[YADC is not under active development any more. If someone would like to take over, contact me.]

YADC: Yet Another (DVSA) cancellations Checker

YADC is yet another checker for driving cancellations in the UK.

why another?: Because all the code I found on github doesn't work with the latest anti-bot protections on dvsa. Besides, it was all rather spaghettified. YADC is, I hope a little cleaner and more modular. It should be easier to modify down the line.
why a bot?: Because I can't get a driving test. Note that I do not think this solution is remotely ideal. See the appeal below.
will you maintain this?: Not if I get a test and pass. But hopefully it should be a better framework for the next generation to come along and do the same.

Features

YADC is:

cleanly written and object-oriented
neat. We use a context manager to handle the browser; I'm pretty proud of that.
easy to extend
easy (hopefully) to get around the next rubbish DVSA put up
pretty immune to blocking (thanks to TOR)
pretty colourful, too. We have coloured logs (thanks to coloredlogs), spinners (thanks to Halo) and well-formatted output. When things go wrong, we save a traceback, a screenshot, and the html the browser was seeing.

YADC is not:

my idea. I started with this repo and extensively rewrote the logic from the ground up. I got the idea of the poisson sleep from this repo, although I didn't take any code directly from anything.
infallible
capable of booking tests for you (thanks to @chrishat34 it will reserve them, though).
always going to work
properly supported

Installation

With pip

YADC is now deployed on PyPi, so you can just

python3 -m pip install YADC

However, since every individual setup is going to vary somewhat, YADC is only a library. You still have to decide how to use it. A starting point is provided in the main.py of this repository: save it somewhere, edit it, and off you go. A dummy CLI is provided which will merely direct you to do this ;)

If anyone wishes to write a proper CLI I will happily merge it.

Note that if you want captcha defeating you need to download buster and unzip it somewhere you can get at from your main.py. Likewise if you want tor, it needs to be installed and executable as the user running your `main.py`.

From git (for development).

clone the repo
run poetry install
run poetry shell
(optional) to download buster and unzip it.
(optional) install tor and ensure it can be run as a user.
edit main.py with your setup.
run python -m yadc.main

FAQ

Chrome/Chromedriver/Tor doesn't start! You need to provide either the complete path to the executable (that's something like "/usr/bin/chromedriver" or r"C:\Program Files\Chrome\chrome.exe "") including any suffix it might have; or you need to make sure that the executables are in your $PATH / %PATH%. Have a look in main.py to see where they should go. Windows users: I'm very sorry your operating system makes this so difficult, it wasn't me who decided to put a space in the path to further confuse things.
Tor doesn't connect to the internet. This is probably a bug in your environment if you have correctly passed the tor executable. If you haven't, do so. Verify that one and only one tor process is running (for some reason multiple tor processes will fail on Windows at least), using Task Manager (Micro$oft Windoze) or pkill (*nix) to stop any which are left over by mistake. If you can't get it working, try manually: start tor with /path/to/tor --SocksPort 48059 and then start chrome with /path/to/chrome --proxy-server="socks4://localhost:48059". If this fails the problem is in your installation: if it succeeds and you can't get it running, open an issue and I'll have a look at it. Note that this code has been confirmed working on both Windows and Linux in the past, but it is not in active development, and it's possible things have broken since.
I don't know where to start! If you can't make sense of the instructions in this readme, open an issue and explain what's confusing. That way I can improve the readme :D
I got this error when it was running! Note that errors when running are perfectly normal. YADC gets detected eventually; sooner or later it will probably stop working entirely. See my appeal to the DVSA below. If you want to open an issue with the error, make sure you include the error dump from the errors dir (the .txt, the .html, and the .png). You are strongly advised to sanitise the files first in case they contain personal data.

Usage

Currently you have to be at the computer to do anything. You will see the browser moving, which should help. If you want to interact with it manually, hit Ctrl-z in the terminal to stop execution of the script, and the browser is yours (note that manual interaction increases the chance of being detected by some anti-bot measures). If the script finds a test it will notify you with the notify function you set (the default is just `print`, so do set something more appropriate! A nice service is `pushover`, though it does have a once-off payment). If the script does find a test it will block

Roadmap / Good first PRs

Work in undetected_browser.py would make it possible to use undetected_chromedriver. A PR implementing this would be welcome.
We have no cli. That's probably not an issue, but it would be trivial to add one, e.g. with click.

Use in other projects

YADC has been written to be reusable. See docs/reusing for pointers.

Appeal to DVSA

The current situation is a mess. Because of the pandemic, there are no tests for months; that is not your fault. So we are all forced either to wait for months, or to book a last-minute cancellation. But we can't do that, because all the companies using bots to find tests snatch them up. So we all use those companies; and then DVSA's website becomes even more useless. There are two easy solutions you could adopt:

Add an auto-booking with queue feature to the website. Once a test is booked you get to specify criteria for a better test, and enter in a queue. When you rise to the top of that queue, you get whatever matches reserved for you. In other words, implement the third-party solution yourself. I would gladly pay a reasonable sum for it. And the third-party firms will go out of business and stop spamming your website.
Add an API, and charge for access. That way the third parties will use the API. You can release tests to the API after x minutes to give humans a chance as well. You can go after anyone who tries to use a bot to get round the limit, and providing your API is reasonable, nobody will mind.

Instead of these, you make the already useless website even harder to use. Bot protection is an uphill battle. So is bot developing. The only people who profit from this are companies like google (via chrome/chromium).

2e0byo/YADC