Corrupt PDF Repair

1. Drop a damaged PDF here, or click browse file.
2. Review Diagnostic findings for header, EOF, cross-reference, object, trailer, stream, and encryption checks.
3. Inspect Recoverable page preview and note any page tiles that failed to render.
4. Click Download repaired PDF when a rebuild is available, then click Download diagnostic report to save the recorded findings.

What corrupt PDF repair checks

The raw scan checks the %PDF- header, extra bytes before it, the final %%EOF marker, and the startxref pointer. That pointer can lead to a classic cross-reference table or a compressed xref stream.

For a classic table, the tool samples offsets and checks whether they land on matching N G obj headers. It checks /Root and /Size in the trailer, compares /Size with the object headers it finds, and reports a stream without a later endstream.

An /Encrypt entry is reported separately. The page can show findings for an encrypted file, but it refuses to remove or bypass its protection.

How damaged PDF repair rebuilds a file

pdf.js receives the full byte array and tries to open it in recovery mode. It renders each page it can read. For an unencrypted document, pdf-lib loads the objects it can parse and saves a new file with fresh cross-reference data, a trailer, and object offsets. Missing or unreadable content is left out.

Input check
/Encrypt 7 0 R

Report
Encryption: Encrypted
Repair outcome: Not repaired

This report appears when the trailer data contains an /Encrypt entry. Raw checks still appear. pdf.js may preview pages if it opens the document without a password.

When corrupt PDF repair cannot produce a file

The rebuild stops when encryption is detected, the parser cannot load the document, or pdf-lib cannot save the parsed objects. The rebuilt file can also contain a different number of pages than pdf.js reports. Objects the rebuild engine cannot parse are omitted rather than recreated.

The drop zone accepts one document at a time. For files over 80 MB, the deep text scan reads only the first 80 MB. The header check reads the first 1,024 bytes, and the EOF check reads the last 2,048 bytes. Parser recovery and the repair attempt still use the full byte array. The preview is capped at 24 pages, and classic xref checking samples up to 500 entries.

Where corrupt PDF repair comes up

• A PDF opens with a cross-reference or object-offset error. Review the findings, then compare the rebuilt copy with the page preview.
• Only part of a damaged PDF opens. Check the page tiles to see which pages pdf.js rendered and which failed.
• A colleague or records team needs the repair attempt. Download the diagnostic report with the structural findings and parser recovery state.

When to use PDF Repair Tools

Use PDF Repair Tools to compare repair tools in this category. Use Corrupt PDF Repair for a damaged PDF when you need structural findings, page preview, and an attempted rebuild in the same result.

Frequently asked questions