An asynchronous API wrapper for Hummus by Ziad87
Project description
Hummus.py
This is an asynchronous (and decorationless) wrapper currently in version 0.6.4!
Installation
You can run pip install hummus2016.py
in shell to install Hummus.py as a package (you can also use the mirror, hmus
), or you can import it manually by downloading the files at the GitLab repository (might be outdated)
Getting started
You can use the following code in the main file to connect to Hummus:
import hummus
from hummus import Client
import asyncio
class Commands(Client):
async def test(self,ctx:hummus.Message):
await ctx.reply(f"<@{ctx.author.id}> activated test!")
Client = Commands(prefix="!",token="INSERT TOKEN HERE", status="online", game="!test")
asyncio.run(Client.RUN())
Adding new commands is as simple as creating new functions under the Commands
class. No need for decorators, just because. Adding custom arguments is as simple as adding parameters in a function, as seen below (remember: In every function, you must have the self
and ctx
args!):
class Commands(Client):
async def test(self,ctx:hummus.Message,test=None): #extra "test" arg
print(test)
print(ctx.content)
await ctx.reply(f"{ctx.author.mention} activated test! Arg 1: {test}") #you can also do "<@{ctx.author.id}>" if you want
You can make these parameters have annotations (test: str
or mention: hummus.User
) or have default values (test=None
).
Arguments
There are 2 ways to have arguments in a command:
- Add parameters to your function
- Use
.split(" ")
to split words in a command into different items on a list
Custom arguments in a function will function differently than using .split(" ")
. The argument system looks for quotation marks in a message, and if there is text within quotation marks, no matter if there are spaces, the entire text (within the quotations!) will be considered as a single argument. This allows for easier usage of commands like !nickname
where you can specify a nickname with spaces, as long as the nickname is within quotation marks.
Quotation marks are not necessary for arguments with no spaces!
Usage
Commands
Hummus.py is in development, so it will not have all the functions existing. Also, please note that the Hummus API itself is unfinished and is missing endpoints, so I cannot have every function either. If you're know about Rotisserie (natsu's site), know that I will not be updating Hummus.py to include things that you can do only on Rotisserie.
Here's an example on how to use the ctx.getUser()
function where you can get any user's avatar:
Without parameter and annotation:
class Commands(Client):
async def avatar(self,ctx:hummus.Message):
if len(ctx.mentions) > 0:
member = ctx.getUser(ctx.mentions[0].id) #this is a bad use case because you can easily use the example shown below, however im too lazy to think of anything else to put here
await ctx.reply(member.avatar.url)
else:
await ctx.reply(ctx.author.avatar.url)
With parameter and annotation:
class Commands(Client):
async def avatar(self,ctx:hummus.Message,mention:hummus.User=None):
if mention: #"None" is equivalent to "False"
await ctx.reply(mention.avatar.url)
else:
await ctx.reply(ctx.author.avatar.url)
As you can see, the above code fetches a member based on the first mention that is in the recieved command, and uses the Member object to get their avatar url.
Moderation Commands (or commands requiring permissions)
You can get access to Hummus.py's permissions checking feature with from hummus.funcs import fullPermsCheck
. Its arguments are the Message
object you recieve on command, your class instance aka self
, the permission you want to check, and optionally, the target user ID to compare permissions with. What the fullPermsCheck()
function does is that it checks to see if the user has the required permission, and it also checks whether it has a higher role than the target ID when specified (permissions comparing). You would use the function like this:
perms = await fullPermsCheck(ctx,self,"kick_members",ctx.mentions[0]) #assume this is in a command function
The code would return a bool
object which you could use to verify that the user executing the command has the required permissions.
So, for commands that require permissions such as kicking members, the optimal code would be the following:
class Commands(Client):
async def kick(self,ctx:hummus.Message,mention:hummus.User=None):
if not mention:
await ctx.reply("Please mention a user to kick them.")
return
perms = await fullPermsCheck(ctx,self,"kick_members",mention)
member = await ctx.getUser(mention)
if perms:
e = await member.kick()
if e.staus_code == 200 or e.status_code == 204: #i dont remember which status code it is lol
await ctx.reply(f"i have kicked <@{member.id}>!")
else:
await ctx.reply(f"Error kicking user. Status code: {e.status_code}")
else:
await ctx.reply(f"You do not have the perms to kick {member.user.username}!")
Existing moderation commands you can use:
member.kick()
member.nick()
member.ban()
member.unban()
member.setRoles()
member.addRoles()
member.removeRoles()
ctx.deleteMessage()
ctx.bulkDelete()
message.delete()
(assuming you are not the author of theMessage
object here)
Hummus's API is very unfinished, which means certain functions cannot be performed or workarounds have to be used. Also, because I am lazy and don't know how Hummus/Discord's permissions integers works, Hummus.py uses Discord.py as a dependency (it has a needed permissions function). However, you do not need to worry about using it somewhere, you just need to have it installed.
