50 changed files with 13121 additions and 0 deletions
@ -0,0 +1,88 @@ |
|||
/// 返回阿里云通义千问的翻译语言配置表（Translation） |
|||
/// |
|||
/// 数据来源：阿里云通义千问翻译支持语种 |
|||
/// 参考： https://bailian.console.aliyun.com/cn-beijing/?spm=5176.12818093_47.overview_recent.2.3be916d0Dc1NWf&tab=doc#/doc/?type=model&url=2983281 |
|||
List<Map<String, String>> getAlibabaTranslationLanguageSpecs() { |
|||
return [ |
|||
// 中文系列 |
|||
{ |
|||
'shortCode': 'zh', |
|||
'chineseName': '中文', |
|||
'englishName': 'Chinese', |
|||
'translationCode': 'zh', |
|||
'ttsVoiceName': 'Cherry', |
|||
}, |
|||
|
|||
// 常见通用语种 |
|||
{ |
|||
'shortCode': 'en', |
|||
'chineseName': '英语', |
|||
'englishName': 'English', |
|||
'translationCode': 'en', |
|||
'ttsVoiceName': 'Cherry', |
|||
}, |
|||
{ |
|||
'shortCode': 'ja', |
|||
'chineseName': '日语', |
|||
'englishName': 'Japanese', |
|||
'translationCode': 'ja', |
|||
'ttsVoiceName': 'Cherry', |
|||
}, |
|||
{ |
|||
'shortCode': 'ko', |
|||
'chineseName': '韩语', |
|||
'englishName': 'Korean', |
|||
'translationCode': 'ko', |
|||
'ttsVoiceName': 'Cherry', |
|||
}, |
|||
{ |
|||
'shortCode': 'fr', |
|||
'chineseName': '法语', |
|||
'englishName': 'French', |
|||
'translationCode': 'fr', |
|||
'ttsVoiceName': 'Cherry', |
|||
}, |
|||
{ |
|||
'shortCode': 'de', |
|||
'chineseName': '德语', |
|||
'englishName': 'German', |
|||
'translationCode': 'de', |
|||
'ttsVoiceName': 'Cherry', |
|||
}, |
|||
{ |
|||
'shortCode': 'es', |
|||
'chineseName': '西班牙语', |
|||
'englishName': 'Spanish', |
|||
'translationCode': 'es', |
|||
'ttsVoiceName': 'Cherry', |
|||
}, |
|||
{ |
|||
'shortCode': 'it', |
|||
'chineseName': '意大利语', |
|||
'englishName': 'Italian', |
|||
'translationCode': 'it', |
|||
'ttsVoiceName': 'Cherry', |
|||
}, |
|||
{ |
|||
'shortCode': 'pt', |
|||
'chineseName': '葡萄牙语', |
|||
'englishName': 'Portuguese', |
|||
'translationCode': 'pt', |
|||
'ttsVoiceName': 'Cherry', |
|||
}, |
|||
{ |
|||
'shortCode': 'ru', |
|||
'chineseName': '俄语', |
|||
'englishName': 'Russian', |
|||
'translationCode': 'ru', |
|||
'ttsVoiceName': 'Cherry', |
|||
}, |
|||
|
|||
// { |
|||
// 'shortCode': 'yue', |
|||
// 'chineseName': '粤语', |
|||
// 'englishName': 'Cantonese', |
|||
// 'translationCode': 'yue', |
|||
// }, |
|||
]; |
|||
} |
|||
@ -0,0 +1,423 @@ |
|||
/// 返回AST配置表 |
|||
List<Map<String, String>> getAstLanguageConfigs() { |
|||
return [ |
|||
// 中文及方言 |
|||
{ |
|||
'shortCode': 'zh', |
|||
'chineseName': '中文', |
|||
'englishName': 'Chinese', |
|||
}, |
|||
|
|||
{ |
|||
'shortCode': 'yue', |
|||
'chineseName': '粤语', |
|||
'englishName': 'Cantonese', |
|||
}, |
|||
|
|||
// 英语各地区 |
|||
{ |
|||
'shortCode': 'en', |
|||
'chineseName': '英语', |
|||
'englishName': 'English', |
|||
}, |
|||
|
|||
// 西班牙语各地区 |
|||
{ |
|||
'shortCode': 'es', |
|||
'chineseName': '西班牙语', |
|||
'englishName': 'Spanish', |
|||
}, |
|||
|
|||
// 其他欧洲语言 |
|||
{ |
|||
'shortCode': 'de', |
|||
'chineseName': '德语', |
|||
'englishName': 'German', |
|||
}, |
|||
|
|||
{ |
|||
'shortCode': 'fr', |
|||
'chineseName': '法语', |
|||
'englishName': 'French', |
|||
}, |
|||
{ |
|||
'shortCode': 'ru', |
|||
'chineseName': '俄语', |
|||
'englishName': 'Russian', |
|||
}, |
|||
{ |
|||
'shortCode': 'it', |
|||
'chineseName': '意大利语', |
|||
'englishName': 'Italian', |
|||
}, |
|||
{ |
|||
'shortCode': 'nl', |
|||
'chineseName': '荷兰语', |
|||
'englishName': 'Dutch', |
|||
}, |
|||
{ |
|||
'shortCode': 'nb', |
|||
'chineseName': '挪威语', |
|||
'englishName': 'Norwegian Bokmål', |
|||
}, |
|||
{ |
|||
'shortCode': 'da', |
|||
'chineseName': '丹麦语', |
|||
'englishName': 'Danish', |
|||
}, |
|||
{ |
|||
'shortCode': 'fi', |
|||
'chineseName': '芬兰语', |
|||
'englishName': 'Finnish', |
|||
}, |
|||
{ |
|||
'shortCode': 'sv', |
|||
'chineseName': '瑞典语', |
|||
'englishName': 'Swedish', |
|||
}, |
|||
{ |
|||
'shortCode': 'pl', |
|||
'chineseName': '波兰语', |
|||
'englishName': 'Polish', |
|||
}, |
|||
{ |
|||
'shortCode': 'tr', |
|||
'chineseName': '土耳其语', |
|||
'englishName': 'Turkish', |
|||
'asrCode': 'tr-TR', |
|||
}, |
|||
{ |
|||
'shortCode': 'el', |
|||
'chineseName': '希腊语', |
|||
'englishName': 'Greek', |
|||
}, |
|||
{ |
|||
'shortCode': 'cs', |
|||
'chineseName': '捷克语', |
|||
'englishName': 'Czech', |
|||
}, |
|||
{ |
|||
'shortCode': 'ca', |
|||
'chineseName': '加泰罗尼亚语', |
|||
'englishName': 'Catalan', |
|||
}, |
|||
{ |
|||
'shortCode': 'cy', |
|||
'chineseName': '威尔士语', |
|||
'englishName': 'Welsh', |
|||
}, |
|||
{ |
|||
'shortCode': 'bg', |
|||
'chineseName': '保加利亚语', |
|||
'englishName': 'Bulgarian', |
|||
}, |
|||
|
|||
// 亚洲语言 |
|||
{ |
|||
'shortCode': 'ja', |
|||
'chineseName': '日语', |
|||
'englishName': 'Japanese', |
|||
}, |
|||
{ |
|||
'shortCode': 'ko', |
|||
'chineseName': '韩语', |
|||
'englishName': 'Korean', |
|||
}, |
|||
{ |
|||
'shortCode': 'hi', |
|||
'chineseName': '印地语', |
|||
'englishName': 'Hindi', |
|||
}, |
|||
{ |
|||
'shortCode': 'th', |
|||
'chineseName': '泰语', |
|||
'englishName': 'Thai', |
|||
}, |
|||
{ |
|||
'shortCode': 'vi', |
|||
'chineseName': '越南语', |
|||
'englishName': 'Vietnamese', |
|||
}, |
|||
{ |
|||
'shortCode': 'id', |
|||
'chineseName': '印尼语', |
|||
'englishName': 'Indonesian', |
|||
}, |
|||
{ |
|||
'shortCode': 'ms', |
|||
'chineseName': '马来语', |
|||
'englishName': 'Malay', |
|||
}, |
|||
{ |
|||
'shortCode': 'bn', |
|||
