test(idn-hostname): reject zero width space

[vtushar06]

Following the methodology I used for ipv4 and uuid, I read RFC 5892 section 2.6 and found the current idn-hostname.json has no test for a zero width space in a label.

U+200B (ZERO WIDTH SPACE) has the DISALLOWED property under IDNA2008. UTS46 processors treat it as ignorable and silently strip it, so a+U+200B+b becomes ab and is accepted - an invisible character that changes the resolved name.

Changes

• Added 1 test case across draft7, draft2019-09, draft2020-12, and v1.
• a + U+200B + b (a zero width space between two letters) is invalid.

Ecosystem Impact

1. python-jsonschema 4.x: PASSES (rejects it). idna.encode raises InvalidCodepoint on U+200B.
2. libidn2, Go x/net/idna, Node (WHATWG), PHP idn_to_ascii, Rust idna, java.net.IDN, ICU4J, Guava: FAIL (accept it). They strip U+200B as ignorable under UTS46 and validate ab. The silent strip is also GNU libidn2 GitLab issue Create tests for new (upcoming?) meta-schema #136.

RFC References

• RFC 5892 section 2.6 (DISALLOWED category): https://www.rfc-editor.org/rfc/rfc5892#section-2.6

Reproduction commands and the idn-hostname cross-implementation matrix are in my evidence repo: https://github.com/vtushar06/JSON-Schema-format-test-Evidence/blob/main/idn-hostname.md

Related: #965

[Copilot]

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Adds additional conformance coverage for the idn-hostname format by asserting that zero-width space (U+200B) is rejected across multiple JSON Schema draft test suites.

Changes:

• Added a new invalid test case for a\u200bb (U+200B inside an IDN label) in the v1 suite.
• Mirrored the same new invalid test case into draft7, draft2019-09, and draft2020-12 optional format suites.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.

File	Description
tests/v1/format/idn-hostname.json	Adds U+200B “DISALLOWED” invalid-case coverage for idn-hostname.
tests/draft7/optional/format/idn-hostname.json	Mirrors the same U+200B invalid-case coverage for draft7 optional tests.
tests/draft2020-12/optional/format/idn-hostname.json	Mirrors the same U+200B invalid-case coverage for draft2020-12 optional tests.
tests/draft2019-09/optional/format/idn-hostname.json	Mirrors the same U+200B invalid-case coverage for draft2019-09 optional tests.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[jviotti]

I double checked and also looks valid to me.