Editing Channel Permissions
You can modify channel permissions like so. The code below adds "mention everyone" permissions to a role.
async def pingEveryone(self,ctx:hummus.Message,role:hummus.Role):
overwrite = PermOverwrites(role)
overwrite.allow.mention_everyone = True
e = await ctx.channel.changePerms(role,overwrite)
print(e)
await ctx.send("@everyone")
Note: Only using the above code is not recommended, it does not have permissions checking. See the section directly above this section on permissions checking.
Editing User Roles
Here is a simple script that adds or removes (depending on whether the user you're pinging has the role or not) the role you mention to/from the user you ping.
class Commands(Client):
async def role(self,ctx:hummus.Message,mention:hummus.User=None,role:hummus.Role=None):
perms = await fullPermsCheck(ctx,self,"manage_roles",mention)
target = await ctx.getGuildUser(mention.id)
if role.id in target.roles:
await target.removeRoles([role])
await ctx.reply(f"Removed <@&{role.id}> from {target.mention}")
else:
await target.addRoles([role])
await ctx.reply(f"Added <@&{role.id}> to {target.mention}")
It is important to note here that you must pass a list as the argument for mention.addRoles()
, mention.removeRoles()
, and mention.setRoles()
.
Events
You can execute code on gateway events, as demonstrated below.
from hummus import Events
class Events(Events):
async def on_message_create(self, message): #activates every time a message is sent in a channel a bot can see
if message.content.startswith("ping"):
await message.reply("pong")
Make sure to put this in your main file, before your Client
class. You will also want to "listen" for this class before running the Client
class, as shown below:
Client = Commands(prefix="!", bottoken=token, status="online", game="!test")
async def bot():
await Client.LISTEN(Events())
await Client.RUN()
asyncio.run(bot())
See the Hummus documentation (go to the Gateway section) for event names, Hummus.py event function names will always be on_{event_name}
(except lowercase).
Embeds
You can create an embed object by importing the Embed
class from hummus
.
from hummus import Embed
class Commands(Client):
async def test(self,ctx:hummus.Message):
embed = Embed(title="Test",description="something")
The Embed
class supports 4 parameters: title
, description
, color
, and timestamp
. The color
parameter uses Discord's integer colors, view the list here. The timestamp
parameter uses the ISO8601 format (what Hummus accepts), so here's the code to get the current time in that format.
from datetime import datetime,timezone
current_time_utc = datetime.now(timezone.utc)
formatted_timestamp = current_time_utc.isoformat()
You can add attributes to the embed such as fields, footers, a thumbnail, an image, or an author like this:
class Commands(Client):
async def test(self,ctx:hummus.Message):
embed = Embed(title="Test",description="something")
await embed.addField(name="field title",value="field value")
await embed.addAuthor(name="author",url="https://google.com/",icon_url="https://www.google.com/favicon.ico")
await embed.addFooter(text="foot",icon_url="https://www.google.com/favicon.ico")
await embed.addThumbnail(url="https://google.com/favicon.ico")
Sending an embed is as expected, you would include it in the ctx.send()
or ctx.reply()
function like this:
class Commands(Client):
async def test(self,ctx:hummus.Message):
embed = Embed(title="Test",description="something")
await embed.addField(name="field title",value="field value")
await ctx.reply("its an embed!",embed=embed)
Note: Embeds will be ignored if an attachment is also passed!
Attachments
Attachments are quite simple. Create a File
object instance, and pass it through the file
parameter in ctx.send()
or ctx.reply()
.
A File
object can take 4 different types of parameters. Here is a list of them and their examples:
str
:file = hummus.File("path/to/file.here")
io.BufferedReader
:file = hummus.File(open("file.png","rb"))
bytes
:file = hummus.File(open("file.png","rb").read())
(.read()
returns abytes
object)BytesIO
:file = hummus.File(BytesIO(b""))
(BytesIO accepts abytes
object,File
gets it)
A use case is the following:
class Commands(Client):
async def test(self,ctx:hummus.Message):
await ctx.reply("its an image!",file=hummus.File("image.png"))
Note: Embeds will be ignored if an attachment is passed!
Support
I am LG125YT#2241 on Hummus, @ytlg on Discord, LG125YT#0001 on Rotisserie, and LG125YT#0001 on Cordcord/Watrlabs. My email is lg125yt@gmail.com, but you might want to let me know somewhere else that you sent me an email. I am most active on Rotisserie and Cordcord.
Roadmap
I'm really just doing what I see as most important for my bot development. I don't think I made the actual package usability great, consider contributing to help me out.
