Skip to content

YueduCoreText 0.3.0

Choose a tag to compare

@CHANG-JUI-LIN CHANG-JUI-LIN released this 12 Sep 11:49
· 91 commits to main since this release

YueduCoreText 0.3.0 adds a standalone native HTML/CSS layout and rendering engine. Consumers can parse HTML/XHTML and ordered stylesheets, lay out blocks and inline content, paginate incrementally or create continuous documents, draw to CGContext, and query UTF-16 text/link/selection geometry without Yuedu Reader sources.

  • Preserves the existing BrowserLayout float/ruby subset, images, backgrounds, borders, inline decoration and semantic descriptors.
  • Retains the YueduCoreTextTypography product and existing selection, content-metrics and tracing APIs.
  • Declares SwiftSoup 2.13.7; minimum requirements remain iOS 17 and Swift tools 6.0.
  • Includes a compiled public-only iOS consumer example.

Validation uses iOS Simulator 27 with Xcode 27 beta: 128 package regressions, 5 independent consumer tests, and Reader integration checks. Exact geometry fingerprints and representative Browser PNGs match the extraction baseline. GitHub CI on Xcode 16.4 compiles the package but fails the existing Typography assertion punctuationStillCompressesWhenSafe (kern == -20, expected > -20). The same assertion failed in the 0.2.1 baseline CI; this release does not claim a green Xcode 16.4 test run. No real-device performance claim is made.

This release does not provide full CSS/EPUB compatibility, vertical HTML layout, table/flex/grid layout, EPUB container management, Reader UI, storage, or playback. Read README.md and docs/EngineIntegration.md for supported scope, explicit resource loading, coordinates, cancellation and executor ownership.

CI evidence: 0.3.0, 0.2.1 baseline.