Reject empty labels in idn-hostname format

[dngr2]

The idn-hostname format accepts host names with a leading or interior empty label:

".example"   // reported valid — should be invalid
"a..b"       // reported valid — should be invalid

RFC5892.isValid skips empty labels with continue, with a comment about a trailing .. But String.split already drops trailing empty strings, so the skip only ever hits a leading or interior empty label, both of which are invalid. Rejecting them matches the JSON-Schema-Test-Suite (idn-hostname.json: "leading dot" and "empty label between two dots is invalid") and the reference validator. Ordinary host names (and trailing dots) are unaffected; the full suite stays green (added a test).

[Copilot]

Pull request overview

Updates idn-hostname format validation to reject hostnames containing leading or interior empty labels (e.g. .example, a..b), aligning behavior with the JSON-Schema-Test-Suite and reference validators.

Changes:

• Change RFC5892.isValid to fail fast on empty labels produced by splitting hostnames into labels.
• Add a unit test asserting .example and a..b are invalid while ordinary hostnames remain valid.

Reviewed changes

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

File	Description
src/main/java/com/networknt/schema/utils/RFC5892.java	Tightens IDN hostname validation by rejecting empty labels during label iteration.
src/test/java/com/networknt/schema/FormatValidatorTest.java	Adds regression coverage for invalid empty-label idn-hostname inputs and confirms valid hostnames still pass.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[dngr2]

Good catch. A value that is only separators (..., or 。。) splits to an empty array, so the loop was skipped and it fell through to valid. Pushed a guard that rejects when there are no labels — that also subsumes the earlier single-separator case, so I dropped that special check. Added the two inputs as regression tests alongside the existing empty-label ones.

[stevehu]

@copilot resolve the merge conflicts in this pull request

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

[Copilot]

Pull request overview

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

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

Suppressed comments (2)

Previously missed (1) — in code that hasn't changed since the last review.

src/test/java/com/networknt/schema/FormatValidatorTest.java:242

• The new test exercises leading/interior empty labels, but it doesn’t cover the edge case where adjacent separators occur at the end (e.g. "a.."), which can be mishandled if the implementation drops trailing empty split tokens. Also, since the implementation explicitly allows a trailing separator, adding a regression assertion for a trailing-dot hostname (e.g. "example.") would lock in the intended behavior described in the PR.

        // A leading or interior empty label (a leading dot, or two adjacent dots) is invalid.
        assertFalse(schema.validate("\".example\"", InputFormat.JSON,
                ec -> ec.executionConfig(c -> c.formatAssertionsEnabled(true))).isEmpty());
        assertFalse(schema.validate("\"a..b\"", InputFormat.JSON,
                ec -> ec.executionConfig(c -> c.formatAssertionsEnabled(true))).isEmpty());

src/main/java/com/networknt/schema/utils/RFC5892.java:106

• String.split(regex) with the default limit drops all trailing empty strings. That means inputs like "a.." (adjacent separators at the end) will be split to ["a"], so the new label.isEmpty() check never runs and the empty label is not rejected. To reliably detect empty labels while still allowing exactly one trailing separator, preserve trailing empties with split(..., -1) and explicitly ignore only a single final empty label.

        // RFC 5892 calls each segment in a host name a label. They are separated by all the recognized label separators.
        String[] labels = value.split(LABEL_SEPARATOR_REGEX);
        if (labels.length == 0) {
            return false; // a value made up only of label separators (e.g. "." or "...") has no labels
        }