Auto Emploi FSM

A small personal project to automate the search for your class in the exams week (if you study in FSM). You can use it at https://fsm.jbara.xyz

How it works

This codes automates logging into FSM Extranet space using your credentials (and ORC on their weak captcha) and retrieves the pdf documents for the time and location of latest day's exams. It then applies OCR on every page of the documents and automates the search for the class you are searching for.

sequenceDiagram
    participant Cronjob
    participant Extrafetcher
    participant Extranet
    participant Refiner
    participant OCR
    participant Database
    participant ImagesFolder

    participant Webapp
    participant User

    loop for every 30 mins
        Cronjob->>+Extrafetcher: Execute every 30 mins
        Extrafetcher->>+Extranet: Scrape PDF files
        Extranet->>-Extrafetcher: PDF files
        loop for each PDF file
            Extrafetcher->>+Refiner: PDF file
            loop for each Page
                Refiner->>+OCR: Feed image to OCR
                OCR->>+Database: Save OCR result
                Refiner->>+ImagesFolder: Save cropped image
            end
        end
    end

    User->>+Webapp: User input their name
    Webapp->>+Database: Find image associated to user 
    Webapp->>+ImagesFolder: Get the image
    Webapp->>+User: Show Salle

Installation

If you wish to use this locally or host it yourself. You can clone the repo and then install the pip requirements by issuing the following command. Note: It is advised you use a python virtual environment for this.

pip install -r requirements.txt

Make sure both Mozilla Firefox and Geckodriver are installed on your system and accessible via PATH as they're both needed by selenium.
Make sure to install tesseract as it is needed by pytesseract.

Note: If you'd like to help port this to use chromium, pull requests are welcome!

Usage

In order to fetch the latest pdf documents and treat them. Note that this will ask you for your login credentials when you use it for the first time. Which they'll be stored in credentials.json

python autodevoir-cli fetch <username> <password>

After that feel free to run the basic web app and test it out

gunicorn webapp:app

In order to update (say, a new pdf document is released upstream) just re-run the python autodevoir-cli fetch

While it's not recommended yet. autodevoir-cli is made to be run in a cronjob to frequently check for updates upstream.

NOTE: Only works on UNIX and UNIX-like Operating systems. If you would like to try this on windows, use WSL or feel free to edit the code (don't forget to PR).

Extra: While this is made for testing purposes. It can be useful to get the filenames that are related from the CLI. (say for an automation script or a custom app you're working on). Run python autodevoir-cli find to use the CLI instead of web.

API

An API is implemented in case you want to use this in whatever you'd like. Just make a POST request to https://jbara.xyz/api with the following content.

{
    "ident": "your name"
}

And you will get a response as an array of images corresponding to your search result as follows.

{
  "result": [
    "https://fsm.jbara.xyz/images/161223-l3-9h.png",
    "https://fsm.jbara.xyz/images/151223-l3-15h30.png"
  ]
}