Add `text_len` to `TextWrapper` for ANSI and variable character width awareness

Feature or enhancement

Proposal:

This is a proposal to add a text_len parameter to textwrap.wrap, textwrap.fill, textwrap.shorten and textwrap.TextWrapper.

That way we can teach these functions how to measure the length of visible text.

The main use-case is ignoring ANSI sequences like so:

>>> import re, textwrap
>>> visible_len = lambda s: len(re.sub(r'\x1b\[[0-9;]*m', '', s))
>>> colored = 'normal \x1b[31mcolored\x1b[0m words here'
>>> lines = textwrap.wrap(colored, width=14, text_len=visible_len)
>>> [re.sub(r'\x1b\[[0-9;]*m', '', line) for line in lines]
['normal colored', 'words here']

This allow flexibility without having to maintain or introduce full ANSI awareness to all the string-manipulating utilities.

And I am pretty sure some users can come up with creative usage other than ANSI, and it provides a future escape-hatch.

An implementation is available at #152702

Has this already been discussed elsewhere?

I have already discussed this feature proposal on Discourse

Links to previous discussion of this feature:

This is a direct follow up on https://github.com/python/cpython/issues/142035 , which introduced a local patch in _split_lines via https://github.com/python/cpython/pull/154634 to address that bug. A new issue has been requested by a project member to track the generalization of that local patch.

Linked PRs

• gh-152702

Commencez par examiner l’implémentation dans la PR #152702 et la discussion associée dans les issues #142035 et #154634. Comparez le paramètre text_len proposé dans textwrap.wrap, fill, shorten et TextWrapper, puis vérifiez que le cas d’utilisation de la largeur visible et le comportement existant sont couverts avant de considérer le travail comme terminé.