Why converting Urdu PDFs to Word is hard
A PDF is made for printing, not for editing. It does not store sentences and paragraphs; it stores instructions such as “draw this shape at this point on the page”. For English that rarely matters, because each letter is a separate shape in left-to-right order. Urdu is different. Nastaliq letters join, change shape by position and are often merged into ligatures, and the line runs from right to left. Many programs that create PDFs record the shapes in the order they are drawn on the page, from left to right, and some record only the shape, not the letter.
That is why most PDF converters fail with Urdu. You get the familiar symptoms:
- words or whole lines reversed, so پاکستان becomes ناتسکاپ;
- broken letters that look right in one program but do not join, cannot be searched and break spell checking, because they are “presentation forms” rather than real Urdu letters;
- Arabic letters in place of Urdu ones (ي instead of ی, ه instead of ہ), which look almost identical but break searching and sorting;
- English words and numbers jumbled inside Urdu sentences.
This converter is built for Urdu. It reads the text with pdf.js, the PDF engine used in Firefox, groups the pieces into lines by their position on the page, and orders each line right to left while keeping English words and numbers left to right. It then runs the text through the same repair steps as our Fix Broken Urdu Text tool: presentation forms become real letters, Arabic look-alikes become the correct Urdu letters, and lines stored backwards are put back into reading order. Finally it writes a real Word document in which every paragraph is marked right-to-left, so the cursor, alignment and punctuation behave as an Urdu writer expects.
Text-based and scanned PDFs
Whether a PDF can be converted depends on how it was made.
| Kind of PDF | How it was made | Result here |
|---|---|---|
| Text-based | Saved or exported from Word, Google Docs, LibreOffice, a website or a phone app | Works: the Urdu text is extracted |
| Printed from a browser | “Print → Save as PDF” in Chrome or Edge | Works, including the repair of broken letters |
| Exported from InPage | InPage’s own PDF export or a PDF printer | Often no usable text, because the fonts do not say which letters they draw |
| Scanned | Photos or scans of printed pages | No text here; use Urdu OCR to read the page images |
A quick test: open the PDF and try to select a single word. If you can highlight words, the PDF has text. If the whole page highlights as one block or nothing highlights, it is a scan.
Worked example
Suppose a web page containing this sentence is saved as a PDF with Chrome’s “Save as PDF”:
میں نے InPage 2000 میں خبر لکھی اور اسے WhatsApp پر بھیج دیا۔
The first box below is the raw text a standard PDF text extractor reads from that file, character for character. The words are in reverse order, most letters are presentation forms, and a space has even moved into the wrong word. The second box is what this tool gives you from the same file:
۔ﺎﯾد ﺞﯿﮭﺑ ﺮپ WhatsApp ےﺳاور ا یﮭکﻟ ﺮﺒﺧ ںﯿﻣ InPage 2000 ےﻧ ںﯿﻣ
میں نے InPage 2000 میں خبر لکھی اور اسے WhatsApp پر بھیج دیا۔
After opening the same PDF here, box 2 shows the sentence exactly as it was written: real Urdu letters, right-to-left order, and “InPage 2000” and “WhatsApp” reading left to right in their correct places. The downloaded Word file contains the sentence as one right-to-left paragraph, set in the font you chose.
Lines in a PDF are wrapped to fit the page. With Join lines that the PDF wrapped into paragraphs turned on, lines that run the full width of the page are joined into flowing paragraphs, while short lines such as poetry, headings and the last line of a paragraph stay as they are. Turn it off to keep every line exactly as it appears in the PDF.
Common problems and fixes
“No text found in this PDF”
The PDF is a scan or was saved as images. Run it through Urdu OCR, which reads the page images on your device: it is very accurate on Naskh-style Urdu and gives a first draft for Nastaliq. Otherwise ask for the original Word or InPage file, or retype it with our Urdu typing tool.
The text is in English letters and symbols instead of Urdu
The PDF’s fonts do not record which Urdu letters they draw. This is common for PDFs exported from InPage and for some old government and newspaper PDFs. The text cannot be recovered reliably from such files. If you have the InPage file, copy from InPage and use the InPage to Unicode converter.
Some words still look wrong
PDFs made by older software sometimes split words in odd places or drop the spaces between them. Correct those words directly in box 2 before downloading, or run the text through the Urdu Spell Checker to find them.
Word shows a plain, non-Nastaliq font
The chosen font is not installed on that computer. Install Jameel Noori Nastaleeq or the free Noto Nastaliq Urdu, or select all the text in Word and choose an Urdu font that is installed.
Who uses this tool
- Students and researchers quoting Urdu books, theses and articles that are only available as PDFs.
- Journalists and editors reusing statements, press releases and notifications sent as PDFs.
- Lawyers and office staff updating Urdu forms, affidavits and letters when only the PDF survives.
- Teachers turning PDF worksheets and past papers into editable documents.
To go the other way and turn your Urdu text into a well-formed PDF, use Urdu Text to PDF. If you need the text in InPage rather than Word, use PDF to InPage.
Private by design
Your PDF never leaves your device. The PDF reader (pdf.js) is downloaded to your browser the first time you open a file, then the reading, repairing and Word file creation all happen locally. Nothing is uploaded, stored or logged, so the tool is suitable for confidential documents.