| OLD | NEW |
| 1 // Copyright (c) 2011 The Chromium Authors. All rights reserved. | 1 // Copyright (c) 2011 The Chromium Authors. All rights reserved. |
| 2 // Use of this source code is governed by a BSD-style license that can be | 2 // Use of this source code is governed by a BSD-style license that can be |
| 3 // found in the LICENSE file. | 3 // found in the LICENSE file. |
| 4 | 4 |
| 5 #include "content/browser/speech/speech_recognition_request.h" | 5 #include "content/browser/speech/speech_recognition_request.h" |
| 6 | 6 |
| 7 #include <vector> | 7 #include <vector> |
| 8 | 8 |
| 9 #include "base/json/json_reader.h" | 9 #include "base/json/json_reader.h" |
| 10 #include "base/string_number_conversions.h" | 10 #include "base/string_number_conversions.h" |
| 11 #include "base/string_util.h" | 11 #include "base/string_util.h" |
| 12 #include "base/values.h" | 12 #include "base/values.h" |
| 13 #include "net/base/escape.h" | 13 #include "net/base/escape.h" |
| 14 #include "net/base/load_flags.h" | 14 #include "net/base/load_flags.h" |
| 15 #include "net/url_request/url_request_context.h" | 15 #include "net/url_request/url_request_context.h" |
| 16 #include "net/url_request/url_request_context_getter.h" | 16 #include "net/url_request/url_request_context_getter.h" |
| 17 #include "net/url_request/url_request_status.h" | 17 #include "net/url_request/url_request_status.h" |
| 18 | 18 |
| 19 namespace { | 19 namespace { |
| 20 | 20 |
| 21 const char* const kDefaultSpeechRecognitionUrl = | 21 const char* const kDefaultSpeechRecognitionUrl = |
| 22 "https://www.google.com/speech-api/v1/recognize?xjerr=1&client=chromium&"; | 22 "https://www.google.com/speech-api/v1/recognize?xjerr=1&client=chromium&"; |
| 23 const char* const kStatusString = "status"; |
| 23 const char* const kHypothesesString = "hypotheses"; | 24 const char* const kHypothesesString = "hypotheses"; |
| 24 const char* const kUtteranceString = "utterance"; | 25 const char* const kUtteranceString = "utterance"; |
| 25 const char* const kConfidenceString = "confidence"; | 26 const char* const kConfidenceString = "confidence"; |
| 26 | 27 |
| 27 // TODO(satish): Remove this hardcoded value once the page is allowed to | 28 // TODO(satish): Remove this hardcoded value once the page is allowed to |
| 28 // set this via an attribute. | 29 // set this via an attribute. |
| 29 const int kMaxResults = 6; | 30 const int kMaxResults = 6; |
| 30 | 31 |
| 31 bool ParseServerResponse(const std::string& response_body, | 32 bool ParseServerResponse(const std::string& response_body, |
| 32 speech_input::SpeechInputResultArray* result) { | 33 speech_input::SpeechInputResult* result) { |
| 33 if (response_body.empty()) { | 34 if (response_body.empty()) { |
| 34 LOG(WARNING) << "ParseServerResponse: Response was empty."; | 35 LOG(WARNING) << "ParseServerResponse: Response was empty."; |
| 35 return false; | 36 return false; |
| 36 } | 37 } |
| 37 DVLOG(1) << "ParseServerResponse: Parsing response " << response_body; | 38 DVLOG(1) << "ParseServerResponse: Parsing response " << response_body; |
| 38 | 39 |
| 39 // Parse the response, ignoring comments. | 40 // Parse the response, ignoring comments. |
| 40 std::string error_msg; | 41 std::string error_msg; |
| 41 scoped_ptr<Value> response_value(base::JSONReader::ReadAndReturnError( | 42 scoped_ptr<Value> response_value(base::JSONReader::ReadAndReturnError( |
| 42 response_body, false, NULL, &error_msg)); | 43 response_body, false, NULL, &error_msg)); |
