full text search doesn't work

[Ourithi]

Hello,

I am trying to run full text search with tesseract OCR on an apache2 nextcloud 20.0.10 webserver. The full text search feature works with documents that were already present on the nextcloud server (the nextcloud manual for example) but when i create a new pdf and add it to the server it doesn't work. The tests and first index work fine. There is nothing in the apache error log.
Any ideas?

[thomasgg23]

Is the adding part or the ocr part done via script or via the nextcloud webinterface?

[Ourithi]

Is the adding part or the ocr part done via script or via the nextcloud webinterface?

it is done via the nextcloud web interface

[Ourithi]

On a side note I meddled here and there and now I get the following error when trying to run the index (after having cleared the indexes table):
"call to a member function getContent() on string in ./files_fulltextsearch/lib/Service/FilesService.php:761"

[Ourithi]

I fixed the error but it still doesn't work. I think it's related to the whole full text search having problems with pdf files and all. Does anybody have an idea?

[thomasgg23]

I checked our installation and we dont use the Full Text tesseract script.
We convert the files with a local script and ocrmypdf and then run the occ command to index them again. Works for several months now.

There is an flow "Workflow OCR" which also uses ocrmypdf to convert any new pdf file uploaded. Perhaps you should check this one out.
https://github.com/R0Wi/workflow_ocr

[sinichi19]

check if you have Group Folder feature enable, try to disabled it
#634

[ArtificialOwl]

does your PDF contain text layer ?

• if yes, check that the PDF option in Admin Settings/Full Text Search is enabled
• if no, check the configuration of your tesseract

[ArtificialOwl]

please paste result of ./occ fulltextsearch:check

[joshtrichards]

Closing due to "needs info" after 3 years. :)