This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Indexing folders?

Is it possible to index folders using the filesystem connector?

The goal is to be able to browse a file structure via IDOL queries

The solution I have come up with doesn't work for traversing through multiple levels of folders with no contents

  • On a per document basis, you can write a Lua script in connector that splits DREREFERENCE into fields that represent the folder structure.   Lua parse_csv() where delimiter is / should work.

    I think this is what you were asking to do.  If not, then please re-phrase. 

  • That's as far as I got, however it isn't quite meeting the requirements - is it possible to index them such that they appear as their own "hits" so that they can be subject to sorting/filtering/paging?

  • Unfortunately, I don’t see any settings in the FileSystem Connector to automatically make an indexable document as you desire.

    A solution is to write a script - pick your language (Powershell, Python, Perl, non-P languages) and produce an a=ingest packet to send to CFS just the connector normally does. Folder deletes is maybe tricky part to handle without the connector data store.

    You could skip CFS and make IDX and send to DIH / Content directly.

    Assuming you have a valid support account you can file a support ticket asking for an enhancement where it will get consideration.
  • Thanks, how would I trigger the script?

    Is it possible to trigger an ingest from a Lua ingest script?

Reply Children
  • Sorry for delay in responding.  Yes, it's been too long.

    send_aci_action() or send_aci_command() sending a=ingest to CFS should work. 

    And after thinking about this a bit more, you'll surely end up with lots of duplicate folders.  This suggests that DRECOMPACT to accommodate the increased DREADDs that end up replacing prior version of "folder" IDOL document should be a consideration.