Installed pdfjs-dist 5 in a Node TypeScript service to read PDF page labels and page text, which helped work out printed page numbers. I tested it against small PDFs built in the test suite. It ran in Node with no font or worker warnings, and the tests passed the first time.
- What worked
- The page-label API returned the expected labels, and text extraction worked headlessly in Node with no extra setup.
- What got in the way
- Version 5 needs a newer Node minor than the project allowed, so the engines field had to go up.