dcdl

This is a Discord bot for downloading attachments from channel messages.

Setup

First, set up a Discord bot application (see here how to do so).

Now, you can go to OAuth2 in the Discord bot application page and go to the sub route URL Generator. There, select the scopes bot and applications.commands. Now, you can copy the Generated Url and pate it into your browser. You might now log in to Discord. After that, select the server you want to have the bot on and invite it.

Using Docker

You can use the provided docker-compose.yml to set up the bot. In the environment configuration, you must set the bot token obtained from the Discord bot application.

You can of course also run the image using the Docker CLI as well.

$ docker run -d \
    --publish "80:80" \
    --name dcdl \
    --env "DCDL_DISCORD_TOKEN=OTEwN..." \
    --env "DCDL_WEBSERVER_PUBLICADDRESS=https://example.com" \
    ghcr.io/zekrotja/dcdl:latest

Using Precompiled Binaries

You can also download precompiled binaries from the actions pipeline.

Then, simply create a configuration file.

./config.yml

discord:
  token: OTEwN...

webserver:
  publicaddress: https://example.com

storage:
  location: ./data

Following, cerate the data directory.

$ mkdir data

After that, start the bot with the following command line.

$ ./dcdl -c config.yml

Usage

Now, go into a text channel and type /collect. Then, the command should pop up.

Here, you can now specify following optional arguments.

• channel: The target channel. This is set to the current channel if not specified.

• limit: The maximum amount of messages to be fetched. This defaults to 0 (equals all messages in the channel) or the globally set limit.

• include-metadata: Whether or not to include a metadata.json in the colelction package containing information about each message containing attachments. This defaults to true.

• include-files: Whether or not download and include the attachment files. You can set this to false if you want to download the files from your system using the metadata.json file. This defaults to true.

Download via metadata.json

You can also download the attachments on your system using the metadata.json file from the archive using the following command.

Therefore, you need curl and jq installed. If you are on windows, use WSL. 😉

$ mkdir files && cat metadata.json | jq -r '.[].attachments[] | [ .archive_filename, .url ] | join(" ")' | xargs -l bash -c 'curl -Lo "files/$0" "$1"'

You can even do stuff like filter by author ID, for example, using the following command.

$ cat metadata.json | jq -r '.[] | select( .author_id == "221905671296253953" )

... and then combine it with the download command to just download attachments sent by that specific user.

$ mkdir files && cat metadata.json | jq -r '.[] | select( .author_id == "221905671296253953" ) | .attachments[] | [ .archive_filename, .url ] | join(" ")' | xargs -l bash -c 'curl -Lo "files/$0" "$1"'

Man, jq is really one of the most useful CLI tool ever created, inst it? 😄