fix(astro): recover island hydration from transient import failures

[0xbejaxer]

Summary

• add astro-island import retry handling with a hash-based retry specifier so transient hydration module import failures can recover
• dispatch a cancelable astro:hydration-error event with error context and keep default logging only when the event is not prevented
• add an E2E regression fixture/test that validates recovery after one failed import and verifies persistent failures emit the event without unhandled page errors

Validation

• Full PR CI matrix is green (build, lint, test, E2E, platform matrix)
• Reviewed issue details and all comments in astro-island: no error event, retry, or recovery when hydration import() fails #16341 (including hash-based retry note)
• Added changeset

Fixes #16341.

[changeset-bot]

🦋 Changeset detected

Latest commit: 2ae3eec

The changes in this PR will be included in the next version bump.

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[codspeed-hq]

Merging this PR will not alter performance

✅ 18 untouched benchmarks

---

_{Comparing 0xbejaxer:fix-16341-hydration-recovery (2ae3eec) with main (15fc3f6)¹}

Footnotes

1. No successful run was found on main (8e0ff3d) during the generation of this report, so 15fc3f6 was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

[0xbejaxer]

Small update now that CI is fully green.

This PR fixes #16341 by handling island hydration import failures in astro-island with:

• one retry path for failed component/renderer dynamic imports
• a cancelable astro:hydration-error event with { error, componentUrl }
• default console logging only when the event is not prevented
• error handling in the directive callback path to avoid unhandled rejection noise

Two follow-up commits were needed during CI hardening:

• f15add9ecf: sync lockfile for the new e2e fixture
• 47bd822e41: stabilize the new retry e2e by waiting for hydration before asserting click behavior

All checks are now passing.

[0xbejaxer]

Hi maintainers — friendly ping on this one whenever you have time.

This PR addresses the hydration import failure recovery path discussed in the issue and includes e2e coverage for both retry-success and persistent-failure event behavior.

If there’s any part you’d like reworked (implementation or test shape), I can update quickly.

[matthewp]

Let's make this just an event, no retries.

[matthewp]

This is a new feature so should be minor.

Also make the changeset a little more in-depth, show a code example of how you might use this.

[matthewp]

please make this a minor change

[github-actions]

This PR is blocked because it contains a minor changeset. A reviewer will merge this at the next release if approved.

Addressed