fix(engine): widen VFR test frame count tolerance for cross-platform FFmpeg

FFmpeg's VFR-to-CFR normalization produces slightly different frame
counts across versions due to timestamp rounding in the fps filter.
The ±1 tolerance was too tight for Linux FFmpeg builds. Widen to ±3
frames — still catches the 25% shortfall regression these tests
guard against.
This commit is contained in:
Miguel Ángel
2026-05-22 23:42:49 -04:00
parent a54953b936
commit e87f5bb769
@@ -395,8 +395,10 @@ describe.skipIf(!HAS_FFMPEG)("extractAllVideoFrames on a VFR source", () => {
expect(result.extracted).toHaveLength(1);
const frames = readdirSync(join(outputDir, "v1")).filter((f) => f.endsWith(".jpg"));
// Pre-fix behavior produced ~90 frames (a 25% shortfall).
expect(frames.length).toBeGreaterThanOrEqual(119);
expect(frames.length).toBeLessThanOrEqual(121);
// ±3 tolerance: FFmpeg's VFR→CFR normalization yields slightly different
// frame counts across versions (timestamp rounding in the fps filter).
expect(frames.length).toBeGreaterThanOrEqual(117);
expect(frames.length).toBeLessThanOrEqual(123);
expect(result.phaseBreakdown).toBeDefined();
expect(result.phaseBreakdown.extractMs).toBeGreaterThan(0);
@@ -656,8 +658,9 @@ describe.skipIf(!HAS_FFMPEG)("extractAllVideoFrames on a VFR source", () => {
const frames = readdirSync(frameDir)
.filter((f) => f.endsWith(".jpg"))
.sort();
expect(frames.length).toBeGreaterThanOrEqual(299);
expect(frames.length).toBeLessThanOrEqual(301);
// ±3 tolerance: same FFmpeg VFR→CFR rounding variance as the mid-segment test.
expect(frames.length).toBeGreaterThanOrEqual(297);
expect(frames.length).toBeLessThanOrEqual(303);
let prevHash: string | null = null;
let duplicates = 0;