Media

Chapter
10 / 12
Time
1 min
Code
4

Bots almost never work with text only. In telegrinder media work usually falls into three common scenarios:

  • sending files
  • downloading files
  • handling media groups

Sending a file#

A simple way to send a local file is to build an InputFile:

import pathlib

from telegrinder import Message
from telegrinder.types import InputFile

cool_bytes = pathlib.Path("examples/assets/satie.jpeg").read_bytes()


@bot.on.message(Text("/photo"))
async def start(message: Message) -> None:
    await message.answer_photo(
        InputFile("satie.jpeg", cool_bytes),
        caption="Erik",
    )

This works well when the file already exists on disk or in memory.

Resolving and downloading a file#

If a user sends a photo, you can get its file_id, resolve the Telegram file path, and then download the bytes.

import pathlib

from telegrinder import Message
from telegrinder.rules import ABCRule

photos_path = pathlib.Path("photos")
photos_path.mkdir(exist_ok=True)


class HasPhoto(ABCRule):
    async def check(self, message: Message) -> bool:
        return bool(message.photo.unwrap_or_none())


@bot.on.message(HasPhoto())
async def downloader(message: Message) -> str:
    photo_file = (await message.api.get_file(file_id=message.photo.unwrap()[-1].file_id)).unwrap()
    photo_path = photo_file.file_path.unwrap()

    path = photos_path / pathlib.Path(photo_path.split("/")[-1])
    path.write_bytes((await message.api.download_file(photo_path)).unwrap())
    return "Photo downloaded!"

Attachment nodes#

You do not always need to unpack media manually. Telegrinder already provides useful built-in nodes:

  • Attachment
  • Photo
  • Video
  • Document
  • Audio
  • Voice
  • Caption
  • File[...]

For example:

from telegrinder import Message, node
from telegrinder.node import File, Photo


@bot.on.message()
async def photo_handler(message: Message, file: File[Photo]) -> None:
    await message.answer(f"File path: {file.file_path.unwrap_or('None')}")

That is much nicer than repeating attachment extraction logic in every handler.

Media groups#

Media groups are a separate case because Telegram sends multiple media items as one set.

For that there is a dedicated media_group view and a MediaGroup node:

from kungfu.library.monad.option import Option

from telegrinder.node import Caption, MediaGroup


@bot.on.media_group()
async def handle_media_group(media_group: MediaGroup, caption: Option[Caption]) -> str:
    return (
        f"Received media group with {len(media_group.items)} items\n"
        f"Caption: {caption.unwrap_or('no caption')}"
    )

That is much better than manually merging updates by media_group_id.

Useful references#