fix IndexError in DateTimeParser.parse with malformed format strings

[Mr-Neutr0n]

Malformed format strings can cause DateTimeParser.parse() to raise an IndexError instead of a ParserMatchError. This happens when the format string contains characters that, after regex escaping and bracket expression handling, produce a token list that doesn't align with the named groups in the generated regex pattern. match.group(token) then blows up with IndexError: no such group.

The fix wraps the match.group() calls in a try/except to catch IndexError and raise ParserMatchError instead, which is the expected exception type for parse failures.

Reproducer from #1191:

from arrow.parser import DateTimeParser
p = DateTimeParser()
p.parse("foo", "[-FFFFFFFFFFFFFFFFFFFFF-[-FFFFFFFFFFFF[ |||||||| 7   v.dG(dG\\][3zasks  &")
# Before: IndexError: no such group
# After: ParserMatchError: Failed to match '...' when parsing 'foo'.

Added a regression test. All 723 existing parser tests still pass.

Fixes #1191

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
⚠️ Please upload report for BASE (master@b247ad0). Learn more about missing BASE report.

Additional details and impacted files

@@            Coverage Diff             @@
##             master     #1243   +/-   ##
==========================================
  Coverage          ?   100.00%           
==========================================
  Files             ?        10           
  Lines             ?      2318           
  Branches          ?       358           
==========================================
  Hits              ?      2318           
  Misses            ?         0           
  Partials          ?         0           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.