These aren't contrived scenarios invented by test authors in total vacuum. They're consequences of the spec's design and reflect real world bugs.
Ok, but you still don’t want to have to change your API to add this
，这一点在夫子中也有详细论述
// We've been tampered with. Refuse to play.
Per-script thresholds would dramatically reduce false positive rates. Treating Mathematical Alphanumeric Symbols with the same urgency as Cyrillic makes no sense when the data shows a 0.145 gap in mean SSIM between them.