Full documentation for processing from scratch

[annahaensch]

Originally opened by @maminian on Dec. 16, 2021

Modify README.md (the main page one) to indicate precisely which python packages are required for scanning the pdf files and processing them. Ideally they should be named by how one would install them using pip/conda/whatever. If anyone wants to go as far as setting up a Python virtual environment from scratch working towards better reproducibility, that'd be great too.

If you're not familiar with github markdown, it's not that important. If you just get the information written down, we can iterate to polish it later.

At a glance, some of the imports in the Python files in the code/ directory that I don't think (or don't recognize) to be default Python modules (but maybe they are?) are below.

• fuzzywuzz
• usaddress
• pandas
• PIL
• pytesseract
• pdf2image

From @maminian:

Doing a little investigating: pytesseract is only a python wrapper for Tesseract, which one needs to install separately. There's the source code here:

https://github.com/tesseract-ocr/tesseract

which has a section about installation.