Handle negative numbers in ordinal() (#192)

[apoorvdarshan]

Addresses the negative-number cases of #192.

Problem

ordinal() mishandles negative numbers:

>>> p = inflect.engine()
>>> p.ordinal(-1)      # TypeError: expected string or bytes-like object, got 'int'
>>> p.ordinal("-1")    # '-1th'   (should be '-1st')
>>> p.ordinal("-21")   # '-21th'  (should be '-21st')

ordinal() uses DIGIT.match(str(num)) to decide whether the input is numeric, but DIGIT (\d) does not match a leading -. So negative numbers fall through to _sub_ord, which is meant for word input (e.g. "one" → "first"). For a negative string it just appends "th"; for a negative int it runs re.sub on the integer and raises TypeError.

Fix

• Accept an optional leading - in the numeric branch.
• Base the suffix on abs(n) — -1 % 10 is 9 in Python, which would give the wrong suffix.

>>> p.ordinal(-1)      # '-1st'
>>> p.ordinal(-2)      # '-2nd'
>>> p.ordinal(-113)    # '-113th'
>>> p.ordinal("-21")   # '-21st'

Positive numbers, floats, and word inputs are unchanged (e.g. ordinal(1) → '1st', ordinal(1.5) → '1.5th', ordinal("one") → 'first').

This addresses the negative-number cases raised in #192. I left the ordinal(1.0) → '1.0st' question alone, since whether that should be '1st', '1.0th', or unchanged is a design call for you rather than a clear defect.

Testing

• Added test_negative_ordinals to tests/test_inflections.py (ints and strings, including the teens which correctly keep "th"). It raises TypeError on main and passes with this change.
• Full test suite: 215 passed, 16 xfailed — no regressions. Doctests pass. ruff format clean and the change introduces no new ruff check findings.
• Added a newsfragments/192.bugfix.rst.

---

Disclosure: this change was prepared with the assistance of an AI tool (Claude Code). I reproduced the issue, implemented and verified the fix and tests, ran the suite and linters, and take responsibility for the contribution and will respond to review feedback personally.