#!/usr/bin/env python # # A library that provides a Python interface to the Telegram Bot API # Copyright (C) 2015-2024 # Leandro Toledo de Souza # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Lesser Public License as published by # the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Lesser Public License for more details. # # You should have received a copy of the GNU Lesser Public License # along with this program. If not, see [http://www.gnu.org/licenses/]. """This module contains the classes that represent Telegram InlineQueryResultAudio.""" from typing import TYPE_CHECKING, Optional, Sequence, Tuple from telegram._inline.inlinekeyboardmarkup import InlineKeyboardMarkup from telegram._inline.inlinequeryresult import InlineQueryResult from telegram._messageentity import MessageEntity from telegram._utils.argumentparsing import parse_sequence_arg from telegram._utils.defaultvalue import DEFAULT_NONE from telegram._utils.types import JSONDict, ODVInput from telegram.constants import InlineQueryResultType if TYPE_CHECKING: from telegram import InputMessageContent class InlineQueryResultAudio(InlineQueryResult): """ Represents a link to an mp3 audio file. By default, this audio file will be sent by the user. Alternatively, you can use :attr:`input_message_content` to send a message with the specified content instead of the audio. .. seealso:: :wiki:`Working with Files and Media ` Args: id (:obj:`str`): Unique identifier for this result, :tg-const:`telegram.InlineQueryResult.MIN_ID_LENGTH`- :tg-const:`telegram.InlineQueryResult.MAX_ID_LENGTH` Bytes. audio_url (:obj:`str`): A valid URL for the audio file. title (:obj:`str`): Title. performer (:obj:`str`, optional): Performer. audio_duration (:obj:`str`, optional): Audio duration in seconds. caption (:obj:`str`, optional): Caption, 0-:tg-const:`telegram.constants.MessageLimit.CAPTION_LENGTH` characters after entities parsing. parse_mode (:obj:`str`, optional): |parse_mode| caption_entities (Sequence[:class:`telegram.MessageEntity`], optional): |caption_entities| .. versionchanged:: 20.0 |sequenceclassargs| reply_markup (:class:`telegram.InlineKeyboardMarkup`, optional): Inline keyboard attached to the message. input_message_content (:class:`telegram.InputMessageContent`, optional): Content of the message to be sent instead of the audio. Attributes: type (:obj:`str`): :tg-const:`telegram.constants.InlineQueryResultType.AUDIO`. id (:obj:`str`): Unique identifier for this result, :tg-const:`telegram.InlineQueryResult.MIN_ID_LENGTH`- :tg-const:`telegram.InlineQueryResult.MAX_ID_LENGTH` Bytes. audio_url (:obj:`str`): A valid URL for the audio file. title (:obj:`str`): Title. performer (:obj:`str`): Optional. Performer. audio_duration (:obj:`str`): Optional. Audio duration in seconds. caption (:obj:`str`): Optional. Caption, 0-:tg-const:`telegram.constants.MessageLimit.CAPTION_LENGTH` characters after entities parsing. parse_mode (:obj:`str`): Optional. |parse_mode| caption_entities (Tuple[:class:`telegram.MessageEntity`]): Optional. |captionentitiesattr| .. versionchanged:: 20.0 * |tupleclassattrs| * |alwaystuple| reply_markup (:class:`telegram.InlineKeyboardMarkup`): Optional. Inline keyboard attached to the message. input_message_content (:class:`telegram.InputMessageContent`): Optional. Content of the message to be sent instead of the audio. """ __slots__ = ( "audio_duration", "audio_url", "caption", "caption_entities", "input_message_content", "parse_mode", "performer", "reply_markup", "title", ) def __init__( self, id: str, # pylint: disable=redefined-builtin audio_url: str, title: str, performer: Optional[str] = None, audio_duration: Optional[int] = None, caption: Optional[str] = None, reply_markup: Optional[InlineKeyboardMarkup] = None, input_message_content: Optional["InputMessageContent"] = None, parse_mode: ODVInput[str] = DEFAULT_NONE, caption_entities: Optional[Sequence[MessageEntity]] = None, *, api_kwargs: Optional[JSONDict] = None, ): # Required super().__init__(InlineQueryResultType.AUDIO, id, api_kwargs=api_kwargs) with self._unfrozen(): self.audio_url: str = audio_url self.title: str = title # Optionals self.performer: Optional[str] = performer self.audio_duration: Optional[int] = audio_duration self.caption: Optional[str] = caption self.parse_mode: ODVInput[str] = parse_mode self.caption_entities: Tuple[MessageEntity, ...] = parse_sequence_arg(caption_entities) self.reply_markup: Optional[InlineKeyboardMarkup] = reply_markup self.input_message_content: Optional[InputMessageContent] = input_message_content