I am trying to use the calibre conversion feature to convert ebook written in Georgian from PDF to Mobi. The effect is however very poor. Someone in 10 lines gets unwrapped. I tried to use different unwrapping factors but to no avail. I tested this feature on a Polish book and the effect was satisfactory with default settings.
Important note about Georgian – its script is unicameral. So I suspect that the lack of capital letters at the beginning of the sentences can cause the unwrapping algorithm to fail. But I didn't look at the code, so maybe someone who knows the algorithm better could confirm.
So my questions are:
- Can it be the unicameral character of the alphabet that causes the problem?
- Can I tweak the algorithm to fix it?
- Is there any other conversion tool that could work for me here.