[saseungmin]

Multiline lineReveal() animations could show fragments of the previous or next line inside another line's reveal window. Each mask translated a full paragraph copy, so its neighboring source lines could enter the visible band during positive or negative reveal offsets.

Crop the intended measured line in a fixed-height, non-collapsible viewport before applying reveal translation. The existing outer frame still owns opacity, scale, pulse, and transform origin; paragraph shaping, accessibility, playback, and public APIs remain unchanged. This adds one native clipping View per rendered line.

Regression coverage

Seven tests follow glyph-band intervals through the actual rendered ancestor tree and clipping styles. They cover positive/negative offsets, large offsets, fractional line geometry, final visibility, and scale/pulse composition. They do not depend on a new wrapper test ID.

• Before the fix: 5 failed, 2 passed, reproducing neighboring-band leaks of 12–21 layout units.
• After the fix: all 7 pass.

Validation

• Full Jest suite: 16 suites, 233 tests pass.
• Typecheck, lint, format check, library build, and declaration-surface check pass.
• Actual iPhone 17 / iOS 26.2 simulator capture: intermediate reveal frames no longer show the duplicated neighboring line; final text remains intact.
• Android 16 / API 36, Pixel 7 ARM64 emulator (1080×2400, 420 dpi), actual Fabric/Hermes native example: 32 frozen captures cover four fixtures × four progress values × two themes, with four recipes per capture (128 recipe states), plus two recorded timing replays.
  • Positive/negative/large reveal offsets and reveal + scale/pulse composition; English, fractional line height/blank lines, Korean/emoji, native wrapping/centered alignment.
  • The original renderer reproduces neighboring-line bleed under the same Android conditions; the fixed renderer excludes those fragments in inspected frames. At 20%, the +48 source band goes from 12,038 non-background pixels to zero. All 0% regions are empty; completed English preview pixels match before/after exactly.
• Independent code and Android capture review found no actionable issues.

Documentation, interactive demos, and refreshed native recordings are maintained separately in #4. This PR contains only the renderer correction and its regression tests. Android rasterization was checked on the emulator above. Physical Android/iOS devices and other Android OS versions remain unverified. Temporary verification harness and source swaps were restored; this validation adds no source changes.

[changeset-bot]

⚠️ No Changeset found

Latest commit: a0188e9

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

Click here to learn what changesets are, and how to add one.

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

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-31T05:03:28.576574Z	a0188e9	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.