Logged In: YES
    user_id=1180246

    i ran exactly this
    htfuzzy -c /apps/htdig/conf/canada.conf accents soundex

    still i see the same error

     
  • Mammai

    Mammai - 2004-12-17

    Logged In: YES
    user_id=1180246

    the htfuzzy , htmerge is in /apps/htdig/bin directory and
    htsearch is in /apps/htdig directory. So what should i do now

     
  • Gilles Detillieux

    Logged In: YES
    user_id=149687

    What files are in the /apps/htdig/canada/ directory? Can you
    show me the output of "ls -l /apps/htdig/canada/" ? Are all
    the files owned by the same user, and are they all readable
    by htsearch?

    How did you build your index in the first place? Did you
    run htdig and htmerge separately yourself, or did you run
    the rundig script, or some other script? Did you use the
    same /apps/htdig/conf/canada.conf file then? What is
    database_dir set to in that configuration file?

     
  • Mammai

    Mammai - 2004-12-17

    Logged In: YES
    user_id=1180246

    Hi

    Looks like the issue is fixed, but we are testing. So do you
    say that after each re-indexing , we need to build the
    accents and soundex databases??

     
  • Gilles Detillieux

    • labels: --> htfuzzy
     
  • Gilles Detillieux

    Logged In: YES
    user_id=149687

    Yes, I thought I was clear on that point. If you use any of
    the soundex, metaphone, or accents databases in your
    search_algorithm attibute definition, you must rebuild these
    after reindexing. This is because these databases are
    generated from the words in your words database, which is
    updated after running htdig and htmerge. If you use rundig
    to reindex, you can uncomment the line in there which is set
    up to call htfuzzy, and then customize it for the databases
    you actually need. If you use another script, you can add
    htfuzzy to that script, with the appropriate arguments,
    after the calls to htdig and htmerge.

    The endings and synonyms databases normally only need to be
    built once, as they're built from static dictionary files,
    rather than the words database.

     
  • Mammai

    Mammai - 2004-12-17

    Logged In: YES
    user_id=1180246

    Thank you very much. All your help is greatly appreciated.

     
  • Gilles Detillieux

    • milestone: --> resolved
    • status: open --> closed-works-for-me
     

Log in to post a comment.

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

JavaScript is required for this form.





No, thanks