'chineseName': '孟加拉语', |
|||
'englishName': 'Bengali', |
|||
}, |
|||
{ |
|||
'shortCode': 'am', |
|||
'chineseName': '阿姆哈拉语', |
|||
'englishName': 'Amharic', |
|||
'asrCode': 'am-ET', |
|||
}, |
|||
{ |
|||
'shortCode': 'as', |
|||
'chineseName': '阿萨姆语', |
|||
'englishName': 'Assamese', |
|||
}, |
|||
|
|||
// 中东与非洲 |
|||
{ |
|||
'shortCode': 'ar', |
|||
'chineseName': '阿拉伯语', |
|||
'englishName': 'Arabic', |
|||
}, |
|||
|
|||
{ |
|||
'shortCode': 'af', |
|||
'chineseName': '南非荷兰语', |
|||
'englishName': 'Afrikaans', |
|||
}, |
|||
|
|||
// 其他 |
|||
{ |
|||
'shortCode': 'az', |
|||
'chineseName': '阿塞拜疆语', |
|||
'englishName': 'Azerbaijani', |
|||
}, |
|||
{ |
|||
'shortCode': 'bs', |
|||
'chineseName': '波斯尼亚语', |
|||
'englishName': 'Bosnian', |
|||
}, |
|||
{ |
|||
'shortCode': 'kn', |
|||
'chineseName': '卡纳达语', |
|||
'englishName': 'kn-IN', |
|||
}, |
|||
{ |
|||
'shortCode': 'et', |
|||
'chineseName': '爱沙尼亚语', |
|||
'englishName': 'Estonian', |
|||
}, |
|||
{ |
|||
'shortCode': 'eu', |
|||
'chineseName': '巴斯克语', |
|||
'englishName': 'Basque', |
|||
}, |
|||
{ |
|||
'shortCode': 'fa', |
|||
'chineseName': '波斯语', |
|||
'englishName': 'Persian', |
|||
}, |
|||
{ |
|||
'shortCode': 'fil', |
|||
'chineseName': '菲律宾语', |
|||
'englishName': 'Filipino', |
|||
}, |
|||
{ |
|||
'shortCode': 'ga', |
|||
'chineseName': '爱尔兰语', |
|||
'englishName': 'Ireland', |
|||
}, |
|||
{ |
|||
'shortCode': 'gl', |
|||
'chineseName': '加利西亚语', |
|||
'englishName': 'Galician', |
|||
}, |
|||
{ |
|||
'shortCode': 'gu', |
|||
'chineseName': '古吉拉特语', |
|||
'englishName': 'Gujarati', |
|||
}, |
|||
{ |
|||
'shortCode': 'he', |
|||
'chineseName': '希伯来语', |
|||
'englishName': 'Hebrew', |
|||
}, |
|||
{ |
|||
'shortCode': 'hr', |
|||
'chineseName': '克罗地亚语', |
|||
'englishName': 'Croatian', |
|||
}, |
|||
{ |
|||
'shortCode': 'hu', |
|||
'chineseName': '匈牙利语', |
|||
'englishName': 'Hungarian', |
|||
}, |
|||
{ |
|||
'shortCode': 'hy', |
|||
'chineseName': '亚美尼亚语', |
|||
'englishName': 'Armenian', |
|||
}, |
|||
{ |
|||
'shortCode': 'is', |
|||
'chineseName': '冰岛语', |
|||
'englishName': 'Icelandic', |
|||
}, |
|||
{ |
|||
'shortCode': 'jv', |
|||
'chineseName': '爪哇语', |
|||
'englishName': 'Javanese', |
|||
}, |
|||
{ |
|||
'shortCode': 'ka', |
|||
'chineseName': '格鲁吉亚语', |
|||
'englishName': 'Georgian', |
|||
}, |
|||
{ |
|||
'shortCode': 'kk', |
|||
'chineseName': '哈萨克语', |
|||
'englishName': 'Kazakh', |
|||
}, |
|||
{ |
|||
'shortCode': 'km', |
|||
'chineseName': '高棉语', |
|||
'englishName': 'Khmer', |
|||
}, |
|||
{ |
|||
'shortCode': 'lo', |
|||
'chineseName': '老挝语', |
|||
'englishName': 'Lao', |
|||
}, |
|||
{ |
|||
'shortCode': 'lt', |
|||
'chineseName': '立陶宛语', |
|||
'englishName': 'Lithuanian', |
|||
}, |
|||
{ |
|||
'shortCode': 'lv', |
|||
'chineseName': '拉脱维亚语', |
|||
'englishName': 'Latvian', |
|||
}, |
|||
{ |
|||
'shortCode': 'mk', |
|||
'chineseName': '马其顿语', |
|||
'englishName': 'Macedonian', |
|||
}, |
|||
{ |
|||
'shortCode': 'ml', |
|||
'chineseName': '马拉雅拉姆语', |
|||
'englishName': 'Malayalam', |
|||
}, |
|||
{ |
|||
'shortCode': 'mn', |
|||
'chineseName': '蒙古语', |
|||
'englishName': 'Mongolian', |
|||
}, |
|||
{ |
|||
'shortCode': 'mr', |
|||
'chineseName': '马拉地语', |
|||
'englishName': 'Marathi', |
|||
}, |
|||
{ |
|||
'shortCode': 'mt', |
|||
'chineseName': '马耳他语', |
|||
'englishName': 'Maltese', |
|||
}, |
|||
{ |
|||
'shortCode': 'my', |
|||
'chineseName': '缅甸语', |
|||
'englishName': 'Burmese', |
|||
}, |
|||
{ |
|||
'shortCode': 'nb', |
|||
'chineseName': '书面挪威语', |
|||
'englishName': 'Norwegian Bokmål', |
|||
}, |
|||
{ |
|||
'shortCode': 'ne', |
|||
'chineseName': '尼泊尔语', |
|||
'englishName': 'Nepali', |
|||
}, |
|||
// { |
|||
// 'shortCode': 'or', |
|||
// 'chineseName': '奥里亚语', |
|||
// 'englishName': 'Oriya', |
|||
// }, |
|||
// { |
|||
// 'shortCode': 'pa', |
|||
// 'chineseName': '旁遮普语', |
|||
// 'englishName': 'Punjabi', |
|||
// }, |
|||
{ |
|||
'shortCode': 'ps', |
|||
'chineseName': '普什图语', |
|||
'englishName': 'Pushto', |
|||
}, |
|||
{ |
|||
'shortCode': 'pt', |
|||
'chineseName': '葡萄牙语', |
|||
'englishName': 'Portuguese', |
|||
}, |
|||
{ |
|||
'shortCode': 'ro', |
|||
'chineseName': '罗马尼亚语', |
|||
'englishName': 'Romanian', |
|||
}, |
|||
{ |
|||
'shortCode': 'si', |
|||
'chineseName': '僧伽罗语', |
|||
'englishName': 'Sinhala', |
|||
}, |
|||
{ |
|||
'shortCode': 'sk', |
|||
'chineseName': '斯洛伐克语', |
|||
'englishName': 'Slovak', |
|||
}, |
|||
{ |
|||
'shortCode': 'sl', |
|||
'chineseName': '斯洛文尼亚语', |
|||
'englishName': 'Slovenian', |
|||
}, |
|||
{ |
|||
'shortCode': 'so', |
|||
'chineseName': '索马里语', |
|||
'englishName': 'Somali', |
|||
}, |
|||
{ |
|||
'shortCode': 'sq', |
|||
'chineseName': '阿尔巴尼亚语', |
|||
'englishName': 'Albanian', |
|||
}, |
|||
{ |
|||
'shortCode': 'sr', |
|||
'chineseName': '塞尔维亚语', |
|||
'englishName': 'Serbian', |
|||
}, |
|||
{ |
|||
'shortCode': 'sv', |
|||
'chineseName': '斯瓦希里语', |
|||
'englishName': 'Swedish', |
|||
}, |
|||
{ |
|||
'shortCode': 'ta', |
|||
'chineseName': '泰米尔语', |
|||
'englishName': 'Tamil', |
|||
}, |
|||
{ |
|||
'shortCode': 'te', |
|||
'chineseName': '泰卢固语', |
|||
'englishName': 'Tamil', |
|||
}, |
|||
{ |
|||
'shortCode': 'uk', |
|||
'chineseName': '乌克兰语', |
|||
'englishName': 'Ukrainian', |
|||
}, |
|||
{ |
|||
'shortCode': 'ur', |
|||
'chineseName': '乌尔都语', |
