-
Notifications
You must be signed in to change notification settings - Fork 5.1k
/
test_audio.py
342 lines (285 loc) · 13.5 KB
/
test_audio.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
#!/usr/bin/env python
#
# A library that provides a Python interface to the Telegram Bot API
# Copyright (C) 2015-2022
# Leandro Toledo de Souza <devs@python-telegram-bot.org>
#
# 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 os
from pathlib import Path
import pytest
from telegram import Audio, Bot, InputFile, MessageEntity, Voice
from telegram.error import TelegramError
from telegram.helpers import escape_markdown
from telegram.request import RequestData
from tests.auxil.bot_method_checks import (
check_defaults_handling,
check_shortcut_call,
check_shortcut_signature,
)
from tests.conftest import data_file
@pytest.fixture(scope="function")
def audio_file():
with open(data_file("telegram.mp3"), "rb") as f:
yield f
@pytest.fixture(scope="class")
async def audio(bot, chat_id):
with data_file("telegram.mp3").open("rb") as f:
thumb = data_file("thumb.jpg")
return (
await bot.send_audio(chat_id, audio=f, read_timeout=50, thumb=thumb.open("rb"))
).audio
class TestAudio:
caption = "Test *audio*"
performer = "Leandro Toledo"
title = "Teste"
file_name = "telegram.mp3"
duration = 3
# audio_file_url = 'https://python-telegram-bot.org/static/testfiles/telegram.mp3'
# Shortened link, the above one is cached with the wrong duration.
audio_file_url = "https://goo.gl/3En24v"
mime_type = "audio/mpeg"
file_size = 122920
thumb_file_size = 1427
thumb_width = 50
thumb_height = 50
audio_file_id = "5a3128a4d2a04750b5b58397f3b5e812"
audio_file_unique_id = "adc3145fd2e84d95b64d68eaa22aa33e"
def test_slot_behaviour(self, audio, mro_slots):
for attr in audio.__slots__:
assert getattr(audio, attr, "err") != "err", f"got extra slot '{attr}'"
assert len(mro_slots(audio)) == len(set(mro_slots(audio))), "duplicate slot"
def test_creation(self, audio):
# Make sure file has been uploaded.
assert isinstance(audio, Audio)
assert isinstance(audio.file_id, str)
assert isinstance(audio.file_unique_id, str)
assert audio.file_id != ""
assert audio.file_unique_id != ""
def test_expected_values(self, audio):
assert audio.duration == self.duration
assert audio.performer is None
assert audio.title is None
assert audio.mime_type == self.mime_type
assert audio.file_size == self.file_size
assert audio.thumb.file_size == self.thumb_file_size
assert audio.thumb.width == self.thumb_width
assert audio.thumb.height == self.thumb_height
@pytest.mark.flaky(3, 1)
async def test_send_all_args(self, bot, chat_id, audio_file, thumb_file):
message = await bot.send_audio(
chat_id,
audio=audio_file,
caption=self.caption,
duration=self.duration,
performer=self.performer,
title=self.title,
disable_notification=False,
protect_content=True,
parse_mode="Markdown",
thumb=thumb_file,
)
assert message.caption == self.caption.replace("*", "")
assert isinstance(message.audio, Audio)
assert isinstance(message.audio.file_id, str)
assert isinstance(message.audio.file_unique_id, str)
assert message.audio.file_unique_id is not None
assert message.audio.file_id is not None
assert message.audio.duration == self.duration
assert message.audio.performer == self.performer
assert message.audio.title == self.title
assert message.audio.file_name == self.file_name
assert message.audio.mime_type == self.mime_type
assert message.audio.file_size == self.file_size
assert message.audio.thumb.file_size == self.thumb_file_size
assert message.audio.thumb.width == self.thumb_width
assert message.audio.thumb.height == self.thumb_height
assert message.has_protected_content
@pytest.mark.flaky(3, 1)
async def test_send_audio_custom_filename(self, bot, chat_id, audio_file, monkeypatch):
