7 Replies Latest reply: Nov 23, 2012 10:16 AM by TimK RSS

    spell dictionary

    BCHK
      Hi,

      anyone try to use the --spellmode paramenter?

      i tried to use both aspell and espell for dgidx,
      set Arguments in workbench for Dgidx: -v compoundDimSearch spellmode espell

      but i cannot find a file called <db_prefix>.worddatn_default

      however, i use the <db_prefix>.spell_config.xml under pipeline folder.

      anyone know how to set aspell_AND_espell in that xml file?

      Thank you
        • 1. Re: spell dictionary
          BCHK
          Hi,

          anyone knows how to set and check it is running with aspell_AND_espell?
          • 2. Re: spell dictionary
            BCHK
            Fixed by setting AppConfig.xml instead of setting in workbench.
            • 3. Re: spell dictionary
              Pravin Chikhale
              Hi BCHK,

              Is changing configuration worked? Did it created that worddat file?

              I tried with french data, but it didn't worked. I think i missed some configuration. Can you please explain configuration you made.

              Thanks,
              Pravin.
              • 4. Re: spell dictionary
                953835
                If you change AppConfig.xml you need to run "initialize_services.sh" to pick up the changes.
                • 5. Re: spell dictionary
                  Michael Peel-Oracle
                  If you are using the Deployment Template, you can add the following to the arguments in AppConfig.xml <dgidx />:
                       <arg>--spellmode</arg>
                       <arg>aspell_AND_espell</arg>

                  This requires a baseline to take effect (and initialize_services.bat if you have removed the --update-definition command from runcommand.bat. You can check that it is executing correctly by looking at the Dgidx log (./logs/dgidxs/Dgidx/Dgidx.log).

                  For French, you may wish to test this versus using just espell as aspell uses English phonetics to perform spell-correction.

                  Michael
                  • 6. Re: spell dictionary
                    975761
                    Hi,

                    thanks for this tip. We've run initialize_services.bat + baseline update. After this all Thesaurus entries in Endeca Workbench were gone.
                    The old entries are still available in this file:
                    /data/web_studio/dgraph_config/mccru.thesaurus.xml

                    But all other files are empty.

                    ./data/web_studio/config/mccru.thesaurus.xml
                    ./data/processing/mccru.thesaurus.xml
                    ./data/complete_index_config/mccru.thesaurus.xml
                    ./data/forge_output/mccru.thesaurus.xml
                    ./config/cas_crawl_pipeline/mccru.thesaurus.xml
                    ./config/pipeline/mccru.thesaurus.xml


                    New entries in Endeca workbench are written into the file
                    ./data/web_studio/config/mccru.thesaurus.xml
                    Where can we put our old file, that Endeca Workbench takes the old Thesaurus words?
                    Thank you in advance.
                    • 7. Re: spell dictionary
                      TimK
                      You can approach this two ways - first, save the thesaurus file somewhere so it won't be overwritten - this is your gold copy for now.

                      The example commands below are in linux format - you can modify to make windows specific. Whatever you do, back up everything as you go in case you need to recover something else.

                      Issue a command to export all workbench managed files to your config/pipeline directory - this will overwrite the same files already there but this should be ok if you have setup all of these files in AppConfig for workbench managed files. (*You could also export to a completely separate location to be safer, then do all the rest of the steps to/from that location).
                      --+
                      echo Extracting $APP workbench files to /<YourPathHere>/$APP/config/pipeline/
                      +emgr_update host localhost:8006 app_name $APP action get_ws_settings prefix $APP --dir /<YourPathHere>/$APP/config/pipeline+
                      --+

                      Copy your gold copy of the thesaurus file into the config/pipeline directory.

                      At this point, I would archive up everything as a saved copy.

                      Approach 1:
                      Issue a command to clear all files from your workbench
                      --+
                      echo "Cleaning pipeline and page builder xml files from $APP"
                      +emgr_update host localhost:8006 app_name $APP action remove_all_settings dir /<YourPathHere>/$APP/config/pipeline --prefix $APP+
                      --+
                      Issue a command to import all files from config/pipeline into your workbench
                      --+
                      echo "Loading pipeline xml files to $APP"
                      +emgr_update host localhost:8006 app_name $APP action update_mgr_settings dir /<YourPathHere>/$APP/config/pipeline --prefix $APP+
                      --+
                      If you use page builder/experience manager templates, don't forget to import those as well.

                      Approach 2:
                      Run initialize services
                      If you use page builder/experience manager templates, don't forget to import those as well.
                      Run baseline updates, etc.