Add support for Wren language

[PureFox48]

This PR adds syntax highlighting support for the Wren programming language.

Wren's github
Wren's website

[Anteru]

Thanks for the contribution! Please check the comments to (dramatically) reduce the number of generated tokens, at the moment, you're matching character-by-character in various places which produces enormous amounts of tokens.

[Anteru]

Can you please take a look at the failing test? --randomly-seed=2087311863? I presume it generated some input which led to an infinite number of tokens being generated.

[PureFox48]

Unfortunately, I don't have a working Windows machine - I only use Linux these days - so I've no idea why the first Windows test is failing. Possibly, it doesn't like the prefixes I've used - r'(?<!\.)' in the keywords section. I could change those to r'\b' without difficulty because, syntactically, keywords can't be used as method names in Wren.

Do you have any clue what the problem could be from past experience?

Incidentally, on my own machine (Ubuntu 22.04), 'make test' runs fine - 4131 passed, 4 skipped.

As regards 'check (pull_request)` failing, this is my first docstring:

"""
    pygments.lexers.wren
    ~~~~~~~~~~~~~~~~~~~~
    Lexer for Wren.
    :copyright: Copyright 2006-2022 by the Pygments team, see AUTHORS.
    :license: BSD, see LICENSE for details.
"""

Looking at other lexers, I suspect I need to insert a blank line before and after 'Lexer for Wren'.

My second docstring looks like this but I can't see anything wrong with it at all:

    """
    For Wren source code, version 0.4.0.

    .. versionadded:: 2.14.0
    """

[birkenfeld]

You're correct regarding the module docstring. It also should begin in line 1, not line 2.

As for the Windows test, the tests have a random component (testing well-behavedness with a random sequence of input characters), so it should be reproducible regardless of OS with the pytest argument that @Anteru provided.

[PureFox48]

OK, thanks.

I'll change the module docstring and see if I can get to the bottom of the other problem by running pytest myself with that random seed though I only have Python 3.10.6 installed.

[PureFox48]

I've corrected the module docstring but other issue not yet resolved.

[PureFox48]

Hopefully, I've fixed the failing test by adding another regex at the end of 'root' to catch an arbitrary series of characters.

I've also taken this opportunity to reclassify 'this' as an ordinary keyword (analogous to 'super') rather than a named built-in as it wasn't really highlighting properly before.

[jeanas]

Hopefully, I've fixed the failing test by adding another regex at the end of 'root' to catch an arbitrary series of characters.

This was papering over the problem, skipping the positions where it would occur ... The underlying problem was non-terminating behavior with unmatched closing parentheses. Your approach assumed popping when seeing (?=\)) would go to 'parenthesized' or 'interpolation', but that was not true in case of invalid input like this (validating input is not a goal, but always terminating is). I've pushed a solution for this to your branch that is also simpler, making use of include. Let me know if it looks OK to you, as this was the last point to check for me before merging this.

[PureFox48]

Ah, I see now where the problem was.

Yes, it looks OK to me and is working fine.

Thanks for your help in finally resolving this issue :)

[jeanas]

Happy to help. Merged now.