feat(grouping): Add ULID parameterization

[JoshFerge]

Summary

• Adds a new ulid parameterization regex that matches 26-character ULID identifiers
• ULIDs use Crockford's Base32 encoding (0-9, A-H, J-K, M-N, P-T, V-Z), deliberately excluding I, L, O, U to avoid ambiguity
• Pattern is case-insensitive (accepts both upper and lowercase)
• Lookaheads require both a letter and a digit to avoid matching pure-alpha words or pure-numeric strings
• (?!0[xX]) negative lookahead prevents matching 0x-prefixed hex values
• Placed before the hex pattern to match before general hex matching

ULID Format

From the ULID spec:

 01ARZ3NDEK TSVO4RRFFQ69G5FAV
|----------|  |----------------|
 Timestamp       Randomness
 (10 chars)      (16 chars)

Test plan

• Added test for uppercase ULID
• Added test for lowercase ULID
• Added test for ULID embedded in a message
• Added negative test for pure-alpha 26-char Crockford Base32 string
• Added negative test for string containing excluded char I
• Verified existing hex tests still pass (including 24-digit 0x-prefixed hex)
• pytest -svv --reuse-db tests/sentry/grouping/test_parameterization.py -- 204 passed

[JoshFerge]

can remove this