Telethon/telethon/_client/uploads.py
2021-09-12 12:16:02 +02:00

468 lines
17 KiB
Python

import hashlib
import io
import itertools
import os
import pathlib
import re
import typing
from io import BytesIO
from ..crypto import AES
from .. import utils, helpers, hints, _tl
try:
import PIL
import PIL.Image
except ImportError:
PIL = None
if typing.TYPE_CHECKING:
from .telegramclient import TelegramClient
class _CacheType:
"""Like functools.partial but pretends to be the wrapped class."""
def __init__(self, cls):
self._cls = cls
def __call__(self, *args, **kwargs):
return self._cls(*args, file_reference=b'', **kwargs)
def __eq__(self, other):
return self._cls == other
def _resize_photo_if_needed(
file, is_image, width=1280, height=1280, background=(255, 255, 255)):
# https://github.com/telegramdesktop/tdesktop/blob/12905f0dcb9d513378e7db11989455a1b764ef75/Telegram/SourceFiles/boxes/photo_crop_box.cpp#L254
if (not is_image
or PIL is None
or (isinstance(file, io.IOBase) and not file.seekable())):
return file
if isinstance(file, bytes):
file = io.BytesIO(file)
before = file.tell() if isinstance(file, io.IOBase) else None
try:
# Don't use a `with` block for `image`, or `file` would be closed.
# See https://github.com/LonamiWebs/Telethon/issues/1121 for more.
image = PIL.Image.open(file)
try:
kwargs = {'exif': image.info['exif']}
except KeyError:
kwargs = {}
if image.width <= width and image.height <= height:
return file
image.thumbnail((width, height), PIL.Image.ANTIALIAS)
alpha_index = image.mode.find('A')
if alpha_index == -1:
# If the image mode doesn't have alpha
# channel then don't bother masking it away.
result = image
else:
# We could save the resized image with the original format, but
# JPEG often compresses better -> smaller size -> faster upload
# We need to mask away the alpha channel ([3]), since otherwise
# IOError is raised when trying to save alpha channels in JPEG.
result = PIL.Image.new('RGB', image.size, background)
result.paste(image, mask=image.split()[alpha_index])
buffer = io.BytesIO()
result.save(buffer, 'JPEG', **kwargs)
buffer.seek(0)
return buffer
except IOError:
return file
finally:
if before is not None:
file.seek(before, io.SEEK_SET)
async def send_file(
self: 'TelegramClient',
entity: 'hints.EntityLike',
file: 'typing.Union[hints.FileLike, typing.Sequence[hints.FileLike]]',
*,
caption: typing.Union[str, typing.Sequence[str]] = None,
force_document: bool = False,
file_size: int = None,
clear_draft: bool = False,
progress_callback: 'hints.ProgressCallback' = None,
reply_to: 'hints.MessageIDLike' = None,
attributes: 'typing.Sequence[_tl.TypeDocumentAttribute]' = None,
thumb: 'hints.FileLike' = None,
allow_cache: bool = True,
parse_mode: str = (),
formatting_entities: typing.Optional[typing.List[_tl.TypeMessageEntity]] = None,
voice_note: bool = False,
video_note: bool = False,
buttons: 'hints.MarkupLike' = None,
silent: bool = None,
background: bool = None,
supports_streaming: bool = False,
schedule: 'hints.DateLike' = None,
comment_to: 'typing.Union[int, _tl.Message]' = None,
ttl: int = None,
**kwargs) -> '_tl.Message':
# TODO Properly implement allow_cache to reuse the sha256 of the file
# i.e. `None` was used
if not file:
raise TypeError('Cannot use {!r} as file'.format(file))
if not caption:
caption = ''
entity = await self.get_input_entity(entity)
if comment_to is not None:
entity, reply_to = await self._get_comment_data(entity, comment_to)
else:
reply_to = utils.get_message_id(reply_to)
