Skip to content

one-among-us/TelegramBackup

Repository files navigation

TelegramBackup

This is the Telegram adapter for tg-blog, a front-end for displaying telegram (or any compatible) channel data as an interactive web page.

Motivation

Telegram has become increasingly fragile, recently revoking usernames for inactive channels, and often randomly banning regular users misclassified as spam. During the username revoking, many channels of deceased individuals have been remove from public space, no longer accessible using t.me links. This tool helps to maintain Telegram channel data in case of an accidental event, and also to publicly display inactive channels that have been revoked.

Demos / Examples

Usage

Installation

First, install Python >= 3.11. Then, run pip install tgc

Then, to support video/animation/sticker conversion, you have to install the following non-python dependencies:

  1. Install Node 19.2 and yarn 1.22
  2. yarn global add puppeteer-lottie-cli
  3. Install ffmpeg using your system package manager

Mode 1: Convert Telegram Export

If you only need a one-time export, you can use mode 1. To do this, you first need to export a channel using tdesktop.

To convert an export file into a format supported by tg-blog, you can run tgce <export path>

Mode 2: Crawl Channel using MTProto API

If you have the permission to add a bot account to a channel, or invite a self-bot account, you can use the MTProto crawler for automatic incremental export updates. (Please, do not log into your own Telegram account for crawling, there's a very high chance of being mis-classified as spam and get banned)

Using this method, it can automatically update the channel backup incrementally, and the information will be more complete. However, it is more difficult to set up than mode 1.

Setup API Keys

  1. Obtain api_id and api_hash by creating your Telegram application (Official Guide)
    1. Log into https://my.telegram.org/apps
    2. Fill out the form to create an application
    3. Or, if you want to go against Telegram's EULA for your convenience, you can leave them blank to use Telegram's official client's API hash at your own risk.
  2. Choose which type of account to log in:
    1. Bot account: Create a bot using the @BotFather bot.
    2. Self-bot account: Leave bot_token blank, it will prompt you to login. You should only use a self-bot when you're not the admin of the channel (because inviting a bot requires admin access).
  3. Fill in the tokens in ~/.config/tgc/config.toml as shown below
# Telegram API id & hash
api_id = 10000000
api_hash = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"

# Telegram bot token (leave blank to log into self-bot account)
bot_token = "0000000000:aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"

After setting up the keys, you can proceed to setting up the channel:

Setup Channels to be Crawled

  1. Either invite your bot to the channel or join the channel on your self-bot account
  2. Forward a channel message to @RawDataBot to obtain the channel ID. (You'll see a JSON response, and you can find the ID from the forward_from_chat field)
  3. Fill in the channel info in ~/.config/tgc/config.toml as shown below
# One export entry in a list of exports
[[exports]]
chat_id = -1001191767119    # Telegram channel chat id
path = "exports/hykilp"     # Output Path

After all setup is complete, you can proceed to running the crawler.

Running the Crawler

Simply run the tgc command.

Additional Config

You can set additional configuration for each export entry like below:

Field Description Type
size_limit_mb Limit downloaded file size (skip large files) float

RSS Feed Generation

If you want to generate RSS feed, you can add the following under the export entry:

[[exports]]
chat_id = -1001191767119    # Telegram channel chat id
path = "exports/hykilp"     # Output Path

[exports.rss]
title = "小桂桂的回忆录 📒"
link = "https://aza.moe/life"
description = "「我们所经历的每个平凡的日常,也许就是连续发生的奇迹」"
language = "zh-cn"
image_url = "https://aza.moe/meru_256px.png"

Automatic Updates using GitHub Actions

If you want to automatically backup/sync telegram channel data using GitHub Actions, you can do this.

  1. Create your config.toml
  2. Make a fork of https://github.com/hykilpikonna/blog-data
  3. In GitHub's Settings tab, create a secret called TGC_CONFIG, and paste your config.toml there.
  4. Remember to set a size_limit_mb to below 100 so that git lfs won't warn you for pushing large files.