Full Text search in Study Inspector#492

Please fix the full-text search. When attempting to search full-text for a nest with ~650 full-text articles, the search crashes the software, usually twice. Wondering if the problem may be related to the large amount of full-text that we have in our nest.

I saw a previous request that was related to this feature: Inspector: Searching Full Text content#387 that indicated full-text search was implemented.

@Eileen Erinoff @Megan Nunemaker fyi

Hi Carol, Thanks for reporting, I’m observing a serious degradation in the performance of this feature. It’s functional, but slow on nests with even 20 FTs, and I get the same behavior as you with 50. I will say, this feature was not developed with searching hundreds of FTs in mind.

Briefly, what’s happening is:

• your filter results in a character-by-character scan of the text content of all the PDFs in your nest
  • this is very slow
• we notice that the result is too slow, and time it out
• when you reload the page, the filter is still set, so we re-attempt to run that filter (repeat from top)

In order to address, we’ll need to build a per-nest index on text content of FT PDFs (which uses an approach much faster than word-by-word scanning).

This should be resolved in release 1.53.4. Searches typically will complete in under a second now, even in large nests. Thanks for filing, and let us know what you think!