|||
'englishName': 'Urdu', |
|||
}, |
|||
{ |
|||
'shortCode': 'uz', |
|||
'chineseName': '乌兹别克语', |
|||
'englishName': 'Uzbek', |
|||
}, |
|||
{ |
|||
'shortCode': 'zu', |
|||
'chineseName': '祖鲁语', |
|||
'englishName': 'Zulu', |
|||
}, |
|||
]; |
|||
} |
|||
File diff suppressed because it is too large
@ -0,0 +1,296 @@ |
|||
/// 返回谷歌的识别语言配置表（ASR） |
|||
List<Map<String, String>> getGoogleASRLanguageSpecs() { |
|||
return [ |
|||
{ |
|||
'shortCode': 'zh', |
|||
'chineseName': '中文', |
|||
'englishName': 'Chinese (simplified)', |
|||
'asrCode': 'zh-CN', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'zh-Hant', |
|||
'chineseName': '中文(繁体)', |
|||
'englishName': 'Chinese (traditional)', |
|||
'asrCode': 'zh-TW', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'zh-Hant-hk', |
|||
'chineseName': '中文(香港繁体)', |
|||
'englishName': 'Chinese (Hongkong traditional)', |
|||
'asrCode': 'zh-HK', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'yue', |
|||
'chineseName': '广东话', |
|||
'englishName': 'Cantonese', |
|||
'asrCode': 'yue-HK', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'en', |
|||
'chineseName': '英语', |
|||
'englishName': 'English', |
|||
'asrCode': 'en-US', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'ja', |
|||
'chineseName': '日语', |
|||
'englishName': 'Japanese', |
|||
'asrCode': 'ja-JP', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'ko', |
|||
'chineseName': '韩语', |
|||
'englishName': 'Korean', |
|||
'asrCode': 'ko-KR', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'fr', |
|||
'chineseName': '法语', |
|||
'englishName': 'French', |
|||
'asrCode': 'fr-FR', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'es', |
|||
'chineseName': '西班牙语', |
|||
'englishName': 'Spanish', |
|||
'asrCode': 'es-ES', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'pt', |
|||
'chineseName': '葡萄牙语', |
|||
'englishName': 'Portuguese', |
|||
'asrCode': 'pt-BR', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'it', |
|||
'chineseName': '意大利语', |
|||
'englishName': 'Italian', |
|||
'asrCode': 'it-IT', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'de', |
|||
'chineseName': '德语', |
|||
'englishName': 'German', |
|||
'asrCode': 'de-DE', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'ru', |
|||
'chineseName': '俄语', |
|||
'englishName': 'Russian', |
|||
'asrCode': 'ru-RU', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'pl', |
|||
'chineseName': '波兰语', |
|||
'englishName': 'Polish', |
|||
'asrCode': 'pl-PL', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'nl', |
|||
'chineseName': '荷兰语', |
|||
'englishName': 'Dutch', |
|||
'asrCode': 'nl-NL', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'sv', |
|||
'chineseName': '瑞典语', |
|||
'englishName': 'Swedish', |
|||
'asrCode': 'sv-SE', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'cs', |
|||
'chineseName': '捷克语', |
|||
'englishName': 'Czech', |
|||
'asrCode': 'cs-CZ', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'el', |
|||
'chineseName': '现代希腊语', |
|||
'englishName': 'Modern Greek', |
|||
'asrCode': 'el-GR', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'ro', |
|||
'chineseName': '罗马尼亚语', |
|||
'englishName': 'Romanian', |
|||
'asrCode': 'ro-RO', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'hu', |
|||
'chineseName': '匈牙利语', |
|||
'englishName': 'Hungarian', |
|||
'asrCode': 'hu-HU', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'uk', |
|||
'chineseName': '乌克兰语', |
|||
'englishName': 'Ukrainian', |
|||
'asrCode': 'uk-UA', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'da', |
|||
'chineseName': '丹麦语', |
|||
'englishName': 'Danish', |
|||
'asrCode': 'da-DK', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'fi', |
|||
'chineseName': '芬兰语', |
|||
'englishName': 'Finnish', |
|||
'asrCode': 'fi-FI', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'no', |
|||
'chineseName': '挪威语', |
|||
'englishName': 'Norwegian', |
|||
'asrCode': 'nb-NO', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'hr', |
|||
'chineseName': '克罗地亚语', |
|||
'englishName': 'Croatian', |
|||
'asrCode': 'hr-HR', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'ca', |
|||
'chineseName': '加泰隆语', |
|||
'englishName': 'Catalan', |
|||
'asrCode': 'ca-ES', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'ar', |
|||
'chineseName': '阿拉伯语', |
|||
'englishName': 'Arabic', |
|||
'asrCode': 'ar-EG', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'th', |
|||
'chineseName': '泰语', |
|||
'englishName': 'Thai', |
|||
'asrCode': 'th-TH', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'vi', |
|||
'chineseName': '越南语', |
|||
'englishName': 'Vietnamese', |
|||
'asrCode': 'vi-VN', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'id', |
|||
'chineseName': '印尼语', |
|||
'englishName': 'Indonesian', |
|||
'asrCode': 'id-ID', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'hi', |
|||
'chineseName': '印地语', |
|||
'englishName': 'Hindi', |
|||
'asrCode': 'hi-IN', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'tr', |
|||
'chineseName': '土耳其语', |
|||
'englishName': 'Turkish', |
|||
'asrCode': 'tr-TR', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'ms', |
|||
'chineseName': '马来语', |
|||
'englishName': 'Malay', |
|||
'asrCode': 'ms-MY', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'bn', |
