This repository contains demos, examples and for using PyMuPDF in the respective folders.

These scripts were written over an extended period of time - each for the then existing PyMuPDF version. I will not (reliably) go over each of them and ensure they still work. Occasionally a script may therefore no longer be compatible with the current version. If you find such inconsistencies, please do not submit issues, but try to repair the script and submit the corrections via a Pull Request instead. Thank you.

Up to version 1.18.x of PyMupdf a major effort was undertaken to rename (almost) all methods and attributes to the snake_case standard. This task is now finished. For the time being and including versions 1.19.x, old and new names will coexist. For example, the old name doc.newPage() can be used as well as the new name doc.new_page() to create a new page. In versions 1.19.x, a deprecation warning will be issued when old method names are used. In versions thereafter, only the new names will remain being valid. To help migrating your scripts to new names, you may want to use in this folder.

OCR Support

While PyMuPDF does not yet support MuPDF's seamless support of Tesseract OCR, there are nonetheless ways to use OCR tools in PyMuPDF scripts.

There are now two demo examples in the new folder OCR which use Tesseract OCR and easyocr respectively.

Advanced TOC Handling

Handling of table of contents (TOC) has been significantly improved in v1.18.6. I have therefore created another new folder dealing specifically with this subject.

Font Replacement

New for PyMuPDF v1.17.6 is the ability to replace selected fonts in existing PDFs. This is a set of two scripts and their documentation in this folder.

Marking Words and Lines

PyMuPDF's features have been extended in this respect. We therefore created this own folder to contain dedicated scripts, descriptions and examples.

Textbox Extraction

PyMuPDF's features have been extended in this respect. We therefore move example scripts and an extended description to its own folder.

Text Extraction, Layout Preservation

Text extraction scripts have been moved into this separate folder. They demonstrate alternate ways extracting text from general documents (not only PDF), bargaining simplicity versus layout-faithful text output.

The most advanced script, produces text which closeley resembles the original layout of the document, including multi-column text, text thats surrounds images, etc.


From the examples folder, here is a synopsis of some scripts that you may find interesting.

  • a really cool GUI script (wxPython) which allows insertion, deletion and modification of images on PDF pages in a visually controlled mode. Requires v1.17.5 and a Phoenix version of wxPython.

  •,, are scripts showing the effect of the morph parameter in text insertions. Each script creates a PDF page, fills a text box and then morphs that box using its upper left corner as fixed point. Each morphing result is put on a new PDF page and the resulting pixmap is shown in an endless loop. Now require PyMuPDF v1.14.5 and can be run with Python v2.7.

  •, require PyMuPDF v1.14.5 and can be run with Python v2.7. They demonstrate how the modified version of drawOval can be used to create a wide range of shapes, both displaying the results in an endless loop.

  • - take all files from a given directory and attach them to pages (as 'FileAttachment' annotations) of a new PDF. Non-file entries of the directory will be skipped.

  • - take all files from a given directory and embed them (as 'EmbeddedFile' entries) in a new PDF. Non-file entries of the directory will be skipped.

  • - take all image (only) files from a given directory and insert them as pages in a new PDF (actually inserts the first page of any supported non-PDF document). Non-file entries of the directory will be skipped.

  • - scan through a PDF and remove all text of all pages. Also erase metadata / XML metadata. This works by eliminating everything enclosed by the string pairs "BT" and "ET" in the pages' /Contents object(s). Text appearing in images cannot be removed with this script -- this includes images made by the PDF operator syntax: some utilities synthesize text on the basis of elementary drawing commands, i.e. every single letter is created by drawing rectangles, lines and curves. This is probably done as a way to enforce copyright protection. In these cases, the script will not work either.

  • - (requires v1.13.17) extract images from a PDF looping through its pages. It takes care of images with a transparency ("stencil") mask and also tries to identify and skip "irrelevant" images. Suppresses extracting the same image multiple times. You may also want to look at below.

  • - (requires v1.13.17) extract all images from a PDF looping through its cross references numbers. Ignores the page tree and generally seeks to be as robust as possible with respect to damaged PDFs. Otherwise similar to the previous script in ignoring irrelevant images and working with stencil masks.

  • - repeatedly copies a fixed screen area to a PDF page. Can be used to page through an e-book (which might be DRM protected ...) and create a PDF consisting of all its pages in image format - very much like making a full book photo copy. You would start an e-book reader to read a book and then trigger this skript to page through the displayed book making images page by page.

  • - a complete GUI document displaying script using Tkinter! It features a zooming facility and can display all document types supported by MuPDF (i.e. not just PDFs!). It requires PySimpleGUI, an awesome pure Python package exclusively based on Tkinter. For response time improvements, we are also using Pillow.

  • - create an output PDF for a given PDF with text and graphics layout analysis. Each text and graphics block is surrounded by a rectangle (only metadata is shown, no image content). Output PDF has page dimensions of the input's /MediaBox. The input's /CropBox is indicated by a gray background rectangle - see example files in this repo: demo1.pdf and layout-demo1.pdf in folder text-extraction.

  • - (requires v1.13.5) Inspect PDF pages whether any have multiple /Contents objects. If not, exit immediately. If yes, check if any contents are shared between pages. If yes, save the document using the "clean" option and exit. If not, join multiple page contents into one and save with option "garbage=3" (which removes now unused objects).

  • demo script: create a PDF with form fields.

  • contains ca. 10 predefined symbols, which can be imported and used in PDF page creation. If invoked standalone, a PDF is created with one page per each of the symbols.

  • creates a list of symbols and their descriptions, which are implemented in -- thus also demonstrating how to use this module.

  • demonstrates the use of Pixmaps be creating Sierpinski's carpet (a fractal).

  • does the same thing, but this time uses a more intuitive recursive function.

  • the animations folder contains more examples which all display some animation in endless loops. If you have installed PyMuPDF and PySimpleGUI, you can execute any of them rightaway.