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
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
content / browser / speech / speech_recognition_browsertest.cc [blame]
// Copyright 2013 The Chromium Authors
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include <stddef.h>
#include <stdint.h>
#include <string.h>
#include <list>
#include <memory>
#include "base/containers/heap_array.h"
#include "base/containers/span.h"
#include "base/functional/bind.h"
#include "base/location.h"
#include "base/memory/raw_ptr.h"
#include "base/numerics/byte_conversions.h"
#include "base/numerics/safe_conversions.h"
#include "base/run_loop.h"
#include "base/strings/string_util.h"
#include "base/strings/utf_string_conversions.h"
#include "base/task/single_thread_task_runner.h"
#include "build/build_config.h"
#include "content/browser/speech/network_speech_recognition_engine_impl.h"
#include "content/browser/speech/speech_recognition_manager_impl.h"
#include "content/browser/speech/speech_recognizer_impl.h"
#include "content/public/browser/browser_task_traits.h"
#include "content/public/browser/browser_thread.h"
#include "content/public/browser/google_streaming_api.pb.h"
#include "content/public/browser/web_contents.h"
#include "content/public/test/browser_test.h"
#include "content/public/test/browser_test_utils.h"
#include "content/public/test/content_browser_test.h"
#include "content/public/test/content_browser_test_utils.h"
#include "content/public/test/test_navigation_observer.h"
#include "content/public/test/test_utils.h"
#include "content/shell/browser/shell.h"
#include "media/audio/audio_system.h"
#include "media/base/audio_capturer_source.h"
#include "media/base/audio_glitch_info.h"
#include "net/test/embedded_test_server/controllable_http_response.h"
#include "net/test/embedded_test_server/embedded_test_server.h"
#include "testing/gmock/include/gmock/gmock.h"
#include "testing/gtest/include/gtest/gtest.h"
#if !BUILDFLAG(IS_FUCHSIA)
#include "base/test/scoped_feature_list.h"
#include "components/soda/soda_util.h"
#include "content/browser/speech/fake_speech_recognition_manager_delegate.h"
#include "content/browser/speech/soda_speech_recognition_engine_impl.h"
#include "media/base/media_switches.h"
#include "media/mojo/mojom/audio_data.mojom.h"
#if BUILDFLAG(IS_CHROMEOS_ASH)
#include "ash/constants/ash_features.h"
#endif // BUILDFLAG(IS_CHROMEOS_ASH)
#endif // !BUILDFLAG(IS_FUCHSIA)
using base::RunLoop;
using CaptureCallback = media::AudioCapturerSource::CaptureCallback;
#if !BUILDFLAG(IS_FUCHSIA)
using testing::_;
using testing::InvokeWithoutArgs;
#endif // !BUILDFLAG(IS_FUCHSIA)
namespace content {
namespace {
#if !BUILDFLAG(IS_FUCHSIA)
const char kWebSpeechExpectGoodResult1[] = "Pictures of the moon";
const char kWebSpeechPageGoodResult1[] = "goodresult1";
#endif // !BUILDFLAG(IS_FUCHSIA)
// TODO(crbug.com/40575807) Use FakeSystemInfo instead.
class MockAudioSystem : public media::AudioSystem {
public:
MockAudioSystem() = default;
MockAudioSystem(const MockAudioSystem&) = delete;
MockAudioSystem& operator=(const MockAudioSystem&) = delete;
// AudioSystem implementation.
void GetInputStreamParameters(const std::string& device_id,
OnAudioParamsCallback on_params_cb) override {
DCHECK_CURRENTLY_ON(BrowserThread::IO);
// Posting callback to allow current SpeechRecognizerImpl dispatching event
// to complete before transitioning to the next FSM state.