| 43 if (response_value == NULL) { | 44 if (response_value == NULL) { |
| 44 LOG(WARNING) << "ParseServerResponse: JSONReader failed : " << error_msg; | 45 LOG(WARNING) << "ParseServerResponse: JSONReader failed : " << error_msg; |
| 45 return false; | 46 return false; |
| 46 } | 47 } |
| 47 | 48 |
| 48 if (!response_value->IsType(Value::TYPE_DICTIONARY)) { | 49 if (!response_value->IsType(Value::TYPE_DICTIONARY)) { |
| 49 VLOG(1) << "ParseServerResponse: Unexpected response type " | 50 VLOG(1) << "ParseServerResponse: Unexpected response type " |
| 50 << response_value->GetType(); | 51 << response_value->GetType(); |
| 51 return false; | 52 return false; |
| 52 } | 53 } |
| 53 const DictionaryValue* response_object = | 54 const DictionaryValue* response_object = |
| 54 static_cast<DictionaryValue*>(response_value.get()); | 55 static_cast<DictionaryValue*>(response_value.get()); |
| 55 | 56 |
| 56 // Get the hypotheses | 57 // Get the status. |
| 58 int status; |
| 59 if (!response_object->GetInteger(kStatusString, &status)) { |
| 60 VLOG(1) << "ParseServerResponse: " << kStatusString |
| 61 << " is not a valid integer value."; |
| 62 return false; |
| 63 } |
| 64 |
| 65 // Process the status. |
| 66 switch (status) { |
| 67 case speech_input::kErrorNone: |
| 68 case speech_input::kErrorNoSpeech: |
| 69 case speech_input::kErrorNoMatch: |
| 70 break; |
| 71 |
| 72 default: |
| 73 // Other status codes should not be returned by the server. |
| 74 VLOG(1) << "ParseServerResponse: unexpected status code " << status; |
| 75 return false; |
| 76 } |
| 77 |
| 78 result->error = static_cast<speech_input::SpeechInputError>(status); |
| 79 |
| 80 // Get the hypotheses. |
| 57 Value* hypotheses_value = NULL; | 81 Value* hypotheses_value = NULL; |
| 58 if (!response_object->Get(kHypothesesString, &hypotheses_value)) { | 82 if (!response_object->Get(kHypothesesString, &hypotheses_value)) { |
| 59 VLOG(1) << "ParseServerResponse: Missing hypotheses attribute."; | 83 VLOG(1) << "ParseServerResponse: Missing hypotheses attribute."; |
| 60 return false; | 84 return false; |
| 61 } | 85 } |
| 86 |
| 62 DCHECK(hypotheses_value); | 87 DCHECK(hypotheses_value); |
| 63 if (!hypotheses_value->IsType(Value::TYPE_LIST)) { | 88 if (!hypotheses_value->IsType(Value::TYPE_LIST)) { |
| 64 VLOG(1) << "ParseServerResponse: Unexpected hypotheses type " | 89 VLOG(1) << "ParseServerResponse: Unexpected hypotheses type " |
| 65 << hypotheses_value->GetType(); | 90 << hypotheses_value->GetType(); |
| 66 return false; | 91 return false; |
| 67 } | 92 } |
| 93 |
| 68 const ListValue* hypotheses_list = static_cast<ListValue*>(hypotheses_value); | 94 const ListValue* hypotheses_list = static_cast<ListValue*>(hypotheses_value); |
| 69 | 95 |
| 70 size_t index = 0; | 96 size_t index = 0; |
| 71 for (; index < hypotheses_list->GetSize(); ++index) { | 97 for (; index < hypotheses_list->GetSize(); ++index) { |
| 72 Value* hypothesis = NULL; | 98 Value* hypothesis = NULL; |
| 73 if (!hypotheses_list->Get(index, &hypothesis)) { | 99 if (!hypotheses_list->Get(index, &hypothesis)) { |
| 74 LOG(WARNING) << "ParseServerResponse: Unable to read hypothesis value."; | 100 LOG(WARNING) << "ParseServerResponse: Unable to read hypothesis value."; |
| 75 break; | 101 break; |
| 76 } | 102 } |
| 77 DCHECK(hypothesis); | 103 DCHECK(hypothesis); |
| 78 if (!hypothesis->IsType(Value::TYPE_DICTIONARY)) { | 104 if (!hypothesis->IsType(Value::TYPE_DICTIONARY)) { |
| 79 LOG(WARNING) << "ParseServerResponse: Unexpected value type " | 105 LOG(WARNING) << "ParseServerResponse: Unexpected value type " |
