PDFs get corrupted by interrupted downloads, buggy exporters, or partial disk writes. Often the content is fine and only the file structure (cross-reference table, trailer) is broken. The Rankato Repair tool parses the PDF stream in tolerant mode, recovers as much content as possible, and rewrites a valid PDF from what it found. Success rate is 80%+ for typical structural damage; catastrophic corruption (missing content streams) can't be recovered.
What gets fixed
Common repairs: rebuilding missing or corrupt cross-reference tables (xref), fixing broken trailers, removing dangling object references, correcting invalid stream lengths, and stripping malformed metadata. If pdf.js can parse enough to walk the object graph, the repair succeeds.
What can't be fixed
If content streams themselves are corrupted (unreadable image data, garbled font tables), the tool can't reconstruct them. It'll drop those objects and repair what remains — so a partially-recovered PDF might be missing some pages or content. Better than a totally broken file.
When to use
Try this when: a PDF fails to open in Preview/Acrobat/Chrome, downloads are interrupted, or a file shows warnings like 'Cross-reference table missing' or 'Trailer not found'. If your PDF opens fine but has content issues (garbled text, missing images), repair won't help — those are content-level problems.