Advanced Usage

wzgram’s API – which consists of well documented methods and types – exists to provide an easier interface to the more complex Telegram API.

In this section, you’ll be shown the alternative way of communicating with Telegram using wzgram: the main “raw” Telegram API with its functions and types.

Telegram Raw API

If you can’t find a high-level method for your needs or if you want complete, low-level access to the whole Telegram API, you have to use the raw functions and types. Both are generated from the TL schema, so they always match the layer wzgram was built against; the names come straight from Telegram’s own schema.

As already hinted, raw functions and types can be less convenient. This section will therefore explain some pitfalls to take into consideration when working with the raw API.

Tip

Every available high-level method in wzgram is built on top of these raw functions.

Invoking Functions

Unlike the methods found in wzgram’s API, which can be called in the usual simple way, functions to be invoked from the raw Telegram API have a different way of usage.

First of all, both raw functions and raw types live in their respective packages (and sub-packages): pyrogram.raw.functions, pyrogram.raw.types. They all exist as Python classes, meaning you need to create an instance of each every time you need them and fill them in with the correct values using named arguments.

Next, to actually invoke the raw function you have to use the invoke() method provided by the Client class and pass the function object you created.

Here’s some examples:

  • Update first name, last name and bio:

    from pyrogram import Client
from pyrogram.raw import functions

async with Client("my_account") as app:
    await app.invoke(
        functions.account.UpdateProfile(
            first_name="First Name", last_name="Last Name",
            about="New bio text"
        )
    )

  • Set online/offline status:

    from pyrogram import Client
from pyrogram.raw import functions, types

async with Client("my_account") as app:
    # Set online status
    await app.invoke(functions.account.UpdateStatus(offline=False))

    # Set offline status
    await app.invoke(functions.account.UpdateStatus(offline=True))

  • Get chat info:

    from pyrogram import Client
from pyrogram.raw import functions, types

async with Client("my_account") as app:
    r = await app.invoke(
        functions.channels.GetFullChannel(
            channel=await app.resolve_peer("username")
        )
    )

    print(r)

Chat IDs

The way Telegram works makes it not possible to directly send a message to a user or a chat by using their IDs only. Instead, a pair of id and access_hash wrapped in a so called InputPeer is always needed. wzgram allows sending messages with IDs only thanks to cached access hashes.

There are three different InputPeer types, one for each kind of Telegram entity. Whenever an InputPeer is needed you must pass one of these:

  • InputPeerUser - Users

  • InputPeerChat - Basic Chats

  • InputPeerChannel - Channels & Supergroups

But you don’t necessarily have to manually instantiate each object because wzgram already provides resolve_peer() as a convenience utility method that returns the correct InputPeer by accepting a peer ID only.

Another thing to take into consideration about chat IDs is the way they are represented: they are all integers and all positive within their respective raw types.

Things are different when working with wzgram’s API because having them in the same space could lead to collisions, and that’s why wzgram uses a slightly different representation for each kind of ID.

For example, given the ID 123456789, here’s how wzgram can tell entities apart:

  • +ID User: 123456789

  • -ID Chat: -123456789

  • -100ID Channel or Supergroup: -100123456789

So, every time you take a raw ID, make sure to translate it into the correct ID when you want to use it with an high-level method.

Note

The raw API tracks the TL layer, and a layer bump can change a field’s type, not only add fields. Code that builds raw objects by hand is what breaks on an upgrade — the high-level methods are updated for you.