cordcgi

README

CordCGI

Do one thing, and one thing terrible.

---

I took the worst parts of WebCGI and turned it into a web framework:

Make a directory cgi-bin, add a bunch of programs to that. At the moment, the only meaningful thing you can do with this is reply to messages. But that's OK! Write your hello world in bash:

$ cat > cgi-bin/hello <<EOF
#!/bin/sh
echo "<@${DISCORD_MESSAGE_AUTHOR}> Hello, World!"
EOF
$ chmod +X cgi-bin/hello

Or maybe you like Python:

#!/usr/bin/env python3
import os

user = os.environ['DISCORD_MESSAGE_AUTHOR']
print(f"<@{user}> Hello, World!")

Perl? Why not!

#!/usr/bin/perl
use Env qw(DISCORD_MESSAGE_AUTHOR);

print "<\@$DISCORD_MESSAGE_AUTHOR> Hello, World!";

Go crazy. Go stupid!!!

#!/usr/bin/tcc -run
#include<stdio.h>
#include<stdlib.h>

int
main(void)
{
	printf("<@%s> Hello, World!\n", getenv("DISCORD_MESSAGE_AUTHOR"));
	return 0;
}

This is really a bad way to write software.

Setup

Copy settings.json.example to settings.json. Write your client ID, bot token, and preferred command prefix. Set CgiBin to the path to the directory that calls your code. I recommend setting it to ./example-cgi-bin and trying out all the commands there. Relative paths are relative to the working directory when you run the program.

go get github.com/bwmarrin/discordgo, go run . and bam! Discord bot!

Your CgiBin directory (which I'm assuming is cgi-bin/ for now) would be a collection of programs -- scripts, binary executables, anything -- that would be run for each command. If someone, for example, says !add 10 20 (assuming the prefix is !), and there is an executable in cgi-bin called add, it is run with the argv arguments 10 and 20. One interesting thing to note is that subdirectories kind-of work as subcommands. The command !directory/command is a valid pathname, and if cgi/directory/command exists, that would be executed.

When a program is run, it will be provided the entire message's content into stdin, it is provided message data as environment varaibles, and it will take all stdout as a reply. The environment varables are:

• DISCORD_MESSAGE - the message's ID
• DISCORD_MESSAGE_AUTHOR - the message's author's ID
• DISCORD_MESSAGE_AUTHOR_AVATAR - the author's avatar ID
• DISCORD_MESSAGE_AUTHOR_LOCALE - the author's preferred locale
• DISCORD_MESSAGE_AUTHOR_USERNAME - the author's username
• DISCORD_MESSAGE_CHANNEL - the message's channel's ID
• DISCORD_MESSAGE_GUILD - the message's guild's ID
• DISCORD_MESSAGE_UNIXNANOS - the timestamp of the message, as the nanoseconds since Unix epoch.

Reflection

In retrospect, if the only discord things you need to do is read the message calling commands and reply to them, this would probably be a really good way to prototype it. Once your bot needs to become any more rich, though, this probably falls apart. Overall, I recommend against using this for serious bot development unless you're doing it for laughs.

Documentation

Overview

* DiscordCGI: run commands from some cgi-bin/ directory as commands * * Copyright (c) 2021 Samuel Hunter * Licensed under GNU AGPL3.0-only.