async def make_assertion(url, request_data: RequestData, *args, **kwargs):
return list(request_data.multipart_data.values())[0][0] == "custom_filename"
monkeypatch.setattr(bot.request, "post", make_assertion)
assert await bot.send_audio(chat_id, audio_file, filename="custom_filename")
@pytest.mark.flaky(3, 1)
async def test_get_and_download(self, bot, audio):
path = Path("telegram.mp3")
if path.is_file():
path.unlink()
new_file = await bot.get_file(audio.file_id)
assert new_file.file_size == self.file_size
assert new_file.file_id == audio.file_id
assert new_file.file_unique_id == audio.file_unique_id
assert str(new_file.file_path).startswith("https://")
await new_file.download_to_drive("telegram.mp3")
assert path.is_file()
@pytest.mark.flaky(3, 1)
async def test_send_mp3_url_file(self, bot, chat_id, audio):
message = await bot.send_audio(
chat_id=chat_id, audio=self.audio_file_url, caption=self.caption
)
assert message.caption == self.caption
assert isinstance(message.audio, Audio)
assert isinstance(message.audio.file_id, str)
assert isinstance(message.audio.file_unique_id, str)
assert message.audio.file_unique_id is not None
assert message.audio.file_id is not None
assert message.audio.duration == audio.duration
assert message.audio.mime_type == audio.mime_type
assert message.audio.file_size == audio.file_size
@pytest.mark.flaky(3, 1)
async def test_resend(self, bot, chat_id, audio):
message = await bot.send_audio(chat_id=chat_id, audio=audio.file_id)
assert message.audio == audio
async def test_send_with_audio(self, monkeypatch, bot, chat_id, audio):
async def make_assertion(url, request_data: RequestData, *args, **kwargs):
return request_data.json_parameters["audio"] == audio.file_id
monkeypatch.setattr(bot.request, "post", make_assertion)
message = await bot.send_audio(audio=audio, chat_id=chat_id)
assert message
@pytest.mark.flaky(3, 1)
async def test_send_audio_caption_entities(self, bot, chat_id, audio):
test_string = "Italic Bold Code"
entities = [
MessageEntity(MessageEntity.ITALIC, 0, 6),
MessageEntity(MessageEntity.ITALIC, 7, 4),
MessageEntity(MessageEntity.ITALIC, 12, 4),
]
message = await bot.send_audio(
chat_id, audio, caption=test_string, caption_entities=entities
)
assert message.caption == test_string
assert message.caption_entities == entities
@pytest.mark.flaky(3, 1)
@pytest.mark.parametrize("default_bot", [{"parse_mode": "Markdown"}], indirect=True)
async def test_send_audio_default_parse_mode_1(self, default_bot, chat_id, audio_file):
test_string = "Italic Bold Code"
test_markdown_string = "_Italic_ *Bold* `Code`"
message = await default_bot.send_audio(chat_id, audio_file, caption=test_markdown_string)
assert message.caption_markdown == test_markdown_string
assert message.caption == test_string
@pytest.mark.flaky(3, 1)
@pytest.mark.parametrize("default_bot", [{"parse_mode": "Markdown"}], indirect=True)
async def test_send_audio_default_parse_mode_2(self, default_bot, chat_id, audio_file):
test_markdown_string = "_Italic_ *Bold* `Code`"
message = await default_bot.send_audio(
chat_id, audio_file, caption=test_markdown_string, parse_mode=None
)
assert message.caption == test_markdown_string
assert message.caption_markdown == escape_markdown(test_markdown_string)
@pytest.mark.flaky(3, 1)
@pytest.mark.parametrize("default_bot", [{"parse_mode": "Markdown"}], indirect=True)
async def test_send_audio_default_parse_mode_3(self, default_bot, chat_id, audio_file):
test_markdown_string = "_Italic_ *Bold* `Code`"
message = await default_bot.send_audio(
chat_id, audio_file, caption=test_markdown_string, parse_mode="HTML"
)
assert message.caption == test_markdown_string
assert message.caption_markdown == escape_markdown(test_markdown_string)
@pytest.mark.flaky(3, 1)
