File: tts_client_impl.cc

package info (click to toggle)
chromium 139.0.7258.127-1~deb13u1
  • links: PTS, VCS
  • area: main
  • in suites:
  • size: 6,122,096 kB
  • sloc: cpp: 35,100,771; ansic: 7,163,530; javascript: 4,103,002; python: 1,436,920; asm: 946,517; xml: 746,709; pascal: 187,653; perl: 88,691; sh: 88,436; objc: 79,953; sql: 51,488; cs: 44,583; fortran: 24,137; makefile: 22,147; tcl: 15,277; php: 13,980; yacc: 8,984; ruby: 7,485; awk: 3,720; lisp: 3,096; lex: 1,327; ada: 727; jsp: 228; sed: 36
file content (219 lines) | stat: -rw-r--r-- 8,089 bytes parent folder | download | duplicates (7)
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
// Copyright 2023 The Chromium Authors
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.

#include "chrome/browser/ash/accessibility/service/tts_client_impl.h"

#include "content/public/browser/browser_context.h"
#include "content/public/browser/tts_controller.h"
#include "services/accessibility/public/mojom/tts.mojom.h"
#include "ui/base/l10n/l10n_util.h"

namespace ash {

namespace {

// The max utterance length allowed by the TTS extension API.
const int kMaxUtteranceLength = 32768;
// The minimum speech rate allowed by the TTS extension API.
const double kMinRate = 0.1;
// The maximum speech rate allowed by the TTS extension API.
const double kMaxRate = 10.0;
// The maximum speech pitch allowed by the TTS extension API.
const double kMaxPitch = 2.0;
// The maximum speech volume allowed by the TTS extension API.
const double kMaxVolume = 1.0;

ax::mojom::TtsEventType ToMojo(content::TtsEventType event_type) {
  switch (event_type) {
    case content::TTS_EVENT_START:
      return ax::mojom::TtsEventType::kStart;
    case content::TTS_EVENT_END:
      return ax::mojom::TtsEventType::kEnd;
    case content::TTS_EVENT_WORD:
      return ax::mojom::TtsEventType::kWord;
    case content::TTS_EVENT_SENTENCE:
      return ax::mojom::TtsEventType::kSentence;
    case content::TTS_EVENT_MARKER:
      return ax::mojom::TtsEventType::kMarker;
    case content::TTS_EVENT_INTERRUPTED:
      return ax::mojom::TtsEventType::kInterrupted;
    case content::TTS_EVENT_CANCELLED:
      return ax::mojom::TtsEventType::kCancelled;
    case content::TTS_EVENT_ERROR:
      return ax::mojom::TtsEventType::kError;
    case content::TTS_EVENT_PAUSE:
      return ax::mojom::TtsEventType::kPause;
    case content::TTS_EVENT_RESUME:
      return ax::mojom::TtsEventType::kResume;
  }
}

// Self-owned, compare to TtsExtensionEventHandler.
class AtpTtsEventHandler : public content::UtteranceEventDelegate {
 public:
  // static creator deals with "new" so clients don't have to think about it.
  static AtpTtsEventHandler* Create() { return new AtpTtsEventHandler(); }
  ~AtpTtsEventHandler() override = default;
  AtpTtsEventHandler(const AtpTtsEventHandler&) = delete;
  AtpTtsEventHandler& operator=(const AtpTtsEventHandler&) = delete;

  // content::UtteranceEventDelegate:
  void OnTtsEvent(content::TtsUtterance* utterance,
                  content::TtsEventType event_type,
                  int char_index,
                  int length,
                  const std::string& error_message) override {
    auto mojom_event = ax::mojom::TtsEvent::New();
    mojom_event->type = ToMojo(event_type);
    mojom_event->char_index = char_index;
    mojom_event->length = length;
    mojom_event->is_final = utterance->IsFinished();
    if (event_type == content::TTS_EVENT_ERROR) {
      mojom_event->error_message = error_message;
    }
    utterance_client_->OnEvent(std::move(mojom_event));
    if (utterance->IsFinished()) {
      // Expected to self-destroy on call to TtsEvent, see
      // tts_utterance_impl.cc.
      delete this;
    }
  }
  mojo::PendingReceiver<ax::mojom::TtsUtteranceClient> PassReceiver() {
    return utterance_client_.BindNewPipeAndPassReceiver();
  }

 private:
  AtpTtsEventHandler() = default;
  mojo::Remote<ax::mojom::TtsUtteranceClient> utterance_client_;
};

}  // namespace

TtsClientImpl::TtsClientImpl(content::BrowserContext* profile)
    : profile_(profile) {
  CHECK(profile_);
}

