avatar Bite 291. Find the fastest speech

In this Bite you will parse an SRT (SubRip Subtitle file / video captions or subtitles). 

This is a snippet that shows how such a file is formatted:

...
5
00:00:18,080 --> 00:00:22,180
Because if you're really new to Python, we really recommend to start with the Newbies,

6
00:00:22,180 --> 00:00:26,260
and otherwise gauge your level doing the 10 Introductory Bites.
...

Each block is separated by two newlines and consists of:
- a section number,
- a start and an end time, separated by -->,
- the actual caption or subtitle text.

Calculate the differences between start and end times for each section, returning a list of section ids ordered by the fastest speech (= most chars spoken per second).

Here is an example how the code should work:

>>> text = """
... 1
... 00:00:00,498 --> 00:00:02,827
... Beautiful is better than ugly.
...
... 2
... 00:00:02,827 --> 00:00:06,383
... Explicit is better than implicit.
...
... 3
... 00:00:06,383 --> 00:00:09,427
... Simple is better than complex.
... """
>>> from srt import get_srt_section_ids
>>> get_srt_section_ids(text)
[1, 3, 2]
Login and get coding
go back Intermediate level
Bitecoin 3X

28 out of 28 users completed this Bite.
Will you be Pythonista #29 to crack this Bite?
Resolution time: ~66 min. (avg. submissions of 5-240 min.)
Pythonistas rate this Bite 3.67 on a 1-10 difficulty scale.
» You can do it! 😌

We use Python 3.8