RE2 support

[Aloso]

any chance for RE2 support? (go and DuckDB for instance).
I'm using pomsky to generate regexes (w/o the features RE2 lacks) so testing with PCRE is equivalent but having this checked off during tests would be nice.

Originally posted by @fundef1 in #112 (comment)

[Aloso]

@fundef1 let's discuss it here, since the other issue is only about testing.

[Aloso]

RE2 is pretty similar to Rust (in that it doesn't support advanced features like lookaround assertions or backreferences, which can have exponential runtime performance).

From the documentation:

• The counting forms x{n,m}, x{n,}, and x{n} reject forms that create a minimum or maximum repetition count above 1000. Unlimited repetitions are not subject to this restriction.
• \w,\d, \s and \b are not Unicode aware (like in JavaScript)
• Supported Unicode properties:
  • General categories (but not LC)
  • Scripts

So there will be some restrictions:

1. To use [word] or %, you need to disable unicode;
2. < and > are unsupported
3. No binary Unicode properties, such as Alphabetic or Emoji
4. No Unicode script extensions
5. No lookaround, recursion, backreferences, or atomic groups
6. No repetition with a bound bigger than 1000
7. No Grapheme

There might be more, which I will find out once I test RE2 more thoroughly.

[fundef1]

Would be great if you could include explicit RE2 support for generating regexes.
I'm already using pomsky with RE2, using the subset/avoiding the limitations you indentified, so no issue there.

[Aloso]

Implemented in a157149