Telethon/README.md

# Telethon
**Telethon** is Telegram client implementation in Python. This project's _core_ is **completely based** on
[TLSharp](https://github.com/sochix/TLSharp). All the files which are fully based on it will have a notice
on the top of the file. Also don't forget to have a look to the original project.

The files without the previously mentioned notice are no longer part of TLSharp itself, or have enough modifications
to make them entirely different.

# Table of contents
- [Why Telethon?](#why-telethon)
- [Requirements](#requirements)
  - [Python modules](#python-modules)
  - [Obtaining your `API ID` and `Hash`](#obtaining-your-api-id-and-hash)
- [Running Telethon](#running-telethon)
- [Advanced uses](#advanced-uses)
  - [Using more than just `TelegramClient`](#using-more-than-just-telegramclient)
  - [Tips for porting Telethon](#tips-for-porting-telethon)
  - [Code generator limitations](#code-generator-limitations)
  - [Updating the `scheme.tl`](#updating-the-schemetl)
- [Plans for the future](#plans-for-the-future)

## Why Telethon?
> Why should I bother with Telethon? You already mentioned [TLSharp](https://github.com/sochix/TLSharp).
> [telegram-cli](https://github.com/vysheng/tg) has also been around for a long while. And we have the
> [official](https://github.com/telegramdesktop/tdesktop) [clients](https://github.com/DrKLO/Telegram)!

With Telethon you don't really need to know anything before using it. Create a client with your settings.
Connect. You're ready to go.

Being written on Python, Telethon can run as a script under any environment you wish, (yes,
[Android too](https://f-droid.org/repository/browse/?fdfilter=termux&fdid=com.termux)). You can schedule it,
or use it in any other script you have. Want to send a message to someone when you're available? Write a script.
Do you want check for new messages at a given time and find relevant ones? Write a script.

An official client has a feature which Telethon doesn't? [Implement it easily](#how-to-add-more-functions-to-telegramclient).

## Requirements
### Python modules
This project requires the following Python modules, which can be installed by issuing `sudo -H pip3 install <module>` on a
Linux terminal:
- `pyaes` ([GitHub](https://github.com/ricmoo/pyaes), [package index](https://pypi.python.org/pypi/pyaes))

### Obtaining your `API ID` and `Hash`
1. Follow [this link](https://my.telegram.org) and login with your phone number.
2. Click under `API Development tools`.
3. A `Create new application` window will appear. Fill in your application details.
There is no need to enter any `URL`, and only the first two fields (`App title` and `Short name`)
can be changed later as long as I'm aware.
4. Click on `Create application` at the end. Now that you have the `API ID` and `Hash`,
head to `api/` directory and create a copy of the `settings_example` file, naming it `settings` (lowercase!).
Then fill the file with the corresponding values (your `api_id`, `api_hash` and phone number in international format).

## Running Telethon
First of all, you need to run the `tl_generator.py` by issuing `python3 tl_generator.py`. This will generate all the
TLObjects from the given `scheme.tl` file. When it's done, you can run `python3 main.py` to start the interactive example.

## Advanced uses
### Using more than just `TelegramClient`
The `TelegramClient` class should be used to provide a quick, well-documented and simplified starting point.
It is **not** meant to be a place for _all_ the available Telegram `Request`'s, because there are simply too many.

However, this doesn't mean that you cannot `invoke` all the power of Telegram's API. Whenever you need to `invoke`
a Telegram `Request`, all you need to do is the following:
```python
result = client.invoke(SomeRequest(...))
```

You have just `invoke`'d `SomeRequest` and retrieved its `result`! That wasn't hard at all, was it? Now you may wonder,
what's the deal with _all the power of Telegram's API_? Have a look under `tl/functions/`.
That is _everything_ you can do. You have **over 200 API `Request`'s** at your disposal.

However, we don't pretty know _how_ that `result` looks like. Easy. `print(str(result))` should give you a quick overview.
Nevertheless, there may be more than a single `result`! Let's have a look at this seemingly innocent `TL` definition:

`messages.getWebPagePreview#25223e24 message:string = MessageMedia;`

Focusing on the end, we can see that the `result` of invoking `GetWebPagePreviewRequest` is `MessageMedia`. But how
can `MessageMedia` exactly look like? It's time to have another look, but this time under `tl/types/`:
```sh
$ tree -P "message_media_*"
.
├── tl
│   └── types
│       ├── message_media_contact.py
│       ├── message_media_document.py
│       ├── message_media_empty.py
│       ├── message_media_geo.py
│       ├── message_media_photo.py
│       ├── message_media_unsupported.py
│       ├── message_media_venue.py
│       └── message_media_web_page.py
```
Those are _eight_ different types! How do we know what exact type it is to determine its properties? A simple
`if type(result) == MessageMediaContact:` or similar will do. Now you're ready to take advantage of Telegram's polymorphism.

### Tips for porting Telethon
First of all, you need to understand how the `scheme.tl` (`TL` language) works. Every object definition is written as follows:

`name#id argument_name:argument_type = CommonType`

This means that in a single line you know what the `TLObject` name is. You know it's unique ID,
and you know what arguments it has. It really isn't that hard to write a generator for generating code to any platform!

The generated code should also be able to _encode_ the `Request` into bytes, so they can be sent over the network.
This isn't a big deal either, because you know how the `TLObject`'s are made.

Once you have your own [code generator](tl_generator.py), start by looking at the
[first release](https://github.com/LonamiWebs/Telethon/releases/tag/v0.1) of Telethon.
The code there is simple to understand, easy to read and hence easy to port. No extra useless features.
Only the bare bones. Perfect for starting a _new implementation_.

P.S.: I may have lied a bit. The `TL` language is not that easy. But it's not that hard either.
You're free to sniff the `parser/` files and learn how to parse other more complex lines.
Or simply use that code and change the [SourceBuilder](parser/source_builder.py)!

### Code generator limitations
The current code generator is not complete, yet adding the missing features would only over-complicate an already hard-to-read code.
Some parts of the `.tl` file _should_ be omitted, because they're "built-in" in the generated code (such as writing booleans, etc.).

In order to make sure that all the generated files will work, please make sure to **always** comment out these lines in `scheme.tl`
(the latest version can always be found
[here](https://github.com/telegramdesktop/tdesktop/blob/master/Telegram/SourceFiles/mtproto/scheme.tl)):

```tl
// boolFalse#bc799737 = Bool;
// boolTrue#997275b5 = Bool;
// true#3fedd339 = True;
// vector#1cb5c415 {t:Type} # [ t ] = Vector t;
```

Also please make sure to rename `updates#74ae4240 ...` to `updates_tg#74ae4240 ...` or similar to avoid confusion between
the `updates` folder and the `updates.py` file! Note that depending on the name, it may break things somewhere else. So
please stick with the suggested name or give one which is still descriptive enough and easy to remember.

### Updating the `scheme.tl`
Have you found a more updated version of the `scheme.tl` file? Those are great news! Updating is as simple as grabbing the
[latest version](https://github.com/telegramdesktop/tdesktop/blob/master/Telegram/SourceFiles/mtproto/scheme.tl) and
replacing the one you can find in this same directory by the updated one. Don't forget to run `python3 tl_generator.py`
afterwards and specifying the new layer number to be used when creating the `TelegramClient`.

If the changes weren't too big, everything should still work the same way as it did before; but with extra features.

## Plans for the future
If everything works well, this probably ends up being a Python package :)

But as of now, and until that happens, help is highly appreciated!