pdf-inspector
pdf-inspector 源码分析：从 PDF 追踪到 Markdown 证据
用夹具阅读检测器、字体、版面、表格、绑定和 OCR。
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
- Fixtures reveal intermediate contracts.
- Font and layout edge cases deserve focused tests.
- OCR integration should stay optional in unit tests.
Read from public contracts
固定 crate 后用列、CID 字体、表格、链接和图像页夹具检查 PdfType、TextItem、PdfRect、Markdown、警告与多语言绑定一致性。
Use the pinned source evidence, record versions and assumptions, and ask a qualified reviewer to validate the result before adopting it.
Follow a fixture
Use synthetic PDFs covering columns, CID fonts, tables, links, page numbers, and image pages. Compare TextItems, PdfRects, PdfType, Markdown, and warnings at each stage.
Contribute safely
Add parser fixtures, encoding regressions, table edge cases, page-selection tests, and JSON schema checks without requiring PDFium, ONNX, or network access.
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
Pin source and inspect public APIs.
- 2
Create tiny PDFs for each edge case.
- 3
Trace intermediate types and warnings.
- 4
Add a deterministic regression.
可复制示例
fixture PDF -> detector -> PdfType/pages
-> TextItem/PdfRect -> layout/tables
-> Markdown/JSON -> binding parity常见问题
Can benchmark output prove internal call paths?
No. Use source symbols, fixtures, and tests to verify implementation details.
How do I test OCR without large runtimes?
Mock the routing boundary and validate the versioned envelope; run end-to-end OCR separately.
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