PDF redaction checker for version comparisons
Compare two releases against your sensitive-string list and find text that becomes extractable again.
Your inputs stay in this tab. Processing runs locally; nothing is uploaded or saved to browser storage.
Nothing processed yet.
Exact outgoing report
No output yet.
How to check PDF redaction across two versions
Use this comparison when you have a previous PDF and a new release and want to check whether specific sensitive text has become extractable again. It checks the text you name; it does not discover every possible secret or redact the document for you.
- Choose the previous PDF and the new PDF. Each must be searchable, unencrypted, at most 20 MB and no more than 200 pages.
- Enter the names, account identifiers or other strings to check, one per line. Matching is case-sensitive; repeated whitespace is collapsed.
- Select Check regression and inspect current matches, newly extractable terms, additional occurrences and page references.
- Review the coverage limits and download the report if needed. It refers to term numbers instead of exporting the sensitive strings.
What does a newly extractable term mean?
If a listed term has no extracted matches in the previous file but appears in the new one, it is flagged as newly extractable. A term already present in both files is still listed in the match results. Page moves can result from pagination changes and do not alone establish a redaction failure.
Can this check scanned PDFs or prove a file is safe?
No. There is no OCR or image inspection. A blank text extraction is not evidence that a page contains no sensitive information. The checker also reports metadata-field changes and attachment counts or name changes, but it does not inspect attachment contents, annotations, forms or old revisions.
After using Edit PDF or Merge PDF, you can compare a searchable output with its previous release. A clean comparison is only one part of reviewing a document before publication.
What this does not prove
Searchable, unencrypted PDFs only, at most 200 pages each. This checks extracted page text, metadata changes and attachment names/counts, not attachment contents, annotations, forms, OCR, pixels, previous revisions or every hidden object. Fonts and text ordering can defeat extraction. A negative result never certifies a PDF as safe to publish. Downloaded reports use term numbers, not sensitive strings.