TtsClientImpl::~TtsClientImpl() = default;

void TtsClientImpl::Bind(mojo::PendingReceiver<Tts> tts_receiver) {
  tts_receivers_.Add(this, std::move(tts_receiver));
}

void TtsClientImpl::Speak(const std::string& utterance,
                          ax::mojom::TtsOptionsPtr options,
                          SpeakCallback callback) {
  auto result = ax::mojom::TtsSpeakResult::New();
  if (utterance.size() > kMaxUtteranceLength) {
    result->error = ax::mojom::TtsError::kErrorUtteranceTooLong;
    std::move(callback).Run(std::move(result));
    return;
  }

  // Check for errors in options.
  // TODO(crbug.com/41278287): Centralize the struct validation.
  if (options->rate < kMinRate || options->rate > kMaxRate) {
    result->error = ax::mojom::TtsError::kErrorInvalidRate;
    std::move(callback).Run(std::move(result));
    return;
  }
  if (options->pitch < 0.0 || options->pitch > kMaxPitch) {
    result->error = ax::mojom::TtsError::kErrorInvalidPitch;
    std::move(callback).Run(std::move(result));
    return;
  }
  if (options->volume < 0.0 || options->volume > kMaxVolume) {
    result->error = ax::mojom::TtsError::kErrorInvalidVolume;
    std::move(callback).Run(std::move(result));
    return;
  }

  // Only make the utterance once we know we aren't going to return early.
  std::unique_ptr<content::TtsUtterance> tts_utterance =
      content::TtsUtterance::Create(profile_);
  tts_utterance->SetText(utterance);
  // TODO(b:277221897): Pass a fake GURL matching the ash extension URL.
  // This will support both UMA and using enhanced network voices in ATP
  // select-to-speak.
  tts_utterance->SetSrcUrl(GURL(""));

  tts_utterance->SetContinuousParameters(options->rate, options->pitch,
                                         options->volume);
  tts_utterance->SetShouldClearQueue(!options->enqueue);
  if (options->lang) {
    std::string lang = options->lang.value();
    if (!lang.empty() && !l10n_util::IsValidLocaleSyntax(lang)) {
      result->error = ax::mojom::TtsError::kErrorInvalidLang;
      std::move(callback).Run(std::move(result));
      return;
    }
    tts_utterance->SetLang(options->lang.value());
  }
  if (options->voice_name) {
    tts_utterance->SetVoiceName(options->voice_name.value());
  }
  if (options->engine_id) {
    tts_utterance->SetEngineId(options->engine_id.value());
  }
  if (options->on_event) {
    auto* atpTtsEventHandler = AtpTtsEventHandler::Create();
    result->utterance_client = atpTtsEventHandler->PassReceiver();
    tts_utterance->SetEventDelegate(atpTtsEventHandler);
  }
  // Note: we don't need desired/required event types because they aren't
  // passed by ChromeVox or STS. We don't need an options_dict, it's redundant,
  // and we don't need a src_id because each ATP utterance has its own callback.

  // Send the callback back to ATP with the utterance client.
  result->error = ax::mojom::TtsError::kNoError;
  std::move(callback).Run(std::move(result));

  // Start speech.
  content::TtsController* controller = content::TtsController::GetInstance();
  controller->SpeakOrEnqueue(std::move(tts_utterance));
}

void TtsClientImpl::Stop() {
  content::TtsController* controller = content::TtsController::GetInstance();
  // TODO(b:277221897): Pass a fake GURL matching the ash extension URL so that
  // extensions cannot clobber other speech.
  controller->Stop(GURL(""));
}

void TtsClientImpl::Pause() {
  content::TtsController::GetInstance()->Pause();
}

void TtsClientImpl::Resume() {
  content::TtsController::GetInstance()->Resume();
}

void TtsClientImpl::IsSpeaking(IsSpeakingCallback callback) {
  std::move(callback).Run(content::TtsController::GetInstance()->IsSpeaking());
}

void TtsClientImpl::GetVoices(GetVoicesCallback callback) {
  std::vector<content::VoiceData> voices;
  // TODO(b:277221897): Pass a fake GURL matching the extension URL so that
  // Select to Speak can get the enhanced network voices.
  content::TtsController::GetInstance()->GetVoices(profile_, GURL(""), &voices);
  std::vector<ax::mojom::TtsVoicePtr> results;
  for (auto& voice : voices) {
    auto result = ax::mojom::TtsVoice::New();
    result->voice_name = voice.name;
    result->lang = voice.lang;
    result->remote = voice.remote;
    result->engine_id = voice.engine_id;
    if (!voice.events.empty()) {
      result->event_types = std::vector<ax::mojom::TtsEventType>();
      for (auto type : voice.events) {
        result->event_types->emplace_back(ToMojo(type));
      }
    }
    results.emplace_back(std::move(result));
  }
  std::move(callback).Run(std::move(results));
}

}  // namespace ash