|||
'chineseName': '孟加拉语', |
|||
'englishName': 'Bengali', |
|||
'asrCode': 'bn-IN', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'ta', |
|||
'chineseName': '泰米尔语', |
|||
'englishName': 'Tamil', |
|||
'asrCode': 'ta-IN', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'te', |
|||
'chineseName': '泰卢固语', |
|||
'englishName': 'Telugu', |
|||
'asrCode': 'te-IN', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'mr', |
|||
'chineseName': '马拉提语', |
|||
'englishName': 'Marathi', |
|||
'asrCode': 'mr-IN', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'ur', |
|||
'chineseName': '乌尔都语', |
|||
'englishName': 'Urdu', |
|||
'asrCode': 'ur-IN', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'fa', |
|||
'chineseName': '波斯语', |
|||
'englishName': 'Persian', |
|||
'asrCode': 'fa-IR', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
{ |
|||
'shortCode': 'sw', |
|||
'chineseName': '斯瓦希里语', |
|||
'englishName': 'Swahili', |
|||
'asrCode': 'sw-KE', |
|||
'ttsVoiceName': '' |
|||
}, |
|||
]; |
|||
} |
|||
|
|||
/// 返回谷歌的翻译语言配置表（Translation） |
|||
List<Map<String, String>> getGoogleTranslationLanguageSpecs() { |
|||
// 谷歌翻译能力覆盖广泛，采用与ASR相同的语言集合 |
|||
return getGoogleASRLanguageSpecs(); |
|||
} |
|||
|
|||
/// 返回谷歌的TTS语言配置表（TTS，保留空语音名以示未配置） |
|||
List<Map<String, String>> getGoogleTTSLanguageSpecs() { |
|||
return getGoogleASRLanguageSpecs(); |
|||
} |
|||
@ -0,0 +1,807 @@ |
|||
/// 返回科大讯飞的识别语言配置表（ASR，大模型） |
|||
/// |
|||
/// 说明：基于“实时语音转写大模型”，`lang` 参数采用如下两种模式： |
|||
/// - `autodialect`：支持中文+英文及202种方言的免切识别 |
|||
/// - `autominor`：支持37种多语种的免切识别（需开通对应能力） |
|||
/// |
|||
/// 配置策略： |
|||
/// - 中文、英文及各中文方言的 `asrCode` 统一映射为 `autodialect` |
|||
/// - 其他小语种的 `asrCode` 统一映射为 `autominor` |
|||
List<Map<String, String>> getIflytekASRLanguageSpecs() { |
|||
return [ |
|||
// 中文/英文（中英+方言免切） |
|||
{ |
|||
'shortCode': 'zh', |
|||
'chineseName': '中文', |
|||
'englishName': 'Chinese', |
|||
'asrCode': 'zh-CN', |
|||
}, |
|||
{ |
|||
'shortCode': 'en', |
|||
'chineseName': '英语', |
|||
'englishName': 'English', |
|||
'asrCode': 'en-US', |
|||
}, |
|||
|
|||
// 中文方言（示例列出常用的11种方言） |
|||
{ |
|||
'shortCode': 'yue', |
|||
'chineseName': '粤语', |
|||
'englishName': 'Cantonese', |
|||
'asrCode': 'yue-HK', |
|||
}, |
|||
{ |
|||
'shortCode': 'sichuan', |
|||
'chineseName': '四川话', |
|||
'englishName': 'Sichuan Dialect', |
|||
'asrCode': 'zh-CN', |
|||
}, |
|||
{ |
|||
'shortCode': 'hunan', |
|||
'chineseName': '湖南话', |
|||
'englishName': 'Hunan Dialect', |
|||
'asrCode': 'zh-CN', |
|||
}, |
|||
{ |
|||
'shortCode': 'henan', |
|||
'chineseName': '河南话', |
|||
'englishName': 'Henan Dialect', |
|||
'asrCode': 'zh-CN', |
|||
}, |
|||
{ |
|||
'shortCode': 'dongbei', |
|||
'chineseName': '东北话', |
|||
'englishName': 'Northeastern Dialect', |
|||
'asrCode': 'zh-CN', |
|||
}, |
|||
{ |
|||
'shortCode': 'shaanxi', |
|||
'chineseName': '陕西话', |
|||
'englishName': 'Shaanxi Dialect', |
|||
'asrCode': 'zh-CN', |
|||
}, |
|||
{ |
|||
'shortCode': 'tw', |
|||
'chineseName': '台湾普通话', |
|||
'englishName': 'Taiwan Mandarin', |
|||
'asrCode': 'zh-TW', |
|||
}, |
|||
{ |
|||
'shortCode': 'shandong', |
|||
'chineseName': '山东话', |
|||
'englishName': 'Shandong Dialect', |
|||
'asrCode': 'zh-CN', |
|||
}, |
|||
{ |
|||
'shortCode': 'hubei', |
|||
'chineseName': '湖北话', |
|||
'englishName': 'Hubei Dialect', |
|||
'asrCode': 'zh-CN', |
|||
}, |
|||
{ |
|||
'shortCode': 'hefei', |
|||
'chineseName': '合肥话', |
|||
'englishName': 'Hefei Dialect', |
|||
'asrCode': 'zh-CN', |
|||
}, |
|||
{ |
|||
'shortCode': 'neimenggu', |
|||
'chineseName': '内蒙古方言', |
|||
'englishName': 'Inner Mongolia Dialect', |
|||
'asrCode': 'zh-CN', |
|||
}, |
|||
|
|||
// 37种多语种（免切识别） |
|||
{ |
|||
'shortCode': 'ko', |
|||
'chineseName': '韩语', |
|||
'englishName': 'Korean', |
|||
'asrCode': 'ko-KR', |
|||
}, |
|||
{ |
|||
'shortCode': 'ja', |
|||
'chineseName': '日语', |
|||
'englishName': 'Japanese', |
|||
'asrCode': 'ja-JP', |
|||
}, |
|||
{ |
|||
'shortCode': 'fr', |
|||
'chineseName': '法语', |
|||
'englishName': 'French', |
|||
'asrCode': 'fr-FR', |
|||
}, |
|||
{ |
|||
'shortCode': 'ru', |
|||
'chineseName': '俄语', |
|||
'englishName': 'Russian', |
|||
'asrCode': 'ru-RU', |
|||
}, |
|||
{ |
|||
'shortCode': 'es', |
|||
'chineseName': '西班牙语', |
|||
'englishName': 'Spanish', |
|||
'asrCode': 'es-ES', |
|||
}, |
|||
{ |
|||
'shortCode': 'hi', |
|||
'chineseName': '印地语', |
|||
'englishName': 'Hindi', |
|||
'asrCode': 'hi-IN', |
|||
}, |
|||
{ |
|||
'shortCode': 'de', |
|||
'chineseName': '德语', |
|||
'englishName': 'German', |
|||
'asrCode': 'de-DE', |
|||
}, |
|||
{ |
|||
'shortCode': 'vi', |
|||
'chineseName': '越南语', |
|||
'englishName': 'Vietnamese', |
|||
'asrCode': 'vi-VN', |
|||
}, |
|||
{ |
|||
'shortCode': 'pt', |
|||
'chineseName': '葡萄牙语', |
|||
'englishName': 'Portuguese', |
|||
'asrCode': 'pt-PT', |
|||
}, |
|||
{ |
|||
'shortCode': 'pt-BR', |
|||
'chineseName': '巴西葡萄牙语', |
|||
'englishName': 'Brazilian Portuguese', |
|||
'asrCode': 'pt-BR', |
|||
}, |
|||
{ |
|||
'shortCode': 'it', |
|||
'chineseName': '意大利语', |
|||
'englishName': 'Italian', |
|||
'asrCode': 'it-IT', |
|||
}, |
|||
{ |
|||
'shortCode': 'th', |
|||
'chineseName': '泰语', |
|||
'englishName': 'Thai', |
|||
'asrCode': 'th-TH', |
|||
}, |
|||
{ |
|||
'shortCode': 'ur', |
|||
'chineseName': '乌尔都语', |
|||
