- Install (latest) Python 3.x with homebrew (http://brew.sh) if not available
brew install python- Install git for version control
brew install git- Clone these codes from git
git clone http://github.com/karnivor/pdf_report_parser.git- Install required tools for to run the code
pip3 install -r requirements.txt- Run the code
python3 run_parser.py /path/to/raw/pdfs/ /where/to/write/results_table.xlsx- If you need to update code
git pulland repeat 5.
Currently I have no access to a Windows machine, so I cannot test or determine what is the optimal way of running Python in windows. Popular options probably include Anaconda https://www.anaconda.com/ or running Windows Subsystem For Linux 2 (WSL2) https://docs.microsoft.com/en-us/windows/wsl/install-win10 with Ubuntu (or other) Linux.
For installing Python3 (and pip) to Linux you can use follow a tutorial like (https://linuxize.com/post/how-to-install-python-3-9-on-ubuntu-20-04/)
- The run_parser.py command reads the list of texts in settings.py to search for similar texts in reports. You may edit settings.py for more texts to be searched. It uses two lists. Some reports have mention "When compared with ECG of XX-AAA-XXXX". The first list is used to compare terms pre-mention and the second list post-mention. By default these are the same. The Excel header texts of post-mention are added with "_v" for the Finnish word "vertailu".
Improve the phrase recognition/comparison part.