Installation:

• Set up a postgres DB and configure DATABASE settings in collected/settings.py
• make install to set up local deps (creates a virtualenv for you).
• source .collected/bin/activate to activate the virtualenv.
• make migrate to apply migrations to your database.
• Download the JSON file from this link and put it in the same directory level as this README.md file.
• make loaddb to load up all the cards.
• Take a short walk or read the rest of this README while the database populates from the JSON file.
• make start to run a local server.

Use:

• You have to sign up and I put approximately 0 effort into making that a nice experience.
• http://127.0.0.1:8000/account/register/
• The activation link prints to the terminal--paste it into your browser.
• http://127.0.0.1:8000/account/login/
• Click the collections link and then Create New Global Cards Collection (I think this is the only one that actually works).
• Marvel at the totally unstyled Django form.
• "Name" can be anything -- it'll be the name of the collection in the list view.
• The check boxes indicate if you care about collecting foils &or nonfoils.
• The border drop down is in case you'd only like to collect a certain border.
• The gigantic unsorted list in the tiny box is every card in the database, best to click in and type what you are looking for. The perfect UX.
• Two good cards to take a look at are Hymn to Tourach and Giant Spider in Nonfoil and Foil.
• I should note it is totally possible to create a collection that has no cards in it if you select a border that has no printings, I don't think it gives any warning about this so if the resulting collection looks blank that is almost certainly why.

Where I have tested this and found it working:

• Python 3.8.5 running on Ubuntu 20.04.2 on WSL2 in Windows
• Postgres 12.6

What is this project?

It's a prototype for a website to track a specific type of Magic: the Gathering collection.

Magic is a trading card game with some ~50,000 unique cards. In Magic, a card can have multiple printings in multiple languages over time.

A dedicated niche of collectors will sometimes choose their favorite card and try to collect a "global set" of that card - one copy of each printing in each language for a specific card or subset of cards.

The idea for the project was to see if the MTGJSON project had high enough quality data that I would feel comfortable maintaining a tool which would generate comprehensive global collection lists for any card you wanted to collect, as these can be very large and full of little-known sets that often aren't included in the existing major card search tools.

Unfortunately, the data isn't quite up to snuff--particularly when it comes to language, which is one of the two most critical datapoints for global collecting. I don't have the time to devote to maintaining a list of this stuff, so until some other nerds on the internet do it I've abandoned this project.

That said, I got the project to the point where it will generate lists of most printings and their languages and will present them to you in a sort of ugly HTML table before deciding to drop it. There's no way to mark cards as collected or to exclude them from the generated list, but I believe there is support for it in the template and the databse.

Which files are interesting?:

The bulk of the code exists on the models:

• collection/models.py
• mtg/models.py

This script loads the database from a JSON file provided by https://mtgjson.com/ It does some sanitization and reorganizing of the data to make it fit my purpose here a bit better.

• mtg/management/commands/loaddb.py

Why on earth are there no tests you dunce???:

• I am loathe to submit a code sample w/ no tests but I hadn't tested this when I wrote it as it was just a prototype and I simply don't have the time to devote to testing it now before shippping it your way. Apologies! Please know that I've been woken up by pagerduty/spent enough time on bug duty in my life to value testing the code I write for things that aren't toy weekend projects.