The four reasons a PDF will not let you select or copy text, how to tell which one you have, and the right fix for each, including Urdu and Arabic PDFs.
You open a PDF, drag the cursor across a paragraph and nothing is selected. Or you paste what you copied and get a string of strange symbols. This is one of the most common PDF frustrations, and it has a short list of causes. Once you know which one you have, the fix is quick.
Open the file and try each of these:
Crisp letters that you can search for mean the PDF has a real text layer. Blurry squares and no search results mean the page is a picture.
If a paper document was scanned or photographed and saved as a PDF, each page is only an image of the text. A picture of a word is not a word, in the same way a photograph of a book is not an ebook. The viewer has nothing to select.
Fix: use OCR, which reads the letters from the picture and turns them into real text. Drop the file into PDF to Text. When it finds no selectable text, it offers to run on-device OCR on each page. Copy the result or save it as a .txt file, and fix anything the reader misjudged.
Some design programs convert text to outlines (curves) when they export a PDF, to make sure the fonts look the same everywhere. The letters still look like letters, but to the file they are drawings. Posters, flyers, certificates and some brochures are often built this way.
Fix: the same as for a scan. OCR reads what the page looks like, so it does not matter that the letters were once outlines. For decorative fonts, expect to proofread more carefully.
Here the PDF has a real text layer, and you can select it, but the pasted result is garbled: boxes, random letters, missing words or letters in the wrong order. The file has an internal map from font shapes to characters, and that map is missing or wrong. This often happens with custom fonts and with scripts that join letters and run right to left, such as Urdu and Arabic. Pasted Urdu may arrive with every letter separated or the words in reverse order.
Fix: ignore the broken text layer and read the page as a picture. Convert the page to an image with PDF to Image and run it through Image to Text, choosing the correct language, or use OCR in PDF to Text. Our guide on reading Urdu text from images has tips for the best results.
A PDF can carry permission flags that say it should not be copied, printed or edited. Many viewers respect these flags. They are a request from the author rather than a lock on your device, but they exist for a reason. If you need the text, ask the owner for a copy without restrictions or for written permission. Only extract text from documents that you own or are allowed to use.
Whatever the cause, plain text loses the page layout. Columns run into each other, tables become lines of words and page numbers appear in the middle of sentences. Read through once, delete headers and footers that repeat on each page, and fix line breaks that fall in the middle of a sentence. For OCR results, also watch for look-alike mistakes such as the letter O for zero and the letter l for the digit 1; this guide explains why they happen.
No. PDF to Text extracts text and runs OCR inside your browser. The first time you use OCR, the language data is downloaded and cached, but your document stays on your device. If you want to check that for yourself, see how to check that a website does not upload your files.
Try to select a word with your cursor, or press Ctrl+F and search for a word you can see. If nothing highlights, the page is probably a picture of text.
The PDF stores its letters in a way that does not map back to normal characters, which is common with some fonts and with right-to-left scripts such as Urdu and Arabic. OCR reads what you can see instead, and often gives a cleaner result.
Some PDFs carry permission settings from their author that limit copying. If that is the case, ask the owner for an unrestricted copy or for permission. Only extract text from documents you have the right to use.
Yes, but it takes longer, because each page is read in turn. On a phone, a long document may take a few minutes.