scanned_pdf_sorter

• Introduction
• Prerequisites
  • Programs / Libraries
  • Pip Packages
• Installation & Running
  • Linux
  • Windows
• Building
• Config.ini
• Notes
• Authors

Introduction

The purpose of this project is to sort and group pages of a scanned pdf file by the selected information that was extracted from each page.

Project Links

• Gitlab
• GitKraken Board

Prerequisites

Programs / Libraries

• Python is required if the project is being built or if being ran without compiling

• ODBC Diver 17 for connection and querying an SQL Database

• Poppler is required for the reading and writing of pdf files.

  • Poppler is included in most Linux Distributions.
  • Poppler for Windows: 7z Archive Download

• Tesseract is required for OCR functionality (Optical Character Recognition)

  • Install Tesseract via your package manager
  • Tesseract building instructions for Windows are located in the Notes section

• Poppler and Tesseract are included in wintools.zip

Pip Packages

• All necessary pip packages are listed in setup.py

• In general, here are the packages that are needed for this program to run

  • matplotlib
  • Pillow
  • pdf2image
  • pytesseract
  • pyodbc

Installation & Running

Linux

• Create and Navigate to the directory for this program

• Create and activate a virtual environment:

  • python -m venv venv
  • source ./venv/bin/activate

• pip install via the git repo, replace BRANCH with the desired git repo branch that you want to install

  pip install git+https://www.gitlab.com/cblacktech/scanned_pdf_sorter@BRANCH
  

• To run the program, run this entry_point command in the terminal

  pdf_sorter_app_run
  

Windows

• Create and navigate into a directory for this program

• Create and activate a virtual environment:

  • python -m venv venv
  • .\env\Scripts\activate

• pip install via the git repo, replace BRANCH with the desired git repo branch that you want to install (omit @BRANCH for default branch)

  pip install git+https://www.gitlab.com/cblacktech/scanned_pdf_sorter@BRANCH
  

• Download wintools.zip from repo and move the zip file to the program directory (Do one of the following):

  • curl -LJO https://gitlab.com/cblacktech/scanned_pdf_sorter/-/raw/BRANCH/wintools.zip
    

  • Download the zip file from this the repo. By clicking the wintools.zip file in the repo, then clicking the download button.

• To run the program, run this entry_point command in the terminal

  pdf_sorter_app_run
  

Building

• Follow the Installation & Running section until the pip installation part

• Pip install the packages needed for development / building purposes

  pip install -e .[dev]
  

  (If project is git cloned, or downloaded)

  or

  pip install git+https://www.gitlab.com/cblacktech/scanned_pdf_sorter@BRANCH[dev]
  

  (where you replace BRANCH with the desired git repo branch that you want to install, omit @BRANCH for default branch)

• Run pyinstaller using a spec file, for example

  pyinstaller pdf_sorter_app.spec
  

• The finished program will be located inside the dist folder

  • Run the executable to start the program after you navigate into the program directory inside of dist
    • Linux: ./pdf_sorter_app
    • Windows: pdf_sorter_app.exe

Config.ini

• image_type determines the image file type that is used (currently supports the values png and jpeg).

• file_initial_search_dir determines where the pdf file selector will first open upo at.

• The CROP_BOX stores the top-left coordinates and the bottom-right coordinates what the images will be cropped to.

• Currently all of the other options are for testing and development purposes (it is not recommended for any of these values to be changed at this time)

Notes

• If you want a custom window icon, have a .png file in the same directory that you are launching your application from.

• To reinstall poppler download it from here and extract it into a folder called wintools.

• tesseract.exe was build from source using the build instructions in the tesseract docs, and then was moved inside the wintools folder.

• The tessdata folder contains the data for recognizing english characters and numbers, eng.traindata. The traindata was downloaded from their git repo.

• Due to not having access to Mac hardware for testing purposes therefore, Mac OS is not supported.

Authors

• Caleb Black - cblacktech | GitLab | GitHub