reject non-hex characters in decodeHex

[netliomax25-code]

1. decodeHex decodes each two-character group with Integer.parseInt(value.substring(i, i + 2), 16), which accepts a leading + or - sign and non-ASCII Unicode digit characters.
2. so malformed input such as "-1", "+a" or "-f" decodes to a byte (0xff, 0x0a, 0xf1) instead of throwing NumberFormatException, which the method javadoc documents for characters that are not valid hexadecimal values.

Replaced the per-group parseInt with a strict per-nibble decode over the ASCII ranges 0-9/a-f/A-F. Valid hex decodes to the same bytes (checked against the old path for every byte value in both letter cases); the rejected inputs are covered by new cases in HexTest. Happy to file a Jira and link it.

[testlens-app]

✅ All tests passed ✅

🏷️ Commit: 5136de2
▶️ Tests: 101826 executed
⚪️ Checks: 23/23 completed

---

Learn more about TestLens at testlens.app.

[Copilot]

Pull request overview

This PR tightens EncodingGroovyMethods.decodeHex(String) so it rejects non-hex characters that Integer.parseInt(..., 16) would otherwise accept (e.g., leading +/- signs and non-ASCII digit characters), aligning behavior with the method’s documented NumberFormatException contract.

Changes:

• Replaced per-byte Integer.parseInt(substring, 16) decoding with strict per-nibble ASCII-range decoding (0-9, a-f, A-F).
• Added regression tests ensuring signed prefixes and non-ASCII digit characters are rejected.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
src/main/java/org/codehaus/groovy/runtime/EncodingGroovyMethods.java	Implements strict ASCII hex decoding via a new hexToNibble helper to reject +/- and non-ASCII digits.
src/test/groovy/groovy/HexTest.groovy	Adds new negative test cases for sign-prefixed and non-ASCII-digit hex strings.

[paulk-asert]

Merged. Thanks!