CoverUP is a free software, developed in Python, designed to provide a secure and straightforward method for redacting PDF files and their subsequent optical character recognition (OCR).
This version is fork of the original work at https://github.com/digidigital/CoverUP. Two major changes are OCR of documents and the possibility to specify name of the document along the program name. This enables to open files in a file manager or from a command line.
Name of the actual app to be used is pdfanon on Linux or pdfanon.exe on Windows.
A PDF document, which is opened in pdfanon, is first converted into images and all textual information is removed. Users can then conceal sensitive text passages visible in the images by overlaying them with black or white bars. Finally, when saving the document, the overlayed parts of the images are deleted and the remaining visible text is OCR-ed.
The OCR feature requires that the Tesseract OCR engine is installed on the system.
Installation requires administration rights. If you are not a sudo user, ask the administrator to install the software.
Steps:
- Download content of this repository (either using
gitor the providedzipfile). - Unzip the zip file, if using it.
- Enter the created directory and run the
install.shscript. Thepdfanonscript will be placed in the/usr/local/bindirectory.
Comment 1: do not delete the directory after installation. It contains the code.
Comment 2: If you want to hide the directory by adding . (dot) as the first character of its name. Do that prior to running the script.
Comment 3: The script was developed on a debian-like system and tested on recent Ubuntu (python 3.12) and older Mint (python 2.8). If you want to use it on e. g. a rpm-based system, you need to do some changes in the script (changing apt for rpm and perhaps other).
Installation requires administration rights. If you do not have them, ask the administrator to install the software.
Steps:
- Install
TesseractOCRfrom Github. During installation select the needed languages - Add path to
TeseractOCR(most likelyC:\Program Files\Tesseract-OCR) in thePathsystem variables (a guide for Windows 11). - Download the most recent zip file from a Google Drive folder, extract the
pdfanon.exefile and place it somewhere, where the system can find it, e.g. in theC:\Program Files\Tesseract-OCRdirectory. - If required, set you default OCR language by setting the
COVERUP_OCR_LANGenvironment variable.
- Install Python
- Repeat first two steps of the previous section
- Add path to python programs to the
pathsystem variable (e.g.C:\Users\username\AppData\Local\Programs\Python\Python312\Scripts). If set correctly, thepipprogram should be accessible in thecmdwindow. - In a
cmdwindow installpyinstallerby runningpip install pyinstaller - Download content of this repository (either using
gitor the providedzipfile). - Unzip the zip file, if using it
- In
cmdwindow enter the directory and install dependencies bypip install -r requirements.txt - In the same window run
pyinstaller --add-data "Fonts\MaterialSymbolsOutlined[FILL,GRAD,opsz,wght].ttf":Fonts --onefile --windowed --splash splash.png -n pdfanon CoverUP.py
The `pdfanon.exe` will be placed in the `dist` subdirectory.
Default OCR language is slk (Slovak), which is hardcoded in the CoverUP.py file. Documents in English seem to be OCR-ed correctly as well. On Linux languages installed by the install.sh script are eng, slk, ces, deu and fra, more can be added by the apt program (or similar). On Windows additional languages can be installed by the TesseractOCR installation program.
Currently, there is no possibility to select language from the program itself.
One can specify OCR language by setting the COVERUP_OCR_LANG environment variable, for example by adding the line
export COVERUP_OCR_LANG=deu
to the user's .bashrc file if using the bash shell or performing an equivalent action for other shells or on Windows. Refer to tesseract documentation to get the proper code for your specific language.
In order to set the language temporarily, in bash you can run it by
export COVERUP_OCR_LANG=deu; pdfanon file.pdf
The actually used OCR language and file name are displayed in the program's title bar.