MohrJonas / obsidian-ocr

Obsidian OCR allows you to search for text in your images and pdfs
GNU General Public License v3.0
281 stars 5 forks source link

Many issues (Windows) #56

Open py9mrg opened 4 months ago

py9mrg commented 4 months ago

Hello,

I've been trying to get this plugin to work on Windows for a couple of hours (haven't tested Linux yet) and am really struggling. First it wouldn't load but fortunately this comment in issue #51 fixed that part.

Then it would load but not do any recognition, hunting through the log I found I needed to do 3 things:

  1. install ghostcript due to a warning about not finding gswin32c.exe
  2. line 56295 - change the command to just magick because convert has been deprecated (error about the command)
  3. line 56310 - move the conversion arguments to after the input file name (but before output file name) because arguments before the input file have been deprecated (error about no input file)

At that point it would work in the sense that checking through the logs I could see the conversion happening and even the converted text would be in the logs. BUT, the search function in Obsidian still doesn't do anything. No idea why - I think the transcripts aren't being saved correctly as I can't see anything in sqlite, or anywhere else, but I don't know why. At this point I'm at a loss as I can't see anything that indicates why.