I tried to test this using ^a+ (.*)$ and this text:

a bc
dac

But it didn't seem to work, am I doing something wrong?

Thanks for testing it! That regex (^a+ (.*)$) would match a group bc for the first line, but it doesn't match any groups for the second line, so the line stays the same dac. The result would be, as you said:

a bc // ordered as bc
dac  // ordered as dac

What was the expected result? A regex that would change the sorting order would be \w ?(.*). That would leave us with the groups bc and ac, resulting in:

dac  // ordered as ac
a bc // ordered as bc

If you have an opinion on improving the underlying implementation and/or user interface I would gladly take it!

Thanks for the example, my mistake. First thing I tried after that was ^([Tt]he|[Aa]) (.*)$ but it didn't seem to work because I think the ([Tt]he|[Aa]) is the first capture group? Do you think we also need a capture group number setting?

A capture group number setting is a great idea to make it really customizable! I'll work on it.

Hey @Tyriar, sorry for taking so long.

Added a linePreprocessorRegexCaptureGroup that defaults to 1, works great and adds tons of extra customization to the extension.

This might be more generally and more flexibly implemented as a comparison customization, as @Tyriar suggested in #40 (comment), eg a.match(/(the|a)\s+(.*)/i)[2] > b.match(/(the|a)\s+(.*)/i)[2].