'englishName': 'Urdu', |
|||
'asrCode': 'ur-PK', |
|||
}, |
|||
{ |
|||
'shortCode': 'pl', |
|||
'chineseName': '波兰语', |
|||
'englishName': 'Polish', |
|||
'asrCode': 'pl-PL', |
|||
}, |
|||
{ |
|||
'shortCode': 'ms', |
|||
'chineseName': '马来语', |
|||
'englishName': 'Malay', |
|||
'asrCode': 'ms-MY', |
|||
}, |
|||
{ |
|||
'shortCode': 'ar', |
|||
'chineseName': '阿拉伯语', |
|||
'englishName': 'Arabic', |
|||
'asrCode': 'ar-SA', |
|||
}, |
|||
{ |
|||
'shortCode': 'kk', |
|||
'chineseName': '哈萨克语', |
|||
'englishName': 'Kazakh', |
|||
'asrCode': 'kk-KZ', |
|||
}, |
|||
{ |
|||
'shortCode': 'sv', |
|||
'chineseName': '瑞典语', |
|||
'englishName': 'Swedish', |
|||
'asrCode': 'sv-SE', |
|||
}, |
|||
{ |
|||
'shortCode': 'fa', |
|||
'chineseName': '波斯语', |
|||
'englishName': 'Persian', |
|||
'asrCode': 'fa-IR', |
|||
}, |
|||
{ |
|||
'shortCode': 'bg', |
|||
'chineseName': '保加利亚语', |
|||
'englishName': 'Bulgarian', |
|||
'asrCode': 'bg-BG', |
|||
}, |
|||
{ |
|||
'shortCode': 'bo', |
|||
'chineseName': '藏语', |
|||
'englishName': 'Tibetan', |
|||
'asrCode': 'bo-CN', |
|||
}, |
|||
{ |
|||
'shortCode': 'uz', |
|||
'chineseName': '乌兹别克语', |
|||
'englishName': 'Uzbek', |
|||
'asrCode': 'uz-UZ', |
|||
}, |
|||
{ |
|||
'shortCode': 'nl', |
|||
'chineseName': '荷兰语', |
|||
'englishName': 'Dutch', |
|||
'asrCode': 'nl-NL', |
|||
}, |
|||
{ |
|||
'shortCode': 'uk', |
|||
'chineseName': '乌克兰语', |
|||
'englishName': 'Ukrainian', |
|||
'asrCode': 'uk-UA', |
|||
}, |
|||
{ |
|||
'shortCode': 'ta', |
|||
'chineseName': '泰米尔语', |
|||
'englishName': 'Tamil', |
|||
'asrCode': 'ta-IN', |
|||
}, |
|||
{ |
|||
'shortCode': 'sw', |
|||
'chineseName': '斯瓦希里语', |
|||
'englishName': 'Swahili', |
|||
'asrCode': 'sw-KE', |
|||
}, |
|||
{ |
|||
'shortCode': 'ro', |
|||
'chineseName': '罗马尼亚语', |
|||
'englishName': 'Romanian', |
|||
'asrCode': 'ro-RO', |
|||
}, |
|||
{ |
|||
'shortCode': 'ha', |
|||
'chineseName': '豪萨语', |
|||
'englishName': 'Hausa', |
|||
'asrCode': 'ha-NG', |
|||
}, |
|||
{ |
|||
'shortCode': 'cs', |
|||
'chineseName': '捷克语', |
|||
'englishName': 'Czech', |
|||
'asrCode': 'cs-CZ', |
|||
}, |
|||
{ |
|||
'shortCode': 'el', |
|||
'chineseName': '希腊语', |
|||
'englishName': 'Greek', |
|||
'asrCode': 'el-GR', |
|||
}, |
|||
{ |
|||
'shortCode': 'bn', |
|||
'chineseName': '孟加拉语', |
|||
'englishName': 'Bengali', |
|||
'asrCode': 'bn-IN', |
|||
}, |
|||
{ |
|||
'shortCode': 'id', |
|||
'chineseName': '印尼语', |
|||
'englishName': 'Indonesian', |
|||
'asrCode': 'id-ID', |
|||
}, |
|||
{ |
|||
'shortCode': 'fil', |
|||
'chineseName': '菲律宾语', |
|||
'englishName': 'Filipino', |
|||
'asrCode': 'fil-PH', |
|||
}, |
|||
{ |
|||
'shortCode': 'tr', |
|||
'chineseName': '土耳其语', |
|||
'englishName': 'Turkish', |
|||
'asrCode': 'tr-TR', |
|||
}, |
|||
{ |
|||
'shortCode': 'ug', |
|||
'chineseName': '维吾尔语', |
|||
'englishName': 'Uyghur', |
|||
'asrCode': 'ug-CN', |
|||
}, |
|||
{ |
|||
'shortCode': 'mn', |
|||
'chineseName': '蒙古语', |
|||
'englishName': 'Mongolian', |
|||
'asrCode': 'mn-MN', |
|||
}, |
|||
]; |
|||
} |
|||
|
|||
/// 返回科大讯飞的翻译语言配置表（Translation） |
|||
List<Map<String, String>> getIflytekTranslationLanguageSpecs() { |
|||
return [ |
|||
{ |
|||
'shortCode': 'zh', |
|||
'chineseName': '中文', |
|||
'englishName': 'Chinese', |
|||
'translationCode': 'cn', |
|||
}, |
|||
{ |
|||
'shortCode': 'yue', |
|||
'chineseName': '粤语', |
|||
'englishName': 'Cantonese', |
|||
'translationCode': 'yue', |
|||
}, |
|||
{ |
|||
'shortCode': 'en', |
|||
'chineseName': '英语', |
|||
'englishName': 'English', |
|||
'translationCode': 'en', |
|||
}, |
|||
{ |
|||
'shortCode': 'es', |
|||
'chineseName': '西班牙语', |
|||
'englishName': 'Spanish', |
|||
'translationCode': 'es', |
|||
}, |
|||
{ |
|||
'shortCode': 'de', |
|||
'chineseName': '德语', |
|||
'englishName': 'German', |
|||
'translationCode': 'de', |
|||
}, |
|||
{ |
|||
'shortCode': 'fr', |
|||
'chineseName': '法语', |
|||
'englishName': 'French', |
|||
'translationCode': 'fr', |
|||
}, |
|||
{ |
|||
'shortCode': 'it', |
|||
'chineseName': '意大利语', |
|||
'englishName': 'Italian', |
|||
'translationCode': 'it', |
|||
}, |
|||
{ |
|||
'shortCode': 'nl', |
|||
'chineseName': '荷兰语', |
|||
'englishName': 'Dutch', |
|||
'translationCode': 'nl', |
|||
}, |
|||
{ |
|||
'shortCode': 'nb', |
|||
'chineseName': '挪威语', |
|||
'englishName': 'Norwegian Bokmål', |
|||
'translationCode': 'nb', |
|||
}, |
|||
{ |
|||
'shortCode': 'da', |
|||
'chineseName': '丹麦语', |
|||
'englishName': 'Danish', |
|||
'translationCode': 'da', |
|||
}, |
|||
{ |
|||
'shortCode': 'fi', |
|||
'chineseName': '芬兰语', |
|||
'englishName': 'Finnish', |
|||
'translationCode': 'fi', |
|||
}, |
|||
{ |
|||
'shortCode': 'sv', |
|||
'chineseName': '瑞典语', |
|||
'englishName': 'Swedish', |
|||
'translationCode': 'sv-SE', |
|||
}, |
|||
{ |
|||
'shortCode': 'pl', |
|||
'chineseName': '波兰语', |
|||
'englishName': 'Polish', |
|||
'translationCode': 'pl-PL', |
|||
}, |
|||
{ |
|||
'shortCode': 'tr', |
|||
'chineseName': '土耳其语', |
|||
'englishName': 'Turkish', |
|||
'translationCode': 'tr', |
|||
}, |
|||
{ |
|||
'shortCode': 'el', |
|||
'chineseName': '希腊语', |
|||
'englishName': 'Greek', |
|||
'translationCode': 'el', |
|||
}, |
|||
{ |
|||
'shortCode': 'cs', |
|||
'chineseName': '捷克语', |
|||
'englishName': 'Czech', |
|||
'translationCode': 'cs-CZ', |
|||
}, |
|||
{ |
