FarCrySolrPro File Contents not indexing

#1

Hello all, I’ve been trying out the FarCrySolrPro plugin. Mostly testing has been pretty smooth, though I’m getting hung up on indexing the contents of documents. I did find a Google Groups conversation with similar symptoms, but the steps leading to the solution weren’t outlined specifically.

It sounds to me like I should tell the plugin to index the dmFile type, the plugin should run everything through Tika, and I should be set right?

If more background helps I’m on FarCrySolrPro 1.4.1, FarCry 7.2.8, and the FarCry CMS plugin (version unknown, but set up in 2013).

I’m suspecting my setup isn’t detecting dmFile uploads. Most of our uploads are PDF’s and Docx files, there is definitely text available to read an I’m finding nothing after what appears to be a successful index.

#2

Hey there. Here are a couple things to consider/try:

  • When configuring the type index, it does not have to be dmFile. It can be any field with a ftType=“file” (which, yes, the dmFile content type does have - see screenshot below).
  • Did you remember to index the field? (sorry if that seems obvious, but just making sure). See screenshot.

  • Check the logs - You can try running Solr from a terminal/command prompt and watch the logs. See if there is an error when you try indexing the document(s).
    – If you’re getting an error, you can try updating Tika to a newer version.

Also, this is unrelated, but if you want to index the file size, Tika can get that info for you as well. Find the section called Document Size and read the hint info for more details. FCSP will return the file size as an extra field.

Let us know how it goes.

#3

Hello Jeff,

Files are indexing for me now. Indexing the filename field was obvious, selecting text wasn’t. I had tried selecting string and phonetic, maybe I should have tried all the recommended options first. It is useful to know that any file type can be indexed, don’t know if we’ll need to know that but good to know.

I didn’t see any errors in the command line today or previously, but everything works now so there wouldn’t be any I expect.

One small piece of feedback on the documentation, it appears I need to place an empty folder labeled ‘work’ at \plugins\farcrysolrpro\packages\custom\cfsolrlib\solr-server as per this stack exchange thread. Unless this issue isn’t common it might be handy to mention adding this folder or add it to the git repo.

Otherwise thanks for the tips, and thanks to yourself and Sean for creating this plugin. Despite some wheel spinning I’m sure this was much faster than integrating search myself.

#4

I’m glad it’s working for you now. Enjoy.