content::GetIOThreadTaskRunner({})->PostTask(
FROM_HERE,
base::BindOnce(std::move(on_params_cb),
media::AudioParameters::UnavailableDeviceParams()));
}
MOCK_METHOD2(GetOutputStreamParameters,
void(const std::string& device_id,
OnAudioParamsCallback on_params_cb));
MOCK_METHOD1(HasInputDevices, void(OnBoolCallback on_has_devices_cb));
MOCK_METHOD1(HasOutputDevices, void(OnBoolCallback on_has_devices_cb));
MOCK_METHOD2(GetDeviceDescriptions,
void(bool for_input,
OnDeviceDescriptionsCallback on_descriptions_cp));
MOCK_METHOD2(GetAssociatedOutputDeviceID,
void(const std::string& input_device_id,
OnDeviceIdCallback on_device_id_cb));
MOCK_METHOD2(GetInputDeviceInfo,
void(const std::string& input_device_id,
OnInputDeviceInfoCallback on_input_device_info_cb));
};
class MockCapturerSource : public media::AudioCapturerSource {
public:
using StartCallback =
base::OnceCallback<void(const media::AudioParameters& audio_parameters,
CaptureCallback* capture_callback)>;
using StopCallback = base::OnceCallback<void()>;
MockCapturerSource(StartCallback start_callback, StopCallback stop_callback) {
start_callback_ = std::move(start_callback);
stop_callback_ = std::move(stop_callback);
}
MockCapturerSource(const MockCapturerSource&) = delete;
MockCapturerSource& operator=(const MockCapturerSource&) = delete;
void Initialize(const media::AudioParameters& params,
CaptureCallback* callback) override {
audio_parameters_ = params;
capture_callback_ = callback;
}
void Start() override {
std::move(start_callback_).Run(audio_parameters_, capture_callback_.get());
}
void Stop() override { std::move(stop_callback_).Run(); }
MOCK_METHOD1(SetAutomaticGainControl, void(bool enable));
MOCK_METHOD1(SetVolume, void(double volume));
MOCK_METHOD1(SetOutputDeviceForAec,
void(const std::string& output_device_id));
protected:
~MockCapturerSource() override = default;
private:
StartCallback start_callback_;
StopCallback stop_callback_;
raw_ptr<CaptureCallback, AcrossTasksDanglingUntriaged> capture_callback_;
media::AudioParameters audio_parameters_;
};
std::string MakeGoodResponse() {
proto::SpeechRecognitionEvent proto_event;
proto_event.set_status(proto::SpeechRecognitionEvent::STATUS_SUCCESS);
proto::SpeechRecognitionResult* proto_result = proto_event.add_result();
media::mojom::WebSpeechRecognitionResultPtr result =
media::mojom::WebSpeechRecognitionResult::New();
result->hypotheses.push_back(media::mojom::SpeechRecognitionHypothesis::New(
u"Pictures of the moon", 1.0F));
proto_result->set_final(!result->is_provisional);
for (size_t i = 0; i < result->hypotheses.size(); ++i) {
proto::SpeechRecognitionAlternative* proto_alternative =
proto_result->add_alternative();
const media::mojom::SpeechRecognitionHypothesisPtr& hypothesis =
result->hypotheses[i];
proto_alternative->set_confidence(hypothesis->confidence);
proto_alternative->set_transcript(base::UTF16ToUTF8(hypothesis->utterance));
}
std::string msg_string;
proto_event.SerializeToString(&msg_string);
// Prepend 4 byte prefix length indication to the protobuf message as
// envisaged by the google streaming recognition webservice protocol.
msg_string.insert(0u, base::as_string_view(base::U32ToBigEndian(
base::checked_cast<uint32_t>(msg_string.size()))));
return msg_string;
}
} // namespace
class SpeechRecognitionBrowserTest : public ContentBrowserTest {
public:
enum StreamingServerState {
kIdle,
kTestAudioCapturerSourceOpened,
kTestAudioCapturerSourceClosed,
};
#if !BUILDFLAG(IS_FUCHSIA)
SpeechRecognitionBrowserTest() {
// Setup the SODA On-Device feature flags.
scoped_feature_list_.InitWithFeatures(
/*enabled_features=*/
{
media::kOnDeviceWebSpeech,
#if BUILDFLAG(IS_CHROMEOS_ASH)
ash::features::kOnDeviceSpeechRecognition,
#endif // BUILDFLAG(IS_CHROMEOS_ASH)
},
/*disabled_features=*/{});
}
#endif // !BUILDFLAG(IS_FUCHSIA)
// Helper methods used by test fixtures.
