Here's a .srt file which both ffmpeg and pysubs2 parse without complaint, which subparse fails on:
1
00:00:06,000 --> 00:00:12,074
text1
2
2
00:11:49,412 --> 00:11:52,858
text2
Both ffmpeg's behavior and pysub2's behavior is to nest the extra line into the previous entry
1
00:00:06,000 --> 00:00:12,074
text1
2
2
00:11:49,412 --> 00:11:52,858
text2
I was working on a python binding for subparse and a file similar to this one triggered an error I wasn't expecting. https://github.com/john-parton/subparsepy