Alass not generating subtitles for non UTF8 encoded input subtitles.

[chadniko]

This is really may not be a problem with subsyncarr....but after adding the alass engine some subs don't get generated in which the other engines do generate there sub versions. I went into the container and ran the alass command from the command line and it complains that it can't parse the input sub.
Going to the alass github it shows that a previous upgrade request to auto covert the input sub to UTF8 format was now incorporated into its latest.
So wondering if your container is using the latest alass? If so maybe alass still cannot parse non-UTF8 input sub files. If this is the case may be possible to convert the non-UTF8 using a set of cli commands like chardet to detect the sub file format and iconv to convert UTF8 prior to running alass as a workaround

TIA for any consideration.

[chadniko]

here's the the issue link in alass kaegi/alass#25
This was done some time ago and seems that it should already be in subsyncarr.

The file is encoded as text/plain; charset=iso-8859-1 according to file -bi
so after converting to UTF8 using : vim +'set nobomb | set fenc=utf8 | x'
subsyncarr processed the file correctly

[johnpc]

The raw alass binary is checked in directly to this repo (see https://github.com/johnpc/subsyncarr/blob/main/bin/alass). I downloaded it from the alass releases page (https://github.com/kaegi/alass/releases/tag/v2.0.0) which is from 2019.

If we check in an updated alass binary with the latest build for alass-linux64 it would probably include a lot of fixes including the one you've mentioned.

[chadniko]

Yes has several updates after v2.0.0 release but has not generated a new release/build and binary which includes new non UTF8 handling...hence the issue.....this issue shows on the alass github....maybe I can convince to build a new release.....