My PDF merge job is corrupting files, what should I check?

If a PDF merge job is producing corrupted files, the usual causes are in the merge tool, source PDFs, or file handling. Here’s a practical checklist to narrow it down:

1. Verify the source PDFs

• Open each input PDF individually to confirm it’s valid.
• Check whether the corruption happens only with certain files.
• Look for:
  • encrypted/password-protected PDFs
  • malformed PDFs
  • PDFs generated by scanners or weird printers
  • incremental-save PDFs or partially downloaded files

2. Check the merge library/tool

• Make sure you’re using a stable, current version.
• Common issues happen when mixing PDFs from different generators.
• If using a library, confirm you’re calling the correct API for:
  • appending pages
  • preserving metadata/bookmarks
  • flattening forms if needed

3. Inspect file I/O handling

• Ensure all files are fully written before merging.
• Make sure you’re not:
  • reading and writing the same file simultaneously
  • merging from a stream that has already been consumed
  • truncating output because the process exits too early
• Confirm binary mode is used when required.

4. Check concurrency and temp files

• If multiple jobs run at once, verify they don’t share temp filenames.
• Look for race conditions with:
  • temporary PDFs
  • output paths
  • cleanup jobs deleting files too early

5. Validate output structure

• Try opening the merged file in multiple viewers.
• Use a PDF validator if available.
• Compare file size to expected size; a suspiciously tiny output often means an incomplete write.

6. Watch for memory issues

• Very large PDFs can fail if the merge job runs out of memory.
• Check logs for:
  • OOM kills
  • timeouts
  • process restarts
  • partial output generation

7. Handle unsupported content

Some PDFs include things that merge tools mishandle:

• embedded files
• annotations
• forms/XFA
• transparency layers
• unusual fonts
• object streams / linearized PDFs

8. Check permissions and filesystem problems

• Ensure the process can write the output path.
• Watch for:
  • disk full
  • network filesystem glitches
  • antivirus locking files
  • permission changes during the job

9. Reproduce with a minimal case

• Merge only 2 PDFs: one known-good, one suspected.
• If the issue disappears, the problem may be a specific file or size threshold.
• If it only fails with many files, suspect resource limits or ordering bugs.

10. Review logs and exceptions carefully

• Look for warnings that are easy to miss.
• Some libraries “succeed” while silently skipping broken pages or objects.

If you want, I can help you debug it more directly if you tell me:

• what language/library you’re using,
• how you merge the PDFs,
• and what “corrupted” means exactly (won’t open, missing pages, blank pages, file size zero, etc.).