#!/usr/bin/env python # # A library that provides a Python interface to the Telegram Bot API # Copyright (C) 2015-2022 # 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/]. import pytest from telegram import ( InlineKeyboardButton, InlineKeyboardMarkup, InlineQueryResultCachedVideo, InlineQueryResultCachedVoice, InputTextMessageContent, MessageEntity, ) @pytest.fixture(scope="class") def inline_query_result_cached_video(): return InlineQueryResultCachedVideo( TestInlineQueryResultCachedVideo.id_, TestInlineQueryResultCachedVideo.video_file_id, TestInlineQueryResultCachedVideo.title, caption=TestInlineQueryResultCachedVideo.caption, parse_mode=TestInlineQueryResultCachedVideo.parse_mode, caption_entities=TestInlineQueryResultCachedVideo.caption_entities, description=TestInlineQueryResultCachedVideo.description, input_message_content=TestInlineQueryResultCachedVideo.input_message_content, reply_markup=TestInlineQueryResultCachedVideo.reply_markup, ) class TestInlineQueryResultCachedVideo: id_ = "id" type_ = "video" video_file_id = "video file id" title = "title" caption = "caption" parse_mode = "Markdown" caption_entities = [MessageEntity(MessageEntity.ITALIC, 0, 7)] description = "description" input_message_content = InputTextMessageContent("input_message_content") reply_markup = InlineKeyboardMarkup([[InlineKeyboardButton("reply_markup")]]) def test_slot_behaviour(self, inline_query_result_cached_video, mro_slots): inst = inline_query_result_cached_video for attr in inst.__slots__: assert getattr(inst, attr, "err") != "err", f"got extra slot '{attr}'" assert len(mro_slots(inst)) == len(set(mro_slots(inst))), "duplicate slot" def test_expected_values(self, inline_query_result_cached_video): assert inline_query_result_cached_video.type == self.type_ assert inline_query_result_cached_video.id == self.id_ assert inline_query_result_cached_video.video_file_id == self.video_file_id assert inline_query_result_cached_video.title == self.title assert inline_query_result_cached_video.description == self.description assert inline_query_result_cached_video.caption == self.caption assert inline_query_result_cached_video.parse_mode == self.parse_mode assert inline_query_result_cached_video.caption_entities == self.caption_entities assert ( inline_query_result_cached_video.input_message_content.to_dict() == self.input_message_content.to_dict() ) assert ( inline_query_result_cached_video.reply_markup.to_dict() == self.reply_markup.to_dict() ) def test_to_dict(self, inline_query_result_cached_video): inline_query_result_cached_video_dict = inline_query_result_cached_video.to_dict() assert isinstance(inline_query_result_cached_video_dict, dict) assert ( inline_query_result_cached_video_dict["type"] == inline_query_result_cached_video.type ) assert inline_query_result_cached_video_dict["id"] == inline_query_result_cached_video.id assert ( inline_query_result_cached_video_dict["video_file_id"] == inline_query_result_cached_video.video_file_id ) assert ( inline_query_result_cached_video_dict["title"] == inline_query_result_cached_video.title ) assert ( inline_query_result_cached_video_dict["description"] == inline_query_result_cached_video.description ) assert ( inline_query_result_cached_video_dict["caption"] == inline_query_result_cached_video.caption ) assert ( inline_query_result_cached_video_dict["parse_mode"] == inline_query_result_cached_video.parse_mode ) assert inline_query_result_cached_video_dict["caption_entities"] == [ ce.to_dict() for ce in inline_query_result_cached_video.caption_entities ] assert ( inline_query_result_cached_video_dict["input_message_content"] == inline_query_result_cached_video.input_message_content.to_dict() ) assert ( inline_query_result_cached_video_dict["reply_markup"] == inline_query_result_cached_video.reply_markup.to_dict() ) def test_equality(self): a = InlineQueryResultCachedVideo(self.id_, self.video_file_id, self.title) b = InlineQueryResultCachedVideo(self.id_, self.video_file_id, self.title) c = InlineQueryResultCachedVideo(self.id_, "", self.title) d = InlineQueryResultCachedVideo("", self.video_file_id, self.title) e = InlineQueryResultCachedVoice(self.id_, "", "") assert a == b assert hash(a) == hash(b) assert a is not b assert a == c assert hash(a) == hash(c) assert a != d assert hash(a) != hash(d) assert a != e assert hash(a) != hash(e)