Vinyl Goblin

Reads your Discogs wantlist and then finds each release in almost every Brisbane record store. It then saves the results in text files that can be searched using a text editor like VS Code or whatever.

Shops supported:

• Blackened Records (Metal)
• Catalog Music
• Dutch Vinyl (Both Brisbane and Melbourne stores)
• Glitter Records
• Hideous Records (*)
• Jet Black Cat Records
• Monster Robot (EDM, lots of 12" EPs)
• Record Exchange (Online only)
• Relove Oxley
• Rockaway Records
• Rocking Horse Records
• Sonic Sherpa
• Spin and Groove Records
• Stash Records
• Waxx Lyrical

Retro World Collective, Bone Yard Collective and Record Wonderland are not here because they don't have online stores

Butter Beats only lists less than 30 records on their online store at any time

JB Hi Fi is not here because it's honestly too hard for very little reward on over priced records

echo & bounce are not here because there is no non-graphql direct search for releases

(*) Hideous records do not list the artist name on the search results, so their name is artificially injected making the search results a bit fuzzy

Running

There are several releases available for Linux, Windows and MacOS (both ARM64 and x64). You can get them from the releases page.

Before running, you will need to create an app in your discogs account. You can find that here

The application name needs to be VinylGoblin for this to work.

Then you need to generate a new access token. You can use the token in each of the methods below.

You'll need to run each in the terminal/command line. Here are examples for each operating system. Replace with the token created from your discogs account. NOTE THAT THIS CAN TAKE A LONG TIME TO RUN, LIKE HOURS IF YOU HAVE HUNDREDS OF WANTED ITEMS:

Linux

First, make the file executable:

chmod +x vinyl_goblin_linux_x64

Then to run:

vinyl_goblin_linux_x64 <DISCOGS TOKEN>

Windows

This application is compiled using pyinstaller. Windows just loves to think that this is a virus. If you manage to get it downloaded, you can run it like this:

vinyl_goblin_windows.exe <DISCOGS TOKEN>

MacOS ARM64 (Like M1/M2 Macbooks)

First, make the file executable:

chmod +x vinyl_goblin_linux_x64

Then to run:

vinyl_goblin_macos_arm64 <DISCOGS TOKEN>

MacOS x64 (Older models)

First, make the file executable:

chmod +x vinyl_goblin_linux_x64

Then to run:

vinyl_goblin_macos_x64 <DISCOGS TOKEN>

Storing your token

Create a text file called .env and put the following into it, again replacing the token:

DISCOGS_TOKEN=<DISCOGS TOKEN>

Running From this repo

Clone the repo first.

This uses uv for its dependency management and virtual environments. If you don't want to install that, you can install the dependencies listed in the pyproject.toml file to get things running.

To install using uv though, change to the project directory and run:

uv sync

Everything will be installed.

It states that it runs on Python 3.13, but there's nothing really stopping it running on earlier versions.

Running

Rename or copy the env.example to .env and put the access token in the file.

Then, to run the searching session:

python src/vinyl_goblin

And it will start finding your wantlist.

IT TAKES A LONG TIME. LIKE HOURS.

It also gives you a progress bar for each store. Once they're all done, you'll get a bunch of text files in the results directory. These can be searched and will give you the shop and price.

Notes

Some shops are more straight forward to scrape than others. This uses a mix of requests/selenium and Beautiful Soup to get things found.

Restricting the number of shops to use

In the src/vinyl_goblin/shops/__init__.py file there is a shops list. Comment out the lines you don't want and it won't search them.

Adding shops

It's fairly straight forward. Copy an existing shop file in the src/vinyl_goblin/shops directory and rename it to the one you want to scrape from. You'll need to figure out how the shop searches for records with the URL, most have some path like /search?artist+blah+blah. You'll see where that goes in the code. Then once you have a search result page you'll need to figure out how to identify each release on the page, usually

elements with stuff in them. Then find the title and price. Good luck. You can ask questions and submit bugs in the repo.

If you're submitting changes, it'll need to have correct types and also linting passed over it. If you installed using uv, you can do:

uv run ./all

And it'll run all the checks you need. Make a fork, create a branch, push it and create a PR and hopefully it'll get accepted.

In the future I might get around to making tests. Then you'll have to do them as well.