Search across many PDF files at once
You know a clause, a customer number, or a date is somewhere in a folder of hundreds of PDFs — but not which file. Opening them one by one is unthinkable. Batch search (sometimes called full-folder or indexed search) lets you type once and scan every document at once. This article explains how it works, why it silently fails on some PDFs, and how to make it reliable.
What batch search actually does
A batch search runs your query against every PDF in a set and returns the files and pages that match. Under the hood:
- The reader extracts or reads the text layer of each document.
- It matches your query against that text (ideally with normalisation for case, ligatures, and hyphens — see why Ctrl+F fails).
- It ranks or lists results with file name and page number so you can jump straight there.
Done well, it turns a day of manual hunting into a few seconds.
The hard dependency: a real text layer
Batch search is only as good as the text it can read. It works perfectly on PDFs with a native text layer — documents produced by Word, LaTeX, or a printer driver that embeds text. It fails completely on:
- Scanned PDFs with no text layer (just images). There is nothing to match.
- PDFs with broken or missing
ToUnicodemaps, where the text exists but maps to the wrong characters, producing garbage matches or none. - Text drawn as vector curves, which has no character identity at all.
On these, batch search returns zero results not because the phrase is absent, but because the file is, to the searcher, a picture.
Fixing the gap with OCR
The practical answer for scans is OCR — recognising the text from the page images (see offline OCR on Windows). Once OCR attaches a text layer, batch search can find the document. Two points matter:
- OCR can be local and offline, so confidential folders never leave your machine.
- OCR results are recognition, not ground truth; a poorly scanned page may still misread a word, so verify critical hits.
Scope and performance considerations
Batch search over many files is heavier than searching one open document:
- A large folder takes longer on first pass; some readers build an index to speed later searches.
- Encrypted PDFs need their password before their text can be read — batch tools either skip them or prompt per file.
- Permissions that disallow copying can block text extraction in strict readers, though a local read-only, offline reader can still search what it is permitted to read.
A results-handling checklist
Good batch search does more than say "found 12":
- Shows the file and page for each hit, not just a count.
- Distinguishes a native-text hit from an OCR hit, so you know how much to trust it.
- Lets you open the exact page, not just the file.
- Handles diacritics and case so
Resumeandrésuméboth surface.
A quick reference table
| Document type | Batch search finds it? | What to do |
|---|---|---|
| Native text PDF | Yes | Search directly |
| Scan, no OCR | No | Run offline OCR first |
| Broken ToUnicode | Unreliable | OCR or fix source |
| Text as curves | No | OCR |
How to make batch search reliable
- Confirm the files have a real text layer before blaming the search.
- OCR your scanned archive once, locally, to make it searchable.
- Use a reader that normalises queries (case, ligatures, hyphens) so matches are not missed.
- Prefer tools that show file + page and label OCR hits, so you can trust and verify results.
Privacy and batch search
Searching a folder of PDFs means reading their contents, which raises a privacy question for confidential archives. A cloud-based search tool would upload every file to build its index — unacceptable for contracts, medical records, or internal reports. The safe approach is local, offline search that reads the files on your disk and never transmits them (see offline, read-only reading). For scanned archives, run OCR locally first so the search is both possible and private. The combination — local OCR plus local batch search — lets you find anything in a sensitive folder without that folder ever leaving your machine.
Conclusion
Batch search across PDFs is transformative — until it meets a file with no usable text. It finds your phrase instantly in native-text documents, and stays silent on scans, curve-text, and broken ToUnicode maps. The fix is to OCR scanned archives locally first, then search with a reader that normalises queries and shows file-and-page hits you can verify. For confidential folders, do it all in an offline reader so nothing leaves your disk while you hunt.
FlyPDF does this locally. Read-only by design: OCR, search, annotations and text export all run on your machine, and the source file is never rewritten.
Try FlyPDF free for 30 days