Bot functionalities include:
- Music playback
- Animated bot emoji handling
- Image storage and text optical character recognition (OCR)
The problem which inspired image storage and text OCR is quick identification and resending of an already sent meme; a short, humoristic expression of ideas and viewpoints of certain individuals/groups.
In this simplified case it would be an image commonly consisting of text + some sort of image meme template.
Animated bot emoji handling was intended as a poor man's alternative to using Discord Nitro.
Official discord.py documentation.
Install the modules:
$ python3 -m pip install -U discord.py[voice] youtube_dl pandas aiosqlite pytesseract Pillow
Install the following dependencies for discord.py[voice] (Debian based systems):
$ sudo apt install libffi-dev libnacl-dev python3-dev
Install the following dependency for the YTDLSource class:
$ sudo apt install ffmpeg
Install the following dependency for the pytesseract module:
$ sudo apt install tesseract-ocr
More about discord.py[voice] can be found on the discordpy install guide.
Each bot has a token which acts as its "key".
The file token.txt should consist of a single line; the bot token.
The following music playback commands are at the users disposal:
!stream url/search_query- plays a youtube url/search query result from a stream!ytd url/search_query- plays a youtube url/search query result from a downloaded audio file, the audio file is permanently downloaded to themedia/folder!queue- lists all the songs currently in the queue (playing+queued)!volume new_volume_integer- sets new audio volume, divides by 100!join- makes the bot join the users voice channel!leave- makes the bot leave the users voice channel!skip- skips current song!pause- pauses current song!resume- resumes current song
Music queueing was implemented through the usage of the asyncio.Queue and asyncio.Event classes.
Event loop blocking is avoided by running the blocking functions in a default loop executor. Example: awaitable loop.run_in_executor(...)
The following animated bot emoji handler commands are at the users disposal:
!areact emoji_name message_id- a(nimated)react reacts with an animated emoji to the specified message id!atag emoji_name user- a(nimated)tag sends a message in the form ofauthor: @user animated_emoji
The following image storage and text OCR command is at the users disposal:
!search_meme substring- returns all the memes containing the given substring (case insensitive)
All the images sent by the users are saved locally by the bot. Each image has its text OCRd.
The image id (which corresponds to the filename, converted to .jpg), as well as its text are then inserted as a single row into a database table for future fetching.
A showcase of some of the bots functionalities.
Click on the gif for a youtube video with sound.