|||
'shortCode': 'ca', |
|||
'chineseName': '加泰罗尼亚语', |
|||
'englishName': 'Catalan', |
|||
'translationCode': 'ca-ES', |
|||
}, |
|||
{ |
|||
'shortCode': 'cy', |
|||
'chineseName': '威尔士语', |
|||
'englishName': 'Welsh', |
|||
'translationCode': 'cy', |
|||
}, |
|||
{ |
|||
'shortCode': 'bg', |
|||
'chineseName': '保加利亚语', |
|||
'englishName': 'Bulgarian', |
|||
'translationCode': 'bg', |
|||
}, |
|||
{ |
|||
'shortCode': 'ja', |
|||
'chineseName': '日语', |
|||
'englishName': 'Japanese', |
|||
'translationCode': 'ja', |
|||
}, |
|||
{ |
|||
'shortCode': 'ko', |
|||
'chineseName': '韩语', |
|||
'englishName': 'Korean', |
|||
'translationCode': 'ko', |
|||
}, |
|||
{ |
|||
'shortCode': 'hi', |
|||
'chineseName': '印地语', |
|||
'englishName': 'Hindi', |
|||
'translationCode': 'hi', |
|||
}, |
|||
{ |
|||
'shortCode': 'th', |
|||
'chineseName': '泰语', |
|||
'englishName': 'Thai', |
|||
'translationCode': 'th', |
|||
}, |
|||
{ |
|||
'shortCode': 'vi', |
|||
'chineseName': '越南语', |
|||
'englishName': 'Vietnamese', |
|||
'translationCode': 'vi', |
|||
}, |
|||
{ |
|||
'shortCode': 'id', |
|||
'chineseName': '印尼语', |
|||
'englishName': 'Indonesian', |
|||
'translationCode': 'id', |
|||
}, |
|||
{ |
|||
'shortCode': 'ms', |
|||
'chineseName': '马来语', |
|||
'englishName': 'Malay', |
|||
'translationCode': 'ms', |
|||
}, |
|||
{ |
|||
'shortCode': 'bn', |
|||
'chineseName': '孟加拉语', |
|||
'englishName': 'Bengali', |
|||
'translationCode': 'bn', |
|||
}, |
|||
{ |
|||
'shortCode': 'am', |
|||
'chineseName': '阿姆哈拉语', |
|||
'englishName': 'Amharic', |
|||
'translationCode': 'am', |
|||
}, |
|||
{ |
|||
'shortCode': 'as', |
|||
'chineseName': '阿萨姆语', |
|||
'englishName': 'Assamese', |
|||
'translationCode': 'as', |
|||
}, |
|||
{ |
|||
'shortCode': 'ar', |
|||
'chineseName': '阿拉伯语', |
|||
'englishName': 'Arabic', |
|||
'translationCode': 'ar', |
|||
}, |
|||
{ |
|||
'shortCode': 'bs', |
|||
'chineseName': '波斯尼亚语(波黑)', |
|||
'englishName': 'Bosnian (Bosnia and Herzegovina)', |
|||
'translationCode': 'bs', |
|||
}, |
|||
]; |
|||
} |
|||
|
|||
/// 返回科大讯飞的TTS语言配置表（TTS） |
|||
List<Map<String, String>> getIflytekTTSLanguageSpecs() { |
|||
return [ |
|||
// 基础语种 |
|||
{ |
|||
'shortCode': 'zh', |
|||
'chineseName': '中文', |
|||
'englishName': 'Chinese', |
|||
'ttsVoiceName': 'x4_yezi', |
|||
}, |
|||
{ |
|||
'shortCode': 'en', |
|||
'chineseName': '英语', |
|||
'englishName': 'English', |
|||
'ttsVoiceName': 'x4_enus_luna_assist', |
|||
}, |
|||
{ |
|||
'shortCode': 'ko', |
|||
'chineseName': '韩语', |
|||
'englishName': 'Korean', |
|||
'ttsVoiceName': 'zhimin', |
|||
}, |
|||
{ |
|||
'shortCode': 'ja', |
|||
'chineseName': '日语', |
|||
'englishName': 'Japanese', |
|||
'ttsVoiceName': 'qianhui', |
|||
}, |
|||
{ |
|||
'shortCode': 'fr', |
|||
'chineseName': '法语', |
|||
'englishName': 'French', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
{ |
|||
'shortCode': 'ru', |
|||
'chineseName': '俄语', |
|||
'englishName': 'Russian', |
|||
'ttsVoiceName': 'x2_RuRu_Keshu', |
|||
}, |
|||
{ |
|||
'shortCode': 'es', |
|||
'chineseName': '西班牙语', |
|||
'englishName': 'Spanish', |
|||
'ttsVoiceName': 'gabriela', |
|||
}, |
|||
{ |
|||
'shortCode': 'hi', |
|||
'chineseName': '印地语', |
|||
'englishName': 'Hindi', |
|||
'ttsVoiceName': 'x2_HiIn_Mohita', |
|||
}, |
|||
{ |
|||
'shortCode': 'de', |
|||
'chineseName': '德语', |
|||
'englishName': 'German', |
|||
'ttsVoiceName': 'x2_DeDe_Christiane', |
|||
}, |
|||
{ |
|||
'shortCode': 'vi', |
|||
'chineseName': '越南语', |
|||
'englishName': 'Vietnamese', |
|||
'ttsVoiceName': 'xiaoyun', |
|||
}, |
|||
{ |
|||
'shortCode': 'pt', |
|||
'chineseName': '葡萄牙语', |
|||
'englishName': 'Portuguese', |
|||
'ttsVoiceName': 'maria', |
|||
}, |
|||
{ |
|||
'shortCode': 'pt-BR', |
|||
'chineseName': '巴西葡萄牙语', |
|||
'englishName': 'Brazilian Portuguese', |
|||
'ttsVoiceName': 'sonia', |
|||
}, |
|||
{ |
|||
'shortCode': 'it', |
|||
'chineseName': '意大利语', |
|||
'englishName': 'Italian', |
|||
'ttsVoiceName': 'x2_ItIt_Anna', |
|||
}, |
|||
{ |
|||
'shortCode': 'th', |
|||
'chineseName': '泰语', |
|||
'englishName': 'Thai', |
|||
'ttsVoiceName': 'yingying', |
|||
}, |
|||
{ |
|||
'shortCode': 'ur', |
|||
'chineseName': '乌尔都语', |
|||
'englishName': 'Urdu', |
|||
'ttsVoiceName': 'x2_UrPk_Noreen', |
|||
}, |
|||
{ |
|||
'shortCode': 'pl', |
|||
'chineseName': '波兰语', |
|||
'englishName': 'Polish', |
|||
'ttsVoiceName': 'x2_PlPl_Malgorzata', |
|||
}, |
|||
{ |
|||
'shortCode': 'ms', |
|||
'chineseName': '马来语', |
|||
'englishName': 'Malay', |
|||
'ttsVoiceName': 'x2_MsMy_Hashim', |
|||
}, |
|||
{ |
|||
'shortCode': 'ar', |
|||
'chineseName': '阿拉伯语', |
|||
'englishName': 'Arabic', |
|||
'ttsVoiceName': 'x2_ArEn_Rania', |
|||
}, |
|||
{ |
|||
'shortCode': 'kk', |
|||
'chineseName': '哈萨克语', |
|||
'englishName': 'Kazakh', |
|||
'ttsVoiceName': 'x2_kzkz_Irina', |
|||
}, |
|||
{ |
|||
'shortCode': 'sv', |
|||
'chineseName': '瑞典语', |
|||
'englishName': 'Swedish', |
|||
'ttsVoiceName': 'x2_SvSe_Michaela', |
|||
}, |
|||
{ |
|||
'shortCode': 'fa', |
|||
'chineseName': '波斯语', |
|||
'englishName': 'Persian', |
|||
'ttsVoiceName': 'x2_FaIr_Saheli', |
|||
}, |
|||
{ |
|||
'shortCode': 'bg', |
|||
'chineseName': '保加利亚语', |
|||
'englishName': 'Bulgarian', |
|||
'ttsVoiceName': 'x2_BgBg_Zlati', |
|||
}, |
|||
{ |
|||
'shortCode': 'bo', |
|||
