Hello, We just completed Indic Wikisource consultation 2018, and one of the most important part of the consultation was the Wikisource tech needs-assessment We have got a few tools and scripts suggestion which Indic TechCom may work on.
Meanwhile, please have a look at: https://meta.wikimedia.org/wiki/Indic-TechCom/Tools/IndicOCR or the tool link directly https://tools.wmflabs.org/jayprakashbot/ This is a web OCR tool. Currently Google OCR does not support the following languages and it will help there
1. Malayalam Wikisource 2. Telugu Wikisource 3. Odiaa Wikisource 4. Gujarati Wikisource 5. Kannada Wikisource 6. Punjabi Wikisource
I'll keep the list informed about further development.
Thanks Tito Dutta Note: If I don't reply to your email in 2 days, please feel free to remind me over email or phone call.