| 80 << hypothesis->GetType(); | 106 << hypothesis->GetType(); |
| 81 break; | 107 break; |
| 82 } | 108 } |
| 83 | 109 |
| 84 const DictionaryValue* hypothesis_value = | 110 const DictionaryValue* hypothesis_value = |
| 85 static_cast<DictionaryValue*>(hypothesis); | 111 static_cast<DictionaryValue*>(hypothesis); |
| 86 string16 utterance; | 112 string16 utterance; |
| 87 if (!hypothesis_value->GetString(kUtteranceString, &utterance)) { | 113 if (!hypothesis_value->GetString(kUtteranceString, &utterance)) { |
| 88 LOG(WARNING) << "ParseServerResponse: Missing utterance value."; | 114 LOG(WARNING) << "ParseServerResponse: Missing utterance value."; |
| 89 break; | 115 break; |
| 90 } | 116 } |
| 91 | 117 |
| 92 // It is not an error if the 'confidence' field is missing. | 118 // It is not an error if the 'confidence' field is missing. |
| 93 double confidence = 0.0; | 119 double confidence = 0.0; |
| 94 hypothesis_value->GetDouble(kConfidenceString, &confidence); | 120 hypothesis_value->GetDouble(kConfidenceString, &confidence); |
| 95 | 121 |
| 96 result->push_back(speech_input::SpeechInputResultItem(utterance, | 122 result->hypotheses.push_back(speech_input::SpeechInputHypothesis( |
| 97 confidence)); | 123 utterance, confidence)); |
| 98 } | 124 } |
| 99 | 125 |
| 100 if (index < hypotheses_list->GetSize()) { | 126 if (index < hypotheses_list->GetSize()) { |
| 101 result->clear(); | 127 result->hypotheses.clear(); |
| 102 return false; | 128 return false; |
| 103 } | 129 } |
| 104 | 130 |
| 105 return true; | 131 return true; |
| 106 } | 132 } |
| 107 | 133 |
| 108 } // namespace | 134 } // namespace |
| 109 | 135 |
| 110 namespace speech_input { | 136 namespace speech_input { |
| 111 | 137 |
| (...skipping 63 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... |
| 175 | 201 |
| 176 void SpeechRecognitionRequest::UploadAudioChunk(const std::string& audio_data, | 202 void SpeechRecognitionRequest::UploadAudioChunk(const std::string& audio_data, |
| 177 bool is_last_chunk) { | 203 bool is_last_chunk) { |
| 178 DCHECK(url_fetcher_.get()); | 204 DCHECK(url_fetcher_.get()); |
| 179 url_fetcher_->AppendChunkToUpload(audio_data, is_last_chunk); | 205 url_fetcher_->AppendChunkToUpload(audio_data, is_last_chunk); |
| 180 } | 206 } |
| 181 | 207 |
| 182 void SpeechRecognitionRequest::OnURLFetchComplete(const URLFetcher* source) { | 208 void SpeechRecognitionRequest::OnURLFetchComplete(const URLFetcher* source) { |
| 183 DCHECK_EQ(url_fetcher_.get(), source); | 209 DCHECK_EQ(url_fetcher_.get(), source); |
| 184 | 210 |
| 185 bool error = | 211 SpeechInputResult result; |
| 186 !source->status().is_success() || source->response_code() != 200; | 212 if (!source->status().is_success() || source->response_code() != 200 || |
| 187 | 213 !ParseServerResponse(source->GetResponseStringRef(), &result)) { |
| 188 SpeechInputResultArray result; | 214 result.error = kErrorNetwork; |
| 189 if (!error) | 215 } |
| 190 error = !ParseServerResponse(source->GetResponseStringRef(), &result); | |
| 191 url_fetcher_.reset(); | |
| 192 | 216 |
| 193 DVLOG(1) << "SpeechRecognitionRequest: Invoking delegate with result."; | 217 DVLOG(1) << "SpeechRecognitionRequest: Invoking delegate with result."; |
| 194 delegate_->SetRecognitionResult(error, result); | 218 url_fetcher_.reset(); |
| 219 delegate_->SetRecognitionResult(result); |
| 195 } | 220 } |
| 196 | 221 |
| 197 } // namespace speech_input | 222 } // namespace speech_input |
| OLD | NEW |