Request: Algorithm description of codepage determination

[Heavygrass]

Hi!
Is there a description of the algorithm used for text codepage determination?
Or which function/file it is implemented in?

I've read whole man page - it mentions some statistical analysis and that's it.
Does ENCA uses language dictionaries?

If we had a description of the algorithm it would be easier both for me and other people to contribute to it to hopefully improve codepage/language recognition accuracy.

Thank you.

[Syed-Shahrukh-OSSRevival]

#87 adds how to add a new language.
Regarding the algorithm its purely statistical in nature.

[Heavygrass]

Thanks, but simply saying its "statistical" doesn't describe the algorithm.

[Syed-Shahrukh-OSSRevival]

@Heavygrass yes this is not enough to describe the algorithm. Project documentation is not complete. So any help in documentation will be helpful. To get a better understanding of how things are working you can start by adding a new language or character set to enca.
enca supports guessing older charsets for languages other than English which is the standard of almost all character sets in range [0x00 - 0x80].
There is now a document available to help add a new language to enca.