Batch correction of scanned archives
Short answer: Archive correction fixes typos, old names and old reference numbers in internal archives without changing the layout of the scans. One keyword table processes a whole folder; each file is saved separately without overwriting the others, and a processing report is produced at the end.
Typical tasks
- Typos that recur in internal manuals and policy documents.
- Old names and model numbers in internal material after a department is renamed or a product model changes.
- Search problems caused by errors in the OCR layer at scanning time (replacement also writes correct, searchable text).
These are all internal documents that your organization has the right to change. Signed or issued files and official certificates are not part of this use case; the software handles them according to the risk levels.
How to do it
- Prepare the keyword table: old text in column A, new text in column B.
- Open the batch window, add the whole folder and choose “Replace with current keywords”.
- Recognition language, font, alignment and other settings are taken from the main window. Try the settings on one or two files in the main window first.
- Run. You can stop at any time; processing stops after the current file is finished.
- Open “Processing report_date.xlsx” in the output folder and filter by read-back result to find the files that need a manual check.
Automation
The same task can run from the command line on a server or as a scheduled task:
PDFTextReplacer.exe --folder <input-folder> <output-folder> <keyword-table.xlsx>Exit codes distinguish success, failure, and “written but verification found problems”, so the tool fits into existing workflows. See Command-line mode.
FAQ
Are the original files overwritten?
No. Each file is saved to the output folder with a suffix added to its name. If a file with the same name already exists, a number is added automatically.
After processing hundreds of files, how do I know which ones need a manual check?
The processing report lists, file by file, the number of items processed, the read-back verification result and the risk level. Start with files whose read-back result is "Needs review" or "Failed".
What if the archive contains signed files?
They are skipped by default and noted in the report. If you really need to, tick "Also process signed or issued files"; the output pages then carry an "EDITED COPY" stamp.