fix(core): reset style fingerprint at each row boundary in diff renderer

[ionfwsrijan]

Description

The differential renderer's style optimization (_lastStyleFingerprint) was only reset once per frame flush, not once per row. When _renderDiffLine renders spans on multiple rows, the style state bled across row boundaries because moveTo positions the cursor at a new row where the terminal has no memory of the previous cell's style. This caused sporadic missing ANSI style sequences, leading to visible color/bold/italic corruption on all rows after the first one.

Fix

Moved this._lastStyleFingerprint = null; inside the for (let r = 0; r < rows; r++) loop in _flush() (Renderer.ts:139-143) so the fingerprint is reset before each row is processed. This ensures the first cell on every row always emits the full ANSI reset+style sequence.

Test

Added 'resets style fingerprint at each row boundary in diff renderer' test that creates two rows with identical boundary styles and asserts the ANSI output contains at least two reset sequences (one per row).

Closes #1910

Summary by CodeRabbit

• Bug Fixes
  • Fixed terminal diff rendering so style resets are correctly emitted at row boundaries.
  • Improved consistency when consecutive rows reuse the same styling, reducing missing or suppressed ANSI reset sequences.

[coderabbitai]

📝 Walkthrough

Walkthrough

In Renderer._flush(), _lastStyleFingerprint is moved from before the row loop to inside it, so the style suppression state resets at each row boundary during differential rendering. A new test verifies that ANSI reset sequences are emitted at row transitions when adjacent rows share the same style fingerprint.

Diff Renderer Row Boundary Fix

Layer / File(s)	Summary
Reset style fingerprint per row + test packages/core/src/terminal/Renderer.ts, packages/core/src/terminal/Renderer.test.ts	_lastStyleFingerprint = null is moved inside the for loop in _flush(). A new test renders two rows with matching styles and asserts \x1b[0m appears at least twice in the output.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~5 minutes

Possibly related PRs

• Karanjot786/TermUI#709: Modifies Renderer._flush() diff-rendering behavior around per-row style fingerprints, overlapping directly with this fix.
• Karanjot786/TermUI#954: Changes _lastStyleFingerprint reset and cell-level style fingerprinting in the same renderer logic as this PR.

Suggested labels

type:bug, level:beginner

Suggested reviewers

• Karanjot786

Poem

🐇 A fingerprint clung past its row,
Painting wrong colors in the glow.
One null per line, placed just right,
Resets the style, restores the light—
Each row now fresh, a clean new show! ✨

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The core description is clear, but the template-required Which package(s) and Type of Change sections are missing.	Add the missing template sections, especially Which package(s), Type of Change, and any checklist/notes items you completed.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title is specific, concise, and accurately describes the row-boundary style fingerprint fix.
Linked Issues check	✅ Passed	The code and test directly address #1910 by resetting the fingerprint per row and verifying reset sequences across rows.
Out of Scope Changes check	✅ Passed	The change stays focused on the renderer fix and its regression test, with no unrelated code paths added.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@packages/core/src/terminal/Renderer.test.ts`:
- Around line 251-263: The regression test in Renderer.test.ts is too weak
because the current reset count can still pass when row 1 does not re-apply
styles after its moveTo call. Tighten the assertion around Renderer.renderNow()
by checking for a reset/style sequence immediately after the row-1 cursor
movement, or by requiring the extra reset that only appears when row 1 correctly
reapplies formatting, so the test fails against the buggy Renderer._flush()
behavior.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: c60b33f3-9495-4d02-8de0-4ccdda46cc81

📥 Commits

Reviewing files that changed from the base of the PR and between 85e8d36 and 198115e.

📒 Files selected for processing (2)

• packages/core/src/terminal/Renderer.test.ts
• packages/core/src/terminal/Renderer.ts

[ionfwsrijan]

@Karanjot786 Please review this