Refactor: factor-out usage of itertools.groupby from generate_identifier_pattern.py script

[jayaddison]

Refactors the generate_identifier_pattern.py script; initially this was done in order to experiment with removal of itertools.groupby code that flake8 pre-commit checks were complaining about (see #1835).

• Resolves flake8 warning: B031: repeat usage of generator #1835

Checklist:

• Add tests that demonstrate the correct behavior of the change. Tests should fail without the change.
  • Please note: I think that the best way to confirm that these changes are correct is to use the script to regenerate jinja's Unicode identifier regex pattern
• Add or update relevant docs, in the docs folder and in code.
• Add an entry in CHANGES.rst summarizing the change and linking to the issue.
• Add .. versionchanged:: entries in any relevant code docs.
• Run pre-commit hooks and fix any issues.
• Run pytest and tox, no tests failed.

[davidism]

The check is getting confused by the name reuse, this does not require a huge refactor. Changing b = list(b) to lb = list(b) fixes the warning. 090a5a9

[jayaddison]

Thanks; that's fair reason, yep.

FYI: there is a single-character difference in the output produced by the refactor compared to the script as-is when I run it locally; the existing script emits a trailing minus symbol (range indicator).

That doesn't appear to produce any change-in-behaviour based on my testing in Py3.11 (I was briefly worried that it could allow the - character to appear in identifiers, or to exclude the preceding character from successfully matching, but those don't appear to be the case).