Detect input (.srt) file encoding about srt-to-vtt HOT 5 CLOSED

What if you put https://github.com/finnp/to-utf-8 inbetween the input the srt-to-vtt?

var utf8 = require('to-utf-8')
...

source.pipe(utf8).pipe(srtToVtt)

from srt-to-vtt.

sure, i can do that in my code, but if you want to use srt-to-vtt from CLI this should be done in srt-to-vtt.

But if the module do srt to vtt and srt COULD BE in non-UTF8 and VTT MUST BE in UTF8, IMHO this should be handled by srt-to-vtt.

I can always send PR and get accepted

from srt-to-vtt.

after some testing i found to-utf-8 doesn't recognize correct encoding because it uses default peek-stream options to get the sample to do the detection. The input is an srt, so the first line is always 1 because its the caption number. So, i need to detect first the encoding from real subtitle text, then pass it to utf8(encoding) to do the conversion

from srt-to-vtt.

PR finnp/to-utf-8#2 for better encoding detection, if that patch is accepted encoding conversion should be easy as change

return pumpify(split(), parse)
to
return pumpify(utf8({newline:false}), split(), parse)
in
https://github.com/mafintosh/srt-to-vtt/blob/master/index.js#L27

from srt-to-vtt.

send a pr for that change once your pr lands in to-utf-8 :)

from srt-to-vtt.