GURL GetTestUrlFromFragment(const std::string& fragment) {
return GURL(GetTestUrl("speech", "web_speech_recognition.html").spec() +
"#" + fragment);
}
std::string GetPageFragment() {
return shell()->web_contents()->GetLastCommittedURL().ref();
}
const StreamingServerState &streaming_server_state() {
return streaming_server_state_;
}
protected:
// ContentBrowserTest methods.
void SetUpOnMainThread() override {
streaming_server_state_ = kIdle;
ASSERT_TRUE(SpeechRecognitionManagerImpl::GetInstance());
audio_system_ = std::make_unique<MockAudioSystem>();
audio_capturer_source_ = base::MakeRefCounted<MockCapturerSource>(
base::BindOnce(&SpeechRecognitionBrowserTest::OnCapturerSourceStart,
base::Unretained(this)),
base::BindOnce(&SpeechRecognitionBrowserTest::OnCapturerSourceStop,
base::Unretained(this)));
SpeechRecognizerImpl::SetAudioEnvironmentForTesting(
audio_system_.get(),
static_cast<media::AudioCapturerSource*>(audio_capturer_source_.get()));
}
void TearDownOnMainThread() override {
SpeechRecognizerImpl::SetAudioEnvironmentForTesting(nullptr, nullptr);
}
#if !BUILDFLAG(IS_FUCHSIA)
// Set SODA On-Device speech recognition features flags.
base::test::ScopedFeatureList scoped_feature_list_;
// Setup mock SODA installer
MockSodaInstaller mock_soda_installer_;
#endif // !BUILDFLAG(IS_FUCHSIA)
private:
void OnCapturerSourceStart(const media::AudioParameters& audio_parameters,
CaptureCallback* capture_callback) {
DCHECK_CURRENTLY_ON(BrowserThread::IO);
ASSERT_EQ(kIdle, streaming_server_state_);
streaming_server_state_ = kTestAudioCapturerSourceOpened;
const int capture_packet_interval_ms =
(1000 * audio_parameters.frames_per_buffer()) /
audio_parameters.sample_rate();
ASSERT_EQ(NetworkSpeechRecognitionEngineImpl::kAudioPacketIntervalMs,
capture_packet_interval_ms);
FeedAudioCapturerSource(audio_parameters, capture_callback, 500 /* ms */,
/*noise=*/false);
FeedAudioCapturerSource(audio_parameters, capture_callback, 1000 /* ms */,
/*noise=*/true);
FeedAudioCapturerSource(audio_parameters, capture_callback, 1000 /* ms */,
/*noise=*/false);
}
void OnCapturerSourceStop() {
DCHECK_CURRENTLY_ON(BrowserThread::IO);
ASSERT_EQ(kTestAudioCapturerSourceOpened, streaming_server_state_);
streaming_server_state_ = kTestAudioCapturerSourceClosed;
// Reset capturer source so SpeechRecognizerImpl destructor doesn't call
// AudioCaptureSourcer::Stop() again.
