Configure Endeca to use Pipeline Thesaurus

Question

Running MDEX 6.4.0 and ToolsAndFrameworks 3.1.1 and created the project with the deployment template.

Have created a single (one-way) thesaurus entry in Experience Manager that sends Nescafe to Nescafé and all was working fine. However after a couple of weeks realised that it no longer works and the target value became completely corrupted, basically replacing the é with a random amount of garbage:

NescafÃÆââââ‚....(repeats alot) 

The system uses import/export_site.bat to migrate content between environments, and it is my understanding that the thesaurus data migrates along with this extracted file. I assume there is some way of using the emgr utility to migrate the thesaurus as well but that adds extra complexity in our migration process and also makes it difficult to commit the thesaurus.xml to source-control.

I've previously worked with the thesaurus.xml at a pipeline level and have been able to add many French/Danish words in there with non-ASCII characters. Thus, knowing that it works I'm keen to revert to using the pipeline .thesaurus.xml file to support the thesaurus via DevStudio. I don't seem to be able to find any documentation on how to switch back to the pipeline thesaurus.

Any help appreciated.

Answer 1

Found an answer on the Oracle Commerce Communities (need to log in).

Unfortunately, the thesaurus is no longer able to be maintained via the pipeline XML and emgr_update settings in 3.1 Tools And Frameworks, since it is now managed within the content repository. The only supported mechanism for importing/exporting the repository contents is via the application's import_site and export_site scripts, or editing entries directly in Workbench itself.

In the next version the configuration is going back to being more file-based instead of only accessible within the repository, so the files that contain thesaurus and rules, etc, should become easier to manage if you preferred the older architecture of XML config.