pdf-inspector
pdf-inspector 部署：Rust、Python、Node、WASM 与 OCR
选择本地、服务端或浏览器边界并保持 OCR 可选。
你将学会
- Trace detection and layout stages
- Run native extraction and OCR safely
- Design provenance checks
开始前需要
- Basic Git and command-line usage
- Comfort reading a project README
You can explain the project, run its documented first step, and decide what to verify before adopting it.
先看结论
- Deployment choice follows residency and runtime needs.
- OCR dependencies should not burden native-text paths.
- Malformed PDFs and missing runtimes need explicit errors.
Choose the boundary
按数据驻留选择 CLI、绑定或 WASM；只有 OCR worker 安装 PDFium、ONNX Runtime 和模型，并测试畸形文件与缺失依赖。
Use the pinned source evidence, record versions and assumptions, and ask a qualified reviewer to validate the result before adopting it.
Keep OCR optional
Default Rust/browser builds remain pure extraction. Native OCR needs separately installed PDFium, ONNX Runtime, and model files only when pages are routed; document those platform dependencies explicitly.
Operate
Pin crate/npm/PyPI versions, model caches, feature flags, and sample PDFs. Test malformed files, large pages, missing OCR libraries, worker cancellation, and disk cleanup before exposing a service.
Use the pinned source evidence, record versions and assumptions, and ask a qualified reviewer to validate the result before adopting it.
如何选择
| 比较维度 | 方案 A | 方案 B |
|---|---|---|
| Best when | You need predictable behavior and easy auditing | You need adaptive optimization and have reliable telemetry |
| Main risk | May leave performance on the table | Can become difficult to explain or debug |
实施步骤
- 1
Select CLI, binding, or WASM boundary.
- 2
Pin versions and optional OCR runtimes.
- 3
Run smoke and failure fixtures.
- 4
Monitor latency, memory, cache, and cleanup.
可复制示例
cargo install pdf-inspector
# OCR build only when needed
cargo install pdf-inspector --features ocr --bin pdf2md
PDFIUM_LIB_PATH=/path/to/libpdfium ORT_DYLIB_PATH=/path/to/libonnxruntime pdf2md scan.pdf --ocr auto --json常见问题
Can browser WASM call PDFium?
The documented browser path uses the Rust parser and embedded CMaps; native OCR dependencies are a separate boundary.
Should OCR libraries be in every container?
No. Keep them in workers that actually receive routed pages.
资料来源
- pdf-inspector README (captured 2026-08-31)来源核查 2026-08-31
- pdf-inspector repository来源核查 2026-08-31