Exports channel content and downloads attachments from Discord guilds.
- Clone the repo into a directory
- Open the solution in Visual Studio (Community 2019 Version 16.7.6 was used) and build the solution.
- (optional) Copy the built application to somewhere with a lot of storage (depending on how big your server is)
- Open
config.json
and configure to your liking. - Ensure the bot is added to the server you want to export channels from.
- Open
DiscordBot.exe
, ensure there are no errors and wait until the gateway is connected before running any commands.
[prefix] update
- Update the channel cache manually for the channel it is ran in.[prefix] export <includeStatistics (true/false)> <channel snowflakes...>
- Update and export the specified channels (or the current channel where none were specified)
The content (messages, embeds and attachments) of channels is cached and changes to messages that are older than the last exported message will not be tracked. It is therefore recommended that you lock the channels you are exporting and hide them from view of regular users to stop messages from being edited/deleted while the process is taking place.
To ensure that a channel caching operation can pick up where it left off, all have their own JSON file for messages. When backing-up multiple channels where many of the author's username's and attachments are shared across them, it makes sense to only store this information once.
Storing the entire URL of an attachment as https://cdn.discordapp.com/attachments/771982307170058270/772001142455533578/vFvNhBa.png
is not space efficient therefore, the URL for a CDN attachment is stored once for each channel in its channel cache and wherever that URL is referenced, it is replaced with the following: !attachment-snowflake/filename.extension
(or for the example above: !772001142455533578/vFvNhBa.png
)
Located at (by default) channelcache/[channel-snowflake].json
, this file stores the constantly updating cache of a channel.
{
"channel-snowflake": 771982307170058270, // The snowflake / ID of the channel which this file pertains to
"newest-snowflake": 772001142938533898, // *usually* the most recent message that has been cached
// A mapping of Discord ID -> Discord Username of all authors who have posted in this channel
"authors": {
"140378625726742528": "Jaymo#1337"
},
// A mapping of attachment ID -> URL of all attachments (with "pointers") in this channel
"attachments": {
"772001142455533578": "https://cdn.discordapp.com/attachments/771982307170058270/772001142455533578/vFvNhBa.png"
},
// A dictionary which stores all messages in this channel where the key is the message ID and value is message data
"messages": {
"772001142938533898": {
"author": 140378625726742528, // The discord ID / snowflake of the author of this message
"content": "", // The text content of this message
"attachments": [ // An array of attachment URLs or "pointers" included in this message
"!772001142455533578/vFvNhBa.png"
],
"embeds": null, // An array of embeds included with this message
"pinned": null // True if this message was pinned in this channel when it was cached
},
"771982360194842645": {
"author": 140378625726742528,
"content": "hi channel 1",
"attachments": null,
"embeds": null,
"pinned": null
}
}
}
To be even more space efficient for an export bundle of multiple channels, the data of all channels in a bundle is collated and organised in a way which takes up the least amount of space possible.
{
// null and default values are ignored on serialization but are displayed in this example for clarity.
// Additionally, these counters are only included when specified as this data is redundant.
// Nice for quickly seeing export statistics though :D
"channelCount": 2,
"messageCount": 2,
"authorCount": 1,
"uniqueAttachmentCount": 1,
// n -> names: Stores the friendly name associated with a Discord snowflake (where applicable)
"n": {
// For a user, the this is their *cached* Discord username + discriminator
"140378625726742528": "Jaymo#1337",
// For a channel, this is the channel name at the time of export
"771982307170058270": "channel-one",
"771982333275275274": "channel-two"
},
// d -> data: stores the channel data in the following format:
// [channel1-snowflake]
// ├────[author1-snowflake]
// │ ├────[message1-snowflake]
// │ ├────[message2-snowflake]
// │ └────[message3-snowflake]
// └────[author2-snowflake]
// ├────[message4-snowflake]
// ├────[message5-snowflake]
// └────[message6-snowflake]
// ...
"d": {
"771982307170058270": {
"140378625726742528": {
"771982360194842645": {
"c": "hi channel 1", // c -> content: for a message with text, this is the message text
"a": null, // a -> attachments: for messages that have attachments, this is an array of URLs OR "attachment pointers"
"e": null, // e -> embeds: for messages which have embeds, this is an array of embed data which is *mostly* kept intact from how the Discord API presents it
"p": null // p -> pinned: is true when this message is pinned in this channel
},
"772001142938533898": {
"c": "",
"a": [
// attachment pointers keep the attachment ID, original file name and extension but point to a file in export directory (or content cache)
"!772001142455533578/vFvNhBa.png"
],
"e": null,
"p": null
}
}
},
"771982333275275274": {
"140378625726742528": {
"771982374534774784": {
"c": "hi channel 2",
"a": null,
"e": null,
"p": null
}
}
},
}
}