@pytest.mark.parametrize("default_bot", [{"protect_content": True}], indirect=True)
async def test_send_audio_default_protect_content(self, default_bot, chat_id, audio):
protected_audio = await default_bot.send_audio(chat_id, audio)
assert protected_audio.has_protected_content
unprotected = await default_bot.send_audio(chat_id, audio, protect_content=False)
assert not unprotected.has_protected_content
@pytest.mark.parametrize("local_mode", [True, False])
async def test_send_audio_local_files(self, monkeypatch, bot, chat_id, local_mode):
try:
bot._local_mode = local_mode
# For just test that the correct paths are passed as we have no local bot API set up
test_flag = False
file = data_file("telegram.jpg")
expected = file.as_uri()
async def make_assertion(_, data, *args, **kwargs):
nonlocal test_flag
if local_mode:
test_flag = data.get("audio") == expected and data.get("thumb") == expected
else:
test_flag = isinstance(data.get("audio"), InputFile) and isinstance(
data.get("thumb"), InputFile
)
monkeypatch.setattr(bot, "_post", make_assertion)
await bot.send_audio(chat_id, file, thumb=file)
assert test_flag
monkeypatch.delattr(bot, "_post")
finally:
bot._local_mode = False
def test_de_json(self, bot, audio):
json_dict = {
"file_id": self.audio_file_id,
"file_unique_id": self.audio_file_unique_id,
"duration": self.duration,
"performer": self.performer,
"title": self.title,
"file_name": self.file_name,
"mime_type": self.mime_type,
"file_size": self.file_size,
"thumb": audio.thumb.to_dict(),
}
json_audio = Audio.de_json(json_dict, bot)
assert json_audio.api_kwargs == {}
assert json_audio.file_id == self.audio_file_id
assert json_audio.file_unique_id == self.audio_file_unique_id
assert json_audio.duration == self.duration
assert json_audio.performer == self.performer
assert json_audio.title == self.title
assert json_audio.file_name == self.file_name
assert json_audio.mime_type == self.mime_type
assert json_audio.file_size == self.file_size
assert json_audio.thumb == audio.thumb
def test_to_dict(self, audio):
audio_dict = audio.to_dict()
assert isinstance(audio_dict, dict)
assert audio_dict["file_id"] == audio.file_id
assert audio_dict["file_unique_id"] == audio.file_unique_id
assert audio_dict["duration"] == audio.duration
assert audio_dict["mime_type"] == audio.mime_type
assert audio_dict["file_size"] == audio.file_size
assert audio_dict["file_name"] == audio.file_name
@pytest.mark.flaky(3, 1)
async def test_error_send_empty_file(self, bot, chat_id):
audio_file = open(os.devnull, "rb")
with pytest.raises(TelegramError):
await bot.send_audio(chat_id=chat_id, audio=audio_file)
@pytest.mark.flaky(3, 1)
async def test_error_send_empty_file_id(self, bot, chat_id):
with pytest.raises(TelegramError):
await bot.send_audio(chat_id=chat_id, audio="")
async def test_error_send_without_required_args(self, bot, chat_id):
with pytest.raises(TypeError):
await bot.send_audio(chat_id=chat_id)
async def test_get_file_instance_method(self, monkeypatch, audio):
async def make_assertion(*_, **kwargs):
return kwargs["file_id"] == audio.file_id
assert check_shortcut_signature(Audio.get_file, Bot.get_file, ["file_id"], [])
assert await check_shortcut_call(audio.get_file, audio.get_bot(), "get_file")
assert await check_defaults_handling(audio.get_file, audio.get_bot())
monkeypatch.setattr(audio._bot, "get_file", make_assertion)
assert await audio.get_file()
def test_equality(self, audio):
a = Audio(audio.file_id, audio.file_unique_id, audio.duration)
b = Audio("", audio.file_unique_id, audio.duration)
c = Audio(audio.file_id, audio.file_unique_id, 0)
d = Audio("", "", audio.duration)
e = Voice(audio.file_id, audio.file_unique_id, audio.duration)
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)