'chineseName': '藏语', |
|||
'englishName': 'Tibetan', |
|||
'ttsVoiceName': 'x2_BoCn_YangJin', |
|||
}, |
|||
{ |
|||
'shortCode': 'uz', |
|||
'chineseName': '乌兹别克语', |
|||
'englishName': 'Uzbek', |
|||
'ttsVoiceName': 'x2_UzUz_Nigina', |
|||
}, |
|||
{ |
|||
'shortCode': 'nl', |
|||
'chineseName': '荷兰语', |
|||
'englishName': 'Dutch', |
|||
'ttsVoiceName': 'x2_NlNl_Robin', |
|||
}, |
|||
{ |
|||
'shortCode': 'uk', |
|||
'chineseName': '乌克兰语', |
|||
'englishName': 'Ukrainian', |
|||
'ttsVoiceName': 'x2_UkUa_Halyna', |
|||
}, |
|||
{ |
|||
'shortCode': 'ta', |
|||
'chineseName': '泰米尔语', |
|||
'englishName': 'Tamil', |
|||
'ttsVoiceName': 'x2_TaIn_Udaya', |
|||
}, |
|||
{ |
|||
'shortCode': 'sw', |
|||
'chineseName': '斯瓦希里语', |
|||
'englishName': 'Swahili', |
|||
'ttsVoiceName': 'x2_SwTz_Hamdan', |
|||
}, |
|||
{ |
|||
'shortCode': 'ro', |
|||
'chineseName': '罗马尼亚语', |
|||
'englishName': 'Romanian', |
|||
'ttsVoiceName': 'x2_RoRo_Miruna', |
|||
}, |
|||
{ |
|||
'shortCode': 'ha', |
|||
'chineseName': '豪萨语', |
|||
'englishName': 'Hausa', |
|||
'ttsVoiceName': 'x2_HaNg_Zainab', |
|||
}, |
|||
{ |
|||
'shortCode': 'cs', |
|||
'chineseName': '捷克语', |
|||
'englishName': 'Czech', |
|||
'ttsVoiceName': 'x2_CsCz_Petra', |
|||
}, |
|||
{ |
|||
'shortCode': 'el', |
|||
'chineseName': '希腊语', |
|||
'englishName': 'Greek', |
|||
'ttsVoiceName': 'x2_ElGr_Dimitra', |
|||
}, |
|||
{ |
|||
'shortCode': 'bn', |
|||
'chineseName': '孟加拉语', |
|||
'englishName': 'Bengali', |
|||
'ttsVoiceName': 'x2_BnBd_Elmy', |
|||
}, |
|||
{ |
|||
'shortCode': 'id', |
|||
'chineseName': '印尼语', |
|||
'englishName': 'Indonesian', |
|||
'ttsVoiceName': 'x2_IdId_Fira', |
|||
}, |
|||
{ |
|||
'shortCode': 'fil', |
|||
'chineseName': '菲律宾语', |
|||
'englishName': 'Filipino', |
|||
'ttsVoiceName': 'x2_TlPh_Grace', |
|||
}, |
|||
{ |
|||
'shortCode': 'tr', |
|||
'chineseName': '土耳其语', |
|||
'englishName': 'Turkish', |
|||
'ttsVoiceName': 'x2_trtr_cenk', |
|||
}, |
|||
{ |
|||
'shortCode': 'ug', |
|||
'chineseName': '维吾尔语', |
|||
'englishName': 'Uyghur', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
{ |
|||
'shortCode': 'mn', |
|||
'chineseName': '蒙古语', |
|||
'englishName': 'Mongolian', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
|
|||
// 方言（部分方言在SDK中可能不支持，仅WebAPI支持） |
|||
{ |
|||
'shortCode': 'yue', |
|||
'chineseName': '粤语', |
|||
'englishName': 'Cantonese', |
|||
'ttsVoiceName': 'x3_xiaoyue', |
|||
}, |
|||
{ |
|||
'shortCode': 'sichuan', |
|||
'chineseName': '四川话', |
|||
'englishName': 'Sichuan Dialect', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
{ |
|||
'shortCode': 'hunan', |
|||
'chineseName': '湖南话', |
|||
'englishName': 'Hunan Dialect', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
{ |
|||
'shortCode': 'henan', |
|||
'chineseName': '河南话', |
|||
'englishName': 'Henan Dialect', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
{ |
|||
'shortCode': 'dongbei', |
|||
'chineseName': '东北话', |
|||
'englishName': 'Northeastern Dialect', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
{ |
|||
'shortCode': 'shaanxi', |
|||
'chineseName': '陕西话', |
|||
'englishName': 'Shaanxi Dialect', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
{ |
|||
'shortCode': 'tw', |
|||
'chineseName': '台湾普通话', |
|||
'englishName': 'Taiwan Mandarin', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
{ |
|||
'shortCode': 'shandong', |
|||
'chineseName': '山东话', |
|||
'englishName': 'Shandong Dialect', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
{ |
|||
'shortCode': 'hubei', |
|||
'chineseName': '湖北话', |
|||
'englishName': 'Hubei Dialect', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
{ |
|||
'shortCode': 'hefei', |
|||
'chineseName': '合肥话', |
|||
'englishName': 'Hefei Dialect', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
{ |
|||
'shortCode': 'neimenggu', |
|||
'chineseName': '内蒙古方言', |
|||
'englishName': 'Inner Mongolia Dialect', |
|||
'ttsVoiceName': '', |
|||
}, |
|||
]; |
|||
} |
|||
@ -0,0 +1,61 @@ |
|||
/// 返回火山引擎的翻译语言配置表（Translation） |
|||
/// |
|||
/// 数据来源：火山引擎机器翻译支持语种 |
|||
/// 参考： https://www.volcengine.com/docs/4640/35107?lang=zh |
|||
List<Map<String, String>> getVolcanoTranslationLanguageSpecs() { |
|||
return [ |
|||
// 中文系列 |
|||
{ |
|||
'shortCode': 'zh', |
|||
'chineseName': '中文', |
|||
'englishName': 'Chinese', |
|||
'translationCode': 'zh', |
|||
}, |
|||
|
|||
// 常见通用语种 |
|||
{ |
|||
'shortCode': 'en', |
|||
'chineseName': '英语', |
|||
'englishName': 'English', |
|||
'translationCode': 'en', |
|||
}, |
|||
|
|||
{ |
|||
'shortCode': 'ja', |
|||
'chineseName': '日语', |
|||
'englishName': 'Japanese', |
|||
'translationCode': 'ja', |
|||
}, |
|||
{ |
|||
'shortCode': 'id', |
|||
'chineseName': '印尼语', |
|||
'englishName': 'Indonesian', |
|||
'translationCode': 'id', |
|||
}, |
|||
{ |
|||
'shortCode': 'es', |
|||
'chineseName': '西班牙语', |
|||
'englishName': 'Spanish', |
|||
'translationCode': 'es', |
|||
}, |
|||
|
|||
{ |
|||
'shortCode': 'pt', |
|||
'chineseName': '葡萄牙语', |
|||
'englishName': 'Portuguese', |
|||
'translationCode': 'pt', |
|||
}, |
|||
{ |
|||
'shortCode': 'fr', |
|||
'chineseName': '法语', |
|||
'englishName': 'French', |
|||
'translationCode': 'fr', |
|||
}, |
|||
{ |
|||
'shortCode': 'de', |
|||
'chineseName': '德语', |
|||
'englishName': 'German', |
|||
'translationCode': 'de', |
|||