SpeechRecognizerImpl::SetAudioEnvironmentForTesting(nullptr, nullptr);
GetUIThreadTaskRunner({})->PostTask(
FROM_HERE, base::BindOnce(&SpeechRecognitionBrowserTest::SendResponse,
base::Unretained(this)));
}
void SendResponse() {}
static void FeedSingleBufferToAudioCapturerSource(
const media::AudioParameters& audio_params,
CaptureCallback* capture_callback,
size_t buffer_size,
bool fill_with_noise) {
DCHECK(capture_callback);
auto audio_buffer = base::HeapArray<uint8_t>::Uninit(buffer_size);
if (fill_with_noise) {
for (size_t i = 0; i < buffer_size; ++i)
audio_buffer[i] =
static_cast<uint8_t>(127 * sin(i * 3.14F / (16 * buffer_size)));
} else {
base::ranges::fill(audio_buffer, 0);
}
std::unique_ptr<media::AudioBus> audio_bus =
media::AudioBus::Create(audio_params);
audio_bus->FromInterleaved<media::SignedInt16SampleTypeTraits>(
reinterpret_cast<int16_t*>(&audio_buffer.data()[0]),
audio_bus->frames());
capture_callback->Capture(audio_bus.get(), base::TimeTicks::Now(), {}, 0.0,
false);
}
void FeedAudioCapturerSource(const media::AudioParameters& audio_params,
CaptureCallback* capture_callback,
int duration_ms,
bool feed_with_noise) {
const size_t buffer_size =
audio_params.GetBytesPerBuffer(media::kSampleFormatS16);
const int ms_per_buffer = audio_params.GetBufferDuration().InMilliseconds();
// We can only simulate durations that are integer multiples of the
// buffer size. In this regard see
// NetworkSpeechRecognitionEngineImpl::GetDesiredAudioChunkDurationMs().
ASSERT_EQ(0, duration_ms % ms_per_buffer);
const int n_buffers = duration_ms / ms_per_buffer;
for (int i = 0; i < n_buffers; ++i) {
base::SingleThreadTaskRunner::GetCurrentDefault()->PostTask(
FROM_HERE,
base::BindOnce(&FeedSingleBufferToAudioCapturerSource, audio_params,
capture_callback, buffer_size, feed_with_noise));
}
}
std::unique_ptr<media::AudioSystem> audio_system_;
scoped_refptr<MockCapturerSource> audio_capturer_source_;
StreamingServerState streaming_server_state_;
};
// Simply loads the test page and checks if it was able to create a Speech
// Recognition object in JavaScript, to make sure the Web Speech API is enabled.
// Flaky on all platforms. http://crbug.com/396414.
IN_PROC_BROWSER_TEST_F(SpeechRecognitionBrowserTest, DISABLED_Precheck) {
NavigateToURLBlockUntilNavigationsComplete(
shell(), GetTestUrlFromFragment("precheck"), 2);
EXPECT_EQ(kIdle, streaming_server_state());
EXPECT_EQ("success", GetPageFragment());
}
// Flaky on mac, see https://crbug.com/794645.
#if BUILDFLAG(IS_MAC)
#define MAYBE_OneShotRecognition DISABLED_OneShotRecognition
#else
#define MAYBE_OneShotRecognition OneShotRecognition
#endif
IN_PROC_BROWSER_TEST_F(SpeechRecognitionBrowserTest, MAYBE_OneShotRecognition) {
// Set up a test server, with two response handlers.
net::test_server::ControllableHttpResponse upstream_response(
embedded_test_server(), "/foo/up?", true /* relative_url_is_prefix */);
net::test_server::ControllableHttpResponse downstream_response(
embedded_test_server(), "/foo/down?", true /* relative_url_is_prefix */);
ASSERT_TRUE(embedded_test_server()->Start());
// Use a base path that doesn't end in a slash to mimic the default URL.
std::string web_service_base_url =
embedded_test_server()->base_url().spec() + "foo";
NetworkSpeechRecognitionEngineImpl::set_web_service_base_url_for_tests(
web_service_base_url.c_str());
// Need to watch for two navigations. Can't use
// NavigateToURLBlockUntilNavigationsComplete so that the
// ControllableHttpResponses can be used to wait for the test server to see
// the network requests, and response to them.
