Try it / Three small examples
A fit. A rejection.
An honest no-match.
Pick an example to see the same matching rules used in the full report. The result comes with the constraints and reasons that produced it.
Documentation-based recommendations. This demo does not open a PDF or execute an extraction tool. For measured outputs, see the separate five-PDF extraction checks.
Choose a situation.
All three examples use Linux, CPU-only hardware, local processing and ruled tables. Nothing runs until you choose one.
Uses only the site's published dataset. No document uploads, outside APIs, model calls or saved inputs.
Documentation-based result
Exact request and full report JSON
Request
Report
Bring your own constraints.
The full matcher shows all source links, configuration limits, license caveats and preflight checks. It never needs the document itself.
Know what this proves.
A successful demo means the reviewed rules produced a result. It does not establish extraction quality, OCR performance or commercial-use rights. “No qualified match” only covers the five configurations in this source snapshot, not every available tool.
The separate extraction checks ran pdfplumber 0.11.9 on five small synthetic PDFs. Those results do not change these source-based recommendations or rankings.