Repository navigation
PDF Reader 0.1.4 improves repeated document reading and native processing reliability.
- Reuse extracted pages across overlapping requests and layout modes with a bounded LRU cache.
- Paginate literal search matches with offsets and continuation information.
- Exclude inherited native-library search overrides from Poppler/Tesseract processes; provide a source-bound dependency evidence collector in the repository.
- Honor cancellation before returning cached OCR; clean temporary OCR input when writing fails; reject malformed PNG headers or invalid render dimensions.
- Test Harness 0.2.1-rc.2 in addition to the existing runtime matrix.
Download dsh-pdf-reader-0.1.4.tgz. The unscoped npm package named dsh-pdf-reader belongs to a different author; use this repository's archive or GitHub source. Poppler and optional Tesseract/language data must be installed separately.
SHA256SUMS.txt and release-manifest.json bind the seven-file archive to the release commit. The exact archive passes 37 tests. All seven CI jobs passed, covering Linux/macOS package tests, four Harness versions and a real disposable-profile install/start/tool-execution/uninstall. Windows native processors, packaged desktop GUIs, update rollback and independent executable provenance remain unverified. DSH Store approval is separate.