About

This is a telegram bot that can be used to download files forwarded to the bot by the owner.

Build with Rust and Teloxide.

file sent directly to the bot

The bot will download the file and save it to the specified directory.

The owner can then react to the messages with emoji to manage the file:

• "👍" | "❤": pin and move the file to favorite directory
• "👎": move the file to trash

file sent to bot managed channel

Initially, the bot owner send /troggle <bypasskey> to the bot to troggle among states:

• paused: pause the bot
• active: sync files and answer reactions
• partially active: answer reactions but not sync files

(the <bypasskey> can be seen in the log, and send /bypasskey to reprint the pwd in the log)

The bot will set "🫡" reaction to the file message to indicate the file is downloading.

Once done, the bot will set "👌". ("😭" if failed, "😨" if canceled, "👾" for inner bug)

People can react to the file with emoji, and the bot will count the score of the file.

Emoji	Score
👍😁🙏😇🤗	+1
❤🔥🥰🎉🍌💋💘😘	+2
❤‍🔥	+3
👎🤯😱😢🥴🌚😐🖕😨	-1
🤬🤮💩🤡💔😡	-2

If the score >= fav_score_limit, the bot will hard-link the file to favorite directory and pin it.

If the score < delete_score_limit, the bot will hard-link the file to trash and delete from channel.

Otherwise, the bot will hard-link the file to normal directory and unpin the file if necessary.

Note: it takes minites to get ReactionCountUpdate, so the bot will not handle reaction from channel immediately.

Deploy

You could create a .env file with the following content:

# Get from botfather
TELOXIDE_TOKEN=xxxxxxxxxx:xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# Your telegram id found in your profile
BYPASS_USERS=xxxxx,xxxxx
# if you want to use local server:
TELEGRAM_API_ID=...
TELEGRAM_API_HASH=...

Deploy:

File size limit 20MB

A telegram bot to sync files to local server.

Usage: telegram_sync_bot <COMMAND>

Commands:
  run     Run the bot
  delete  Delete files by file_name in the data dir, and delete the record in the database, delete the message in the channel. The database should not be locked by other process, and there should not be any other bot instance.

telegram_sync_bot run -d /path/to/data

No file size limit (local server)

You need to apply for telegram api id and hash from Telegram first. (If you always get Error during applying, try cloudflare warp as VPN)

All methods below running local server in container first.

(You can also run local server natively, just omit -c and -i args when start telegram_sync_bot. I'll just skip this method here)

Get local server image first:

Prepare (Windows and MacOS with podman only):

podman machine init -v /path/to/output:/path/to/output bot_machine
podman machine start bot_machine

You can use the following command to build the telegram api bot local server image:

podman build --target server -t server --network host server

Or download and load from the release page (server.tar.gz), I've built one through GitHub Action for you.

We provide 4 ways here:

• native
• pod
• podman kube play
• k8s

normal way: server in container but bot native

podman run --name server -itd --env-file .env -p 8081:8081 server

telegram_sync_bot run -d /path/to/output -l http://127.0.0.1:8081 -c podman -i server

run as pod

Build telegram_sync_bot into container image:

# build bot image
podman build --target bot -t bot:$(cargo pkgid -p telegram_sync_bot | sed -n "s/.*@//p") --network host bot

Or download and load from the release page (bot.tar.gz).

Start server and bot in a pod:

podman pod create sync_bot

podman run --pod sync_bot --name server -itd --env-file .env \
    -v /path/to/data:/app/data server
podman run --pod sync_bot --name bot -itd --env-file .env --stop-signal SIGINT. \
    -v /path/to/data:/app/data.  \
    bot:0.X.0. \
    run -d /app/data -l http://server:8081

run with podman kube play

Modify sync-bot.yaml to fit your need.

You can download and load server.tar.gz and bot.tar.gz from the release page first. Or command below will automatically build the images for you which cost a lot of time.

podman kube play sync-bot.yaml

run with k8s

Build and save the image as .tar.gz or download from the release page first.

Modify .env and k8s/pv.yaml and so on to fit your need.

# load local images
sudo ctr -n=k8s.io images import /tmp/server.tar.gz
sudo ctr -n=k8s.io images import /tmp/bot.tar.gz

sudo crictl image
# you should see localhost/bot and localhost/server available

kubectl apply -k .

Systemd Service

Native without local server:

# /etc/systemd/system/sync-bot.service
[Unit]
Description=Telegram file sync bot
After=network-online.target

[Service]
Type=simple
User=<...>
WorkingDirectory=</path/to/output>
ExecStart=/usr/local/bin/telegram_sync_bot run
Restart=on-failure
Environment="TELOXIDE_TOKEN=<...>"
Environment="BYPASS_USERS=<...>"

[Install]
WantedBy=multi-user.target

or with local server container and native telegram_sync_bot (after the first setup):

# /etc/systemd/system/sync-bot.service
[Unit]
Description=Telegram file sync bot
After=network-online.target

[Service]
Type=simple
User=<...>
WorkingDirectory=</path/to/output>
ExecStartPre=/usr/bin/podman restart server
ExecStart=/usr/local/bin/telegram_sync_bot run -l http://127.0.0.1:8081 -c podman -i server
ExecStop=/bin/bash -c 'kill -SIGINT $MAINPID; for i in {1..5}; do sleep 1; kill -0 $MAINPID 2>/dev/null || exit 0; done; kill -SIGKILL $MAINPID'
ExecStopPost=/usr/bin/podman stop server
Restart=on-failure
Environment="TELOXIDE_TOKEN=<...>"
Environment="BYPASS_USERS=<...>"

[Install]
WantedBy=multi-user.target

or with pure pod (after images are built or loaded):

# /etc/container/systemd/users/<UserID>/sync-bot.kube
[Unit]
Description=Telegram file sync bot
After=network-online.target

[Kube]
Yaml=/etc/containers/systemd/users/<UserID>/sync-bot.yaml

[Install]
WantedBy=default.target

Search podman quadlet for using podman kube play as systemd service.

systemctl --user daemon-reload
systemctl start --user sync-bot

Note: you can use /usr/lib/systemd/system-generators/podman-system-generator --user --dryrun to check the generated service file.

Tips

Use fd to delete database, channel message, and files in the data dir,

fd ".*\.[jpg|mp4]" '/path/to/data' -X podman run --name bot -it --env-file .env -v /path/to/data:/app/data --replace bot:0.X.0 delete -d /app/data {/}

Development

Rust 2024 is essencial

Set DATABASE_URL in .env to generate entity crate.

# .env
DATABASE_URL=sqlite://data/data.db

Then you can run the following command to create the database and generate the entity:

cargo install sea-orm-cli
mkdir data
sea-orm-cli migrate refresh
sea-orm-cli generate entity --expanded-format -o bot/src/storage/entity/inner