Closed uhuntu closed 1 year ago
I'm sorry but meanwhile I've modified the devel branch due too Robocop tests. Pull the updated devel branch and push your changes again.
Yes, I re-do it again, please help to review, thanks.
As the final step, fix the errors in Rubocop tests and we can merge it :-)
I was thinking for a while about it and found a better approach, I think. We can add a filter to omindex
command to parse image files directly. E.g.:
_extra/xapianindexer.rb
...
OMINDEX = "/usr/bin/omindex --filter=image/png:'tesseract %f -'"
...
# unless File.exist?(OMINDEX)
# my_log "#{OMINDEX} does not exist, exiting...", true
# exit 1
# end
...
PNG files are then indexed too.
An output on my computer. I have just one PNG with "Karel" in the image.
So, I suggest following this approach because:
Looks awesome, it is more elegant!
I'm sorry for not accepting your pull request. I've just extended the documentation in #135. Thank you for your effort and contribution to the plugin.
Hi there, this PR is an interesting idea to OCR the png/jpg files.