# First check if the user passed an iterable, in which case
# we may want to send grouped.
if utils.is_list_like(file):
if utils.is_list_like(caption):
captions = caption
else:
captions = [caption]
result = []
while file:
result += await self._send_album(
entity, file[:10], caption=captions[:10],
progress_callback=progress_callback, reply_to=reply_to,
parse_mode=parse_mode, silent=silent, schedule=schedule,
supports_streaming=supports_streaming, clear_draft=clear_draft,
force_document=force_document, background=background,
)
file = file[10:]
captions = captions[10:]
for doc, cap in zip(file, captions):
result.append(await self.send_file(
entity, doc, allow_cache=allow_cache,
caption=cap, force_document=force_document,
progress_callback=progress_callback, reply_to=reply_to,
attributes=attributes, thumb=thumb, voice_note=voice_note,
video_note=video_note, buttons=buttons, silent=silent,
supports_streaming=supports_streaming, schedule=schedule,
clear_draft=clear_draft, background=background,
**kwargs
))
return result
if formatting_entities is not None:
msg_entities = formatting_entities
else:
caption, msg_entities =\
await self._parse_message_text(caption, parse_mode)
file_handle, media, image = await self._file_to_media(
file, force_document=force_document,
file_size=file_size,
progress_callback=progress_callback,
attributes=attributes, allow_cache=allow_cache, thumb=thumb,
voice_note=voice_note, video_note=video_note,
supports_streaming=supports_streaming, ttl=ttl
)
# e.g. invalid cast from :tl:`MessageMediaWebPage`
if not media:
raise TypeError('Cannot use {!r} as file'.format(file))
markup = self.build_reply_markup(buttons)
request = _tl.fn.messages.SendMedia(
entity, media, reply_to_msg_id=reply_to, message=caption,
entities=msg_entities, reply_markup=markup, silent=silent,
schedule_date=schedule, clear_draft=clear_draft,
background=background
)
return self._get_response_message(request, await self(request), entity)
async def _send_album(self: 'TelegramClient', entity, files, caption='',
progress_callback=None, reply_to=None,
parse_mode=(), silent=None, schedule=None,
supports_streaming=None, clear_draft=None,
force_document=False, background=None, ttl=None):
"""Specialized version of .send_file for albums"""
# We don't care if the user wants to avoid cache, we will use it
# anyway. Why? The cached version will be exactly the same thing
# we need to produce right now to send albums (uploadMedia), and
# cache only makes a difference for documents where the user may
# want the attributes used on them to change.
#
# In theory documents can be sent inside the albums but they appear
# as different messages (not inside the album), and the logic to set
# the attributes/avoid cache is already written in .send_file().
entity = await self.get_input_entity(entity)
if not utils.is_list_like(caption):
caption = (caption,)
captions = []
for c in reversed(caption): # Pop from the end (so reverse)
captions.append(await self._parse_message_text(c or '', parse_mode))
reply_to = utils.get_message_id(reply_to)
# Need to upload the media first, but only if they're not cached yet
media = []
for file in files:
# Albums want :tl:`InputMedia` which, in theory, includes
# :tl:`InputMediaUploadedPhoto`. However using that will
# make it `raise MediaInvalidError`, so we need to upload
# it as media and then convert that to :tl:`InputMediaPhoto`.
fh, fm, _ = await self._file_to_media(
file, supports_streaming=supports_streaming,
force_document=force_document, ttl=ttl)
if isinstance(fm, (_tl.InputMediaUploadedPhoto, _tl.InputMediaPhotoExternal)):
r = await self(_tl.fn.messages.UploadMedia(
entity, media=fm
))
fm = utils.get_input_media(r.photo)
elif isinstance(fm, _tl.InputMediaUploadedDocument):
r = await self(_tl.fn.messages.UploadMedia(
entity, media=fm
))
fm = utils.get_input_media(
r.document, supports_streaming=supports_streaming)
if captions:
caption, msg_entities = captions.pop()
else:
caption, msg_entities = '', None
media.append(_tl.InputSingleMedia(
fm,
message=caption,
entities=msg_entities
# random_id is autogenerated
))
# Now we can construct the multi-media request
request = _tl.fn.messages.SendMultiMedia(
entity, reply_to_msg_id=reply_to, multi_media=media,
silent=silent, schedule_date=schedule, clear_draft=clear_draft,
background=background
)
result = await self(request)
random_ids = [m.random_id for m in media]
return self._get_response_message(random_ids, result, entity)
async def upload_file(
self: 'TelegramClient',
file: 'hints.FileLike',
*,
part_size_kb: float = None,
file_size: int = None,
file_name: str = None,
use_cache: type = None,
key: bytes = None,
iv: bytes = None,
progress_callback: 'hints.ProgressCallback' = None) -> '_tl.TypeInputFile':
if isinstance(file, (_tl.InputFile, _tl.InputFileBig)):
return file # Already uploaded
pos = 0
async with helpers._FileStream(file, file_size=file_size) as stream:
# Opening the stream will determine the correct file size
file_size = stream.file_size
if not part_size_kb:
part_size_kb = utils.get_appropriated_part_size(file_size)
if part_size_kb > 512:
raise ValueError('The part size must be less or equal to 512KB')
part_size = int(part_size_kb * 1024)
if part_size % 1024 != 0:
raise ValueError(
'The part size must be evenly divisible by 1024')
# Set a default file name if None was specified
file_id = helpers.generate_random_long()
if not file_name:
file_name = stream.name or str(file_id)
