Cannot select or copy text in a PDF: the 3 causes and their fixes

The 1-minute diagnosis

Drop the PDF on Inspect above and look at 2 boxes, Pages and Security. These are the readings from 3 of our test files:

What Inspect shows	Scanned lease	Restricted invoice	Invoice with a broken font
Pages with text	0 of 1 checked	3 of 3 checked	1 of 1 checked
Scanned pages (image, no text)	1	0	0
Fonts	none (no text)	Helvetica, not embedded	LiberationSans, embedded
Encrypted	no	yes, revision 6	no
Copy allowed	yes	no	yes
The cause	no text to copy	copying restricted	font without a map
The fix	OCR PDF	Unlock PDF	OCR PDF, reading again

The third column is the tricky one: everything looks normal, because the problem is inside the font. You only find it by pasting a sentence.

Cause 1: it is a scan, so there is no text

A scanner, a phone scanning app or a fax makes a PDF whose pages are photographs. You see letters; the file holds pixels. Nothing selects, or the selection is 1 box around the whole page, and search finds nothing.

You can spot it without any tool: zoom in to 400 %. Real text stays sharp at any zoom, because it is drawn from the font; scanned letters go soft and blocky, and the paper often shows grain or a slight tilt.

Inspect says it in so many words: “Pages with text: 0 of 1 checked” and “Scanned pages (image, no text): 1”. On a mixed file, a typed contract with a scanned signature page, it counts them: our 4-page test file showed “3 of 4 checked” and 1 scanned page. To learn which page, run it through PDF to text: its warning names it (“has no text on page 4”).

A rarer cousin: pages whose letters were turned into drawn shapes, as design programs do when text is “converted to outlines”. Those pages show 0 pages with text and also 0 scanned pages, because there is no picture either. The fix is the same.

The fix: OCR PDF. Pick the language, press Make it searchable, and it adds an invisible text layer where each word sits, leaving the page looking exactly as before. Then text selects, searches and copies. OCR is a reading, not a certainty; it reports its confidence per page, and on a scanned bank statement we measured 94 %. The guide to making scans searchable has the details.

Cause 2: copying is restricted

The file opens without a password, the text may even select, but Copy is greyed out or pastes nothing. Whoever made the file set an owner password and switched off copying. That is a permission flag, stored in the file’s encryption settings, that asks viewers to block copying; Inspect shows it as “Copy allowed: no” next to “Encrypted: yes”. Chrome’s built-in viewer obeys the flag; Firefox’s ignores it by default, which is why the same file can copy in one browser and not the other. The guide to print and copy blocks shows where that comes from, and user and owner passwords explains the 2 kinds of password.

The fix, for files you have the right to use:

1. Open Unlock PDF and drop the file.
2. If it opens without a password, there is nothing to type. Press Unlock.
3. Download the copy. The result says “Removed the owner restrictions from 1 file” and “Pages, text and images untouched”.

A file that needs a password to open is different: Unlock PDF asks for that password and cannot work without it. It does not guess or crack passwords.

Worth knowing: getPDF’s own converters read the text of a copy-restricted file without unlocking it. On our restricted test invoice, Inspect said “Copy allowed: no” and PDF to text still returned every line, because the engine does not apply the flag. The same rule goes with it as with Unlock.

The honest part about permissions. A copy restriction is a request, not protection: anyone who can open the file can read the text, and some viewers, Firefox’s built-in one among them, ignore the flag. But it is also the author saying what they agreed to. Your own files, your employer’s files you work with, a form you need to fill in: unlock them. A bought e-book, a licensed standard or a paid report: the restriction may be part of what you agreed to, and lifting it is between you and the publisher. This explains the technology; it is not legal advice.

Cause 3: it copies, but pastes as nonsense

The text selects and copies, but what lands in your document is RIILFH WKH OHDVH instead of “office the lease”, or boxes, or the wrong alphabet. The font inside the PDF has no correct map from its drawn shapes back to letters, so what you copy are the shapes’ internal numbers. In our test file, “Łódź office: the lease starts on 1 March.” pasted as ăµGļRIILFHWKHOHDVHVWDUWVRQ0DUFK.

Inspect cannot see this: the font shows as embedded and the page as having text, both true. The fix is OCR with Pages with text set to “read again too”; on our test file that read the line back at 84 % confidence. The guide to jumbled PDF text covers this case and its relatives, wrong order and missing fi and fl, in full.

Still stuck?

• The text selects but in odd pieces or the wrong order: the file stores its text out of reading order; see the jumbled text guide.
• It pastes, but with a break after every line: see the line break fixes.
• You need the whole text, not a selection: PDF to text takes all of it at once, page by page, and names any page with no text.

All of these tools run in your browser tab. The file is not uploaded, which matters for the kind of files people most often cannot copy from: contracts, statements and official letters.