Support for more document languages

[eikek]

Languages are currently english, german and french. This is what is supported by stanford-nlp. But other languages good be added easily without the nlp support. For these a fallback could be provided, then adding more languages is not hard. Maybe get rid of NLP alltogether.

See: #461

[mrtnggnn]

Docspell 0.17.1 installed with the docker-compose method seems not to have the french language installed for tesseract.
French is the default language for my collective.

2020-12-31T17:03:10: ocrmypdf stderr: ERROR - The installed version of tesseract does not have language data for the following requested languages: fra
2020-12-31T17:03:10: PDF conversion failed: Command result=3. No output file found.. Go without PDF file

[eikek]

Thanks @mrtnggnn – this is missing indeed! I'll create a new issue from your comment to fix this bug in the docker file.

[eikek]

Next release will include the following languages for document processing: Spanish, Italian, Portuguese, Czech, Dutch, Danish, Finnish, Norwegian, Swedish, Russian, Romanian

If you'd like others to be included, please let me know.

[maciekb]

Please consider to add Polish language processing support.

[eikek]

Please consider to add Polish language processing support.

@maciekb Yes, of course! Let's create a new issue #1345 for this, otherwise it gets lost too easily. Can I ping you for help, because unfortunately I don't know any Polish.