# If the file name lacks extension, add it if possible.
# Else Telegram complains with `PHOTO_EXT_INVALID_ERROR`
# even if the uploaded image is indeed a photo.
if not os.path.splitext(file_name)[-1]:
file_name += utils._get_extension(stream)
# Determine whether the file is too big (over 10MB) or not
# Telegram does make a distinction between smaller or larger files
is_big = file_size > 10 * 1024 * 1024
hash_md5 = hashlib.md5()
part_count = (file_size + part_size - 1) // part_size
self._log[__name__].info('Uploading file of %d bytes in %d chunks of %d',
file_size, part_count, part_size)
pos = 0
for part_index in range(part_count):
# Read the file by in chunks of size part_size
part = await helpers._maybe_await(stream.read(part_size))
if not isinstance(part, bytes):
raise TypeError(
'file descriptor returned {}, not bytes (you must '
'open the file in bytes mode)'.format(type(part)))
# `file_size` could be wrong in which case `part` may not be
# `part_size` before reaching the end.
if len(part) != part_size and part_index < part_count - 1:
raise ValueError(
'read less than {} before reaching the end; either '
'`file_size` or `read` are wrong'.format(part_size))
pos += len(part)
# Encryption part if needed
if key and iv:
part = AES.encrypt_ige(part, key, iv)
if not is_big:
# Bit odd that MD5 is only needed for small files and not
# big ones with more chance for corruption, but that's
# what Telegram wants.
hash_md5.update(part)
# The SavePartRequest is different depending on whether
# the file is too large or not (over or less than 10MB)
if is_big:
request = _tl.fn.upload.SaveBigFilePart(
file_id, part_index, part_count, part)
else:
request = _tl.fn.upload.SaveFilePart(
file_id, part_index, part)
result = await self(request)
if result:
self._log[__name__].debug('Uploaded %d/%d',
part_index + 1, part_count)
if progress_callback:
await helpers._maybe_await(progress_callback(pos, file_size))
else:
raise RuntimeError(
'Failed to upload file part {}.'.format(part_index))
if is_big:
return _tl.InputFileBig(file_id, part_count, file_name)
else:
return _tl.custom.InputSizedFile(
file_id, part_count, file_name, md5=hash_md5, size=file_size
)
async def _file_to_media(
self, file, force_document=False, file_size=None,
progress_callback=None, attributes=None, thumb=None,
allow_cache=True, voice_note=False, video_note=False,
supports_streaming=False, mime_type=None, as_image=None,
ttl=None):
if not file:
return None, None, None
if isinstance(file, pathlib.Path):
file = str(file.absolute())
is_image = utils.is_image(file)
if as_image is None:
as_image = is_image and not force_document
# `aiofiles` do not base `io.IOBase` but do have `read`, so we
# just check for the read attribute to see if it's file-like.
if not isinstance(file, (str, bytes, _tl.InputFile, _tl.InputFileBig))\
and not hasattr(file, 'read'):
# The user may pass a Message containing media (or the media,
# or anything similar) that should be treated as a file. Try
# getting the input media for whatever they passed and send it.
#
# We pass all attributes since these will be used if the user
# passed :tl:`InputFile`, and all information may be relevant.
try:
return (None, utils.get_input_media(
file,
is_photo=as_image,
attributes=attributes,
force_document=force_document,
voice_note=voice_note,
video_note=video_note,
supports_streaming=supports_streaming,
ttl=ttl
), as_image)
except TypeError:
# Can't turn whatever was given into media
return None, None, as_image
media = None
file_handle = None
if isinstance(file, (_tl.InputFile, _tl.InputFileBig)):
file_handle = file
elif not isinstance(file, str) or os.path.isfile(file):
file_handle = await self.upload_file(
_resize_photo_if_needed(file, as_image),
file_size=file_size,
progress_callback=progress_callback
)
elif re.match('https?://', file):
if as_image:
media = _tl.InputMediaPhotoExternal(file, ttl_seconds=ttl)
else:
media = _tl.InputMediaDocumentExternal(file, ttl_seconds=ttl)
else:
bot_file = utils.resolve_bot_file_id(file)
if bot_file:
media = utils.get_input_media(bot_file, ttl=ttl)
if media:
pass # Already have media, don't check the rest
elif not file_handle:
raise ValueError(
'Failed to convert {} to media. Not an existing file, '
'an HTTP URL or a valid bot-API-like file ID'.format(file)
)
elif as_image:
media = _tl.InputMediaUploadedPhoto(file_handle, ttl_seconds=ttl)
else:
attributes, mime_type = utils.get_attributes(
file,
mime_type=mime_type,
attributes=attributes,
force_document=force_document and not is_image,
voice_note=voice_note,
video_note=video_note,
supports_streaming=supports_streaming,
thumb=thumb
)
if not thumb:
thumb = None
else:
if isinstance(thumb, pathlib.Path):
thumb = str(thumb.absolute())
thumb = await self.upload_file(thumb, file_size=file_size)
media = _tl.InputMediaUploadedDocument(
file=file_handle,
mime_type=mime_type,
attributes=attributes,
thumb=thumb,
force_file=force_document and not is_image,
ttl_seconds=ttl
)
return file_handle, media, as_image