Changelog
Verson 0.6.4:
- Some bug in guild update got fixed, thats kinda it
- Excuses to pretend that I actively update hummus.py!!!! (real)
Version 0.6.3:
- ok guys, i tested starting up, so stuff should not break now (i love forgetting yet another bug making this unusable)
Version 0.6.2:
- i am so idiot i literally didnt test this version out and realize one of the files was missing a parenthesis! (the whole package broke
Version 0.6.1:
- Bug fixes with
setRoles()
,addRoles()
, andremoveRoles()
in theMember
andUser
objects - Forgot to add
ban()
andunban()
functions in theMember
andUser
objects, also their respective gateway events now exist inhummus.Events
- README update, more up-to-date ig
Version 0.6.0:
AllGuild
now inherits fromGuild
instead of making aGuild
attribute in the class.- New
Channel
andEmoji
objects exist, which arw available in anAllGuild
object.PermOverwrites
is also a new object that you can create instances of to edit channel permissions. AllMost events are added. In the middle of this update, banning has been added to Hummus and Oldground, and the events and functions do not exist for it yet.- Hummus.py now supports catching websocket errors, and auto-restarts if an error occurs (such as heartbeat expiring).
fullPermsCheck()
now correctly checks for server owner ID before returning whether the user has a specific permission.Channel
object attributes now have support for both voice and text channels.User
andMember
now have role-related functions (addRoles
,removeRoles
, andsetRoles
)- Many of these new objects have a
toDict()
function, however, this should be mainly internal. Message
objects now have aguild
andchannel
attribute containing the entireGuild
andChannel
objects (respectively). The old attributes that only had the ID are now calledguild_id
andchannel_id
.getGuildChannels()
exists as a function in aMessage
object.- The
Permissions
object now supports empty/fake role creation and the creation of a role with only a permissions integer. - The
Role
object now has aPermissions
object for theirpermissions
attribute. - The
bottoken
parameter name has been changed totoken
when creating aClient
/Commands
object. - Hummus.py will now get the cdn url from the base url provided, no need to provide a cdn url when creating a
Client
/Commands
object now.
Version 0.5.2:
- Again, no real changes to hummus.py, but I'm pretty sure I figured out how to fix the previous error. Also,
hmus
exists as a mirror package now if you don't want to type inhummus2016.py
to install the package.
Version 0.5.1:
- No real changes, I'm just trying to fix the pypi issue where
import hummus
doesnt work. Also updating this file a little to be more up-to-date.
Version 0.5.0:
- You now need to upload files by creating a
File
object with your file in it. This allows for less cluttering inmessage.py
for thesend
function, and it allows me to easily add more freedom to uploading files. - Continuation of above note, you can now provide a file path,
bytes
object,BytesIO
object, or anio.BufferedReader
object (the object you get when you assign a variable toopen("name.ext","rb")
) - All
Message
objects now have adelete()
function, which sends a request to delete that message on Hummus. (Previously, we only haddeleteMessage(id)
, where you needed to provide a message ID to delete.) - For some reason, the v0.4.x versions have not been updating the
hummus
folder, and only thehummus2016
folder. If it persists, consider switching allfrom hummus import *
andimport hummus
tofrom hummus2016 import *
andimport hummus2016
. This may also require changing some other lines, or you can useimport hummus2016 as hummus
.
Version 0.4.1:
- Updates in 0.4.0 didn't push for some reason lol. Here they are (again).
- Turns out that the Image and Thumbnail objects do work, they just require
width
andheight
parameters that Hummus.py will automatically take care for you. Nothing has changed, you can continue uploading files as before. - Fixed bug not letting Hummus.py work, its just me forgetting to include headers when making the request to get the websocket url.
Version 0.4.0:
- Added embed support (Image and Thumbnail objects to not work on Hummus, so they are not included in Hummus.py)
- Added attachment support (You only have to include the filepath)
Version 0.3.4:
- Fixed
ctx.edit()
so that it edits only the "content" section of the "reply" fromctx.reply()
(basically it doesn't edit the whole message) - Added a new
is_reply
(boolean),original_reply
(content of the message being replied to), andoriginal_author
(User
class) attributes to theMessage
objectis_reply
is entirely local, it is only true for a message object where thectx.reply()
function is used, and cannot detect replies outside ofctx.reply()
- Errors caught by the package exception handler now print out the traceback
- Minor changes to the README
Version 0.3.3:
- Updated the README (this thingy you're reading!) to not have the wackiest indents (frick you replit)
Version 0.3.2:
- Fixed hummus.py not working when you didn't manually register an
Events
class to listen.
Version 0.3.1:
- Quick update just added
ctx.deleteMessage(id)
because i forgor to add it before :sku:
Version 0.3.0:
- More events in the
Events
class work now, alsoself.allGuilds
in theClient
class now updates on related events like guild member join/leave and guild add/remove Client
class initialization redone a little, see the "Getting Started" and "Events" section in this README- Command args should now fully work, you do not need to add a default value to parameters and some annotations (namely
str
,int
, andhummus.member.User
) now work - The
Author
class has been renamed to theUser
class because most use cases within the package do not reflect the "Author" status of a user (this should not affect much of anything since annotations did not work before this update)
Too lazy to like do anything from before, if you want to do it for me, go for it.
Contributing
Contribute if you want, you can make a pull request on the GitLab repository, comment on the Replit project. Note that the Replit project is the most recent version of Hummus.py, because it is where I test new features. You can see upcoming features on the Replit project if you want.
Authors
This wrapper was made by LG125YT. Contact me on Hummus (LG125YT#2241) or Discord (@ytlg)
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.