test(email): expand is_idn_email coverage for the address-literal branch

[vtushar06]

I added 39 tests to idn_email_test.cc covering the address-literal branch of is_idn_email. Core already passes all of them, so this is coverage, not a fix.

What is new: the branch had almost nothing behind it. is_general_address_literal had no tests at all, and is_snum's leading-zero behaviour - the thing that separates it from the RFC 3986 dec-octet behind is_ipv4 - was not pinned anywhere, despite being called out in a comment.

• Snum (13 tests): leading zeros in each position, padded zero octets, the 255 maximum, one over, four digits, wrong octet counts, and the dot-placement errors.
• General-address-literal (9 tests): the tag:content form, the shortest valid form, a leading-hyphen tag, a trailing-hyphen tag, empty tag and empty content, and the dcontent exclusions for SP and [.
• IPv6 literals (5 tests): the fully compressed form, an uppercase IPV6: tag, IPv6v4-comp, IPv6v4-full, and the empty payload.
• Literal shape (5 tests): empty, unterminated, doubled brackets, a label after the closing bracket, and a trailing space inside.
• The @ delimiter (2 tests): U+FF20 and U+FE6B, neither of which is the delimiter.
• Local-part code points (4 tests): a C1 control, two noncharacters, and a supplementary private-use code point - all admitted by UTF8-non-ascii with no property filter.
• Quoted-string structure (3 tests): unterminated, text after the closing DQUOTE, and a quoted-pair escaping the DQUOTE.

How I checked the expected values: every input was run against the built is_idn_email and is_email before being written, and each expectation is the RFC verdict rather than the observed one - the two agreed on all 43 candidates I probed. sourcemeta_core_email_unit is 393 passed / 0 failed with these added.

Two of my initial expectations were wrong and Core was right, which is worth recording. user@[-tag:abc] is valid, because Ldh-str = *( ALPHA / DIGIT / "-" ) Let-dig only constrains the last character, unlike sub-domain = Let-dig [Ldh-str]. And user@[IPv4:192.0.2.1] is a syntactically valid General-address-literal, since IPv4 is not a tag RFC 5321 defines. Both are now tests with the reasoning in the comment.

RFC references:

• RFC 5321 section 4.1.3 (Address Literals - Snum, dcontent, General-address-literal, the IPv6 forms): https://www.rfc-editor.org/rfc/rfc5321#section-4.1.3
• RFC 5321 section 4.1.2 (Mailbox, Ldh-str, Quoted-string, quoted-pairSMTP): https://www.rfc-editor.org/rfc/rfc5321#section-4.1.2
• RFC 5234 section 2.3 (ABNF literal strings are case-insensitive, so the IPv6: tag matches in any case): https://www.rfc-editor.org/rfc/rfc5234#section-2.3
• RFC 6532 section 3.1 and RFC 3629 section 4 (UTF8-non-ascii, for the local-part code-point tests): https://www.rfc-editor.org/rfc/rfc6532#section-3.1

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[augmentcode]

🤖 Augment PR Summary

Summary: Expands test coverage for the RFC 5321 address-literal branch in sourcemeta::core::is_idn_email, with matching assertions for is_email where applicable.

Changes:

• Adds a focused Snum (IPv4 address-literal) suite, including leading-zero behavior, range bounds (0–255), digit-count limits, and dot-placement errors.
• Adds General-address-literal tests covering minimal forms, hyphen rules in Standardized-tag, and invalid empty tag/content cases.
• Adds dcontent exclusion tests (e.g., SP and [) within general address literal content.
• Adds IPv6 address-literal tests for fully compressed form, case-insensitive IPv6: tag, and IPv6v4 (comp/full) forms, plus an empty-payload invalid case.
• Adds “shape” negative tests for malformed literals (empty, unterminated, doubled brackets, trailing text/space).
• Adds tests ensuring only ASCII U+0040 acts as the mailbox delimiter (rejects fullwidth/small commercial at code points).
• Adds IDN-local-part UTF-8 edge cases (C1 control, noncharacters, supplementary private-use), valid for is_idn_email but invalid for is_email.
• Adds quoted-string structure tests (unterminated, trailing local-part text, and escaped DQUOTE via quoted-pair).

Notes: This PR is test-only (coverage expansion); expectations are pinned to RFC-defined behavior and were cross-checked against current Core behavior.

_{🤖 Was this summary useful? React with 👍 or 👎}

[augmentcode]

Review completed. No suggestions at this time.

Comment augment review to trigger a new review at any time.

[cubic-dev-ai]

No issues found across 1 file

_{Re-trigger cubic}

[jviotti]

Can you remove section comments?

[jviotti]

Looks good, but there are still lots of section comments!