Menu

Full text index not working

Mee
2015-01-30
2015-02-14
  • Mee

    Mee - 2015-01-30

    I love SeedDMS, even though the full text index is driving me crazy. It was working when I set it up, but then I realized that refreshing the index does not really add those documents to the index (even though it says that those documents were added). When I rebuild the entire index, the subsequent page shows that all documents were added, but the only content included in the full text info is the content of a single text file, which consists of two words. Then, some days later, I was able to re-create the index and the content was shown correctly in the full text info. Right now it is not working again. I tried different users (admin and another user with admin rights), I tried to reach the server via IP, FQDN or locally - nothing solved my problem.

    Unfortunately, I am not able to reproduce a working recreation of the index and there are no debug or error messages. Can anyone help me to have the full text index working reliably?
    Thanks in advance!

     

    Last edit: Mee 2015-01-30
  • Daniel

    Daniel - 2015-02-01

    What system are you using to run SeedDMS on? Within the settings (tab Advanced) you can find the command lines used for indexing. These tools must exist on your system in order to have the fulltext indexing working properly. Furthermore, there are some PDF files (PDF saved as image) which may not be read out because pdftotext (standard setting for conversion) can't read out the content of the image. In this case, text in this PDF will not be added to the fulltext index.

     
  • Mee

    Mee - 2015-02-02

    Hi Daniel,

    thanks for your reply. I am running SeedDMS successfully on my QNAP-NAS and the indexing/parsing of documents generally works, but the parsed/retrieved data is not taken over into the searchable index-database (the content in 'Volltext Index Info') only shows two words which are contained in a single text file - even though more than 2,500 documents have been passed and added when I recreate the full text index...

    Am I able to explain my problem?

     
  • Mee

    Mee - 2015-02-04

    Update: I am getting the followin php warning, when I try to run indexer.php from the command line
    (php ./indexer.php --config /share/Web/dms/www/conf/settings.xml)

    PHP Warning: getopt(): No support for long options in this build in /share/CACHEDEV1_DATA/Web/dms/seeddms-4.3.14/utils/indexer.php on line 20
    You does not seem to have a valid configuration. Run the install tool first.

    Can this be the reason for the indexer not working properly?

     
  • Daniel

    Daniel - 2015-02-11

    Hi Mee,

    That looks strange, especially the message to run the install tool first. Probably an issue with your configuration file which contains wrong (or even missing) entries.

    Can you check your settings.xml file and see what you find within the converters section?

    Anyway, if possible for you, i would recommend to create a backup (files and database), rename the settings.xml file and run the installation procedure again. Don't forget to put the ENABLE_INSTALL_TOOL file before starting.

    This process will ensure that you have a proper settings.xml file at the end of the process.

     
  • Uwe Steinmann

    Uwe Steinmann - 2015-02-14

    The getopt() warning causes the trouble. Because long options are not supported on your system, the configuration file cannot be specified. As a consequence the programm doesn't know anything about your seeddms installation.

     

Log in to post a comment.