[JunaidAhamed-7777]

The case-insensitive IBAN regex currently accepts Unicode case equivalents of ASCII letters. For example, iban("\u0131T60X0542811101000000123456") and iban("GB82WE\u017fT12345698765432") return True: dotless i and long s match [a-z] and become ASCII letters during the checksum calculation.

Add re.ASCII to the format check and regression tests covering the country code and account identifier. Existing lowercase and mixed-case ASCII IBAN support is preserved.

Validation on Python 3.14:

• Both new regression cases failed before the fix.
• All 8 IBAN tests pass after the fix.
• The full local test run passed 1,097 tests; 18 Ethereum-address tests failed because the optional Ethereum hashing dependency is not installed. CI includes that extra.
• git diff --check passed. Ruff and Pyright are left to CI.

AI-assisted using GPT-6; human review follows draft submission.

[nandgator]

Thanks, the fix is correct and I'd like to merge it. CI's lint step (ruff check .) fails on tests/test_iban.py:28, though: E501, the line is 102 characters and the limit is 100. Running ruff format and pushing should fix it.

[JunaidAhamed-7777]

Thanks, Ran Ruff and pushed the formatting fix. Lint, type checks, and all tests pass locally
@nandgator set alle?