What's Changed
Lots of changes and improvements in this release: new capabilities, fidelity fixes, and cross-browser support.
feat
- feat: emulate backdrop-filter at capture time #457
- feat: clip option + snapdom.viewport() — region capture with offscreen culling 328226b
fix
- fix: shadow fidelity in WebKit's SVG rasterization b7e9af2
- Fix rotated-root captures clipped at edges (d31) 64c5a8f
- fix: key the embedded-fonts cache by document identity 05ef9a9
- fix: guard content-sized boxes against sub-pixel width truncation (text re-wrap) 4b92392
- Fix type definitions drift: compress default, quality default, missing options (excludeStyleProps, fontStylesheetDomains, safariWarmupAttempts, debug, filename), toRaw/to on result, LocalFont.stretchPct 7a473c4
- fix: keep root filter blur() and always expand bbox for its bleed 3f6e6e7
- fix: stop compensating the root's stripped translation in the viewBox bbox 3241481
- fix: repair the broken htmlInCanvas re-export in the plugins barrel 1b74a1a
- fix: stop stabilizeLayout from leaving a permanent border on the source element cf3da9b
- fix: guard the container padding offset against Chromium 140 all:initial expansion 0b24929
other (docs)
- FEATURES_CN.md: '按设计跳过' → '有意跳过'. The phrasing fix was independently #453
- docs: refine Chinese documentation dce83ab
- docs: add Chinese translation maintenance guide 4e53063
- Update FEATURES_CN.md to change '按设计跳过' to '有意跳过' 56ff167
other (test)
- Make the test suite pass on Firefox and WebKit 74de348
- Add cross-browser test runs (BROWSER=firefox|webkit|all) with per-engine visual baselines f2b72d8
- test: wait for KaTeX CDN + fonts in the d454 visual demo 8205e1b
other (chore/refactor/merge)
- chore: add cross-engine visual report (npm run report:cross) 5b6e8cd
- chore: update snapdom plugins version ed984ff
- refactor(api): drop snapdom.viewport() — clip: 'viewport' is the one spelling 4a0f301
- Merge pull request #455 from mosuzi/codex/discussion-450-chinese-docs-review 5b0d149
New Contributors
Full Changelog: v2.16.0...v2.18.0