TestNavigationObserver navigation_observer(shell()->web_contents(), 2);
shell()->LoadURL(GetTestUrlFromFragment("oneshot"));
// Wait for the upstream HTTP request to be completely received, and return an
// empty response.
upstream_response.WaitForRequest();
EXPECT_FALSE(upstream_response.http_request()->content.empty());
EXPECT_EQ(net::test_server::METHOD_POST,
upstream_response.http_request()->method);
EXPECT_EQ("chunked",
upstream_response.http_request()->headers.at("Transfer-Encoding"));
EXPECT_EQ("audio/x-flac; rate=16000",
upstream_response.http_request()->headers.at("Content-Type"));
upstream_response.Send("HTTP/1.1 200 OK\r\n\r\n");
upstream_response.Done();
// Wait for the downstream HTTP request to be received, and response with a
// valid response.
downstream_response.WaitForRequest();
EXPECT_EQ(net::test_server::METHOD_GET,
downstream_response.http_request()->method);
downstream_response.Send("HTTP/1.1 200 OK\r\n\r\n" + MakeGoodResponse());
downstream_response.Done();
navigation_observer.Wait();
EXPECT_EQ(kTestAudioCapturerSourceClosed, streaming_server_state());
EXPECT_EQ("goodresult1", GetPageFragment());
// Remove reference to URL string that's on the stack.
NetworkSpeechRecognitionEngineImpl::set_web_service_base_url_for_tests(
nullptr);
}
#if !BUILDFLAG(IS_FUCHSIA)
IN_PROC_BROWSER_TEST_F(SpeechRecognitionBrowserTest,
OnDeviceWebSpeechRecognition) {
// Speech On-Device not supported.
if (!speech::IsOnDeviceSpeechRecognitionSupported()) {
return;
}
std::unique_ptr<MockOnDeviceWebSpeechRecognitionService> mock_speech_service =
std::make_unique<MockOnDeviceWebSpeechRecognitionService>(
shell()->web_contents()->GetBrowserContext());
std::unique_ptr<FakeSpeechRecognitionManagerDelegate>
fake_speech_recognition_mgr_delegate =
std::make_unique<FakeSpeechRecognitionManagerDelegate>(
mock_speech_service.get());
SodaSpeechRecognitionEngineImpl::
SetSpeechRecognitionManagerDelegateForTesting(
fake_speech_recognition_mgr_delegate.get());
mock_soda_installer_.NotifySodaInstalledForTesting();
mock_soda_installer_.NotifySodaInstalledForTesting(
speech::LanguageCode::kEnUs);
EXPECT_CALL(mock_soda_installer_, GetAvailableLanguages())
.WillRepeatedly(InvokeWithoutArgs([]() {
std::vector<std::string> langs;
langs.push_back("en-US");
return langs;
}));
bool has_reponsed = false;
EXPECT_CALL(*mock_speech_service, SendAudioToSpeechRecognitionService(_))
.WillRepeatedly(testing::Invoke([&](media::mojom::AudioDataS16Ptr data) {
if (!has_reponsed) {
has_reponsed = true;
media::SpeechRecognitionResult result =
media::SpeechRecognitionResult(kWebSpeechExpectGoodResult1, true);
GetIOThreadTaskRunner({})->PostTask(
FROM_HERE,
base::BindOnce(&MockOnDeviceWebSpeechRecognitionService::
SendSpeechRecognitionResult,
mock_speech_service->GetWeakPtr(),
std::move(result)));
}
}));
TestNavigationObserver navigation_observer(shell()->web_contents(), 2);
shell()->LoadURL(GetTestUrlFromFragment("oneshot"));
navigation_observer.Wait();
EXPECT_EQ(kTestAudioCapturerSourceClosed, streaming_server_state());
EXPECT_EQ(kWebSpeechPageGoodResult1, GetPageFragment());
base::RunLoop().RunUntilIdle();
// clean
SodaSpeechRecognitionEngineImpl::
SetSpeechRecognitionManagerDelegateForTesting(nullptr);
fake_speech_recognition_mgr_delegate->Reset(nullptr);
fake_speech_recognition_mgr_delegate.reset();
// Clear raw_ptr<content::BrowserContext> before object released.
GetIOThreadTaskRunner({})->PostTask(
FROM_HERE,
base::BindOnce([](std::unique_ptr<MockOnDeviceWebSpeechRecognitionService>
mock_service) { mock_service.reset(); },
std::move(mock_speech_service)));
base::RunLoop().RunUntilIdle();
}
#endif // !BUILDFLAG(IS_FUCHSIA)
} // namespace content