}, |
|||
]; |
|||
} |
|||
@ -0,0 +1,14 @@ |
|||
gradle-wrapper.jar |
|||
/.gradle |
|||
/captures/ |
|||
/gradlew |
|||
/gradlew.bat |
|||
/local.properties |
|||
GeneratedPluginRegistrant.java |
|||
.cxx/ |
|||
|
|||
# Remember to never publicly share your keystore. |
|||
# See https://flutter.dev/to/reference-keystore |
|||
key.properties |
|||
**/*.keystore |
|||
**/*.jks |
|||
@ -0,0 +1,44 @@ |
|||
plugins { |
|||
id("com.android.application") |
|||
id("kotlin-android") |
|||
// The Flutter Gradle Plugin must be applied after the Android and Kotlin Gradle plugins. |
|||
id("dev.flutter.flutter-gradle-plugin") |
|||
} |
|||
|
|||
android { |
|||
namespace = "com.example.azure_speech" |
|||
compileSdk = flutter.compileSdkVersion |
|||
ndkVersion = flutter.ndkVersion |
|||
|
|||
compileOptions { |
|||
sourceCompatibility = JavaVersion.VERSION_11 |
|||
targetCompatibility = JavaVersion.VERSION_11 |
|||
} |
|||
|
|||
kotlinOptions { |
|||
jvmTarget = JavaVersion.VERSION_11.toString() |
|||
} |
|||
|
|||
defaultConfig { |
|||
// TODO: Specify your own unique Application ID (https://developer.android.com/studio/build/application-id.html). |
|||
applicationId = "com.example.azure_speech" |
|||
// You can update the following values to match your application needs. |
|||
// For more information, see: https://flutter.dev/to/review-gradle-config. |
|||
minSdk = flutter.minSdkVersion |
|||
targetSdk = flutter.targetSdkVersion |
|||
versionCode = flutter.versionCode |
|||
versionName = flutter.versionName |
|||
} |
|||
|
|||
buildTypes { |
|||
release { |
|||
// TODO: Add your own signing config for the release build. |
|||
// Signing with the debug keys for now, so `flutter run --release` works. |
|||
signingConfig = signingConfigs.getByName("debug") |
|||
} |
|||
} |
|||
} |
|||
|
|||
flutter { |
|||
source = "../.." |
|||
} |
|||
@ -0,0 +1,7 @@ |
|||
<manifest xmlns:android="http://schemas.android.com/apk/res/android"> |
|||
<!-- The INTERNET permission is required for development. Specifically, |
|||
the Flutter tool needs it to communicate with the running application |
|||
to allow setting breakpoints, to provide hot reload, etc. |
|||
--> |
|||
<uses-permission android:name="android.permission.INTERNET"/> |
|||
</manifest> |
|||
@ -0,0 +1,45 @@ |
|||
<manifest xmlns:android="http://schemas.android.com/apk/res/android"> |
|||
<application |
|||
android:label="azure_speech" |
|||
android:name="${applicationName}" |
|||
android:icon="@mipmap/ic_launcher"> |
|||
<activity |
|||
android:name=".MainActivity" |
|||
android:exported="true" |
|||
android:launchMode="singleTop" |
|||
android:taskAffinity="" |
|||
android:theme="@style/LaunchTheme" |
|||
android:configChanges="orientation|keyboardHidden|keyboard|screenSize|smallestScreenSize|locale|layoutDirection|fontScale|screenLayout|density|uiMode" |
|||
android:hardwareAccelerated="true" |
|||
android:windowSoftInputMode="adjustResize"> |
|||
<!-- Specifies an Android theme to apply to this Activity as soon as |
|||
the Android process has started. This theme is visible to the user |
|||
while the Flutter UI initializes. After that, this theme continues |
|||
to determine the Window background behind the Flutter UI. --> |
|||
<meta-data |
|||
android:name="io.flutter.embedding.android.NormalTheme" |
|||
android:resource="@style/NormalTheme" |
|||
/> |
|||
<intent-filter> |
|||
<action android:name="android.intent.action.MAIN"/> |
|||
<category android:name="android.intent.category.LAUNCHER"/> |
|||
</intent-filter> |
|||
</activity> |
|||
<!-- Don't delete the meta-data below. |
|||
This is used by the Flutter tool to generate GeneratedPluginRegistrant.java --> |
|||
<meta-data |
|||
android:name="flutterEmbedding" |
|||
android:value="2" /> |
|||
</application> |
|||
<!-- Required to query activities that can process text, see: |
|||
https://developer.android.com/training/package-visibility and |
|||
https://developer.android.com/reference/android/content/Intent#ACTION_PROCESS_TEXT. |
|||
|
|||
In particular, this is used by the Flutter engine in io.flutter.plugin.text.ProcessTextPlugin. --> |
|||
<queries> |
|||
<intent> |
|||
<action android:name="android.intent.action.PROCESS_TEXT"/> |
|||
<data android:mimeType="text/plain"/> |
|||
</intent> |
|||
</queries> |
|||
</manifest> |
|||
@ -0,0 +1,5 @@ |
|||
package com.example.azure_speech |
|||
|
|||
import io.flutter.embedding.android.FlutterActivity |
|||
|
|||
class MainActivity : FlutterActivity() |
|||
@ -0,0 +1,12 @@ |
|||
<?xml version="1.0" encoding="utf-8"?> |
|||
<!-- Modify this file to customize your launch splash screen --> |
|||
<layer-list xmlns:android="http://schemas.android.com/apk/res/android"> |
|||
<item android:drawable="?android:colorBackground" /> |
|||
|
|||
<!-- You can insert your own image assets here --> |
|||
<!-- <item> |
|||
<bitmap |
|||
android:gravity="center" |
|||
android:src="@mipmap/launch_image" /> |
|||
</item> --> |
|||
</layer-list> |
|||
@ -0,0 +1,12 @@ |
|||
<?xml version="1.0" encoding="utf-8"?> |
|||
<!-- Modify this file to customize your launch splash screen --> |
|||
<layer-list xmlns:android="http://schemas.android.com/apk/res/android"> |
|||
<item android:drawable="@android:color/white" /> |
|||
|
|||
<!-- You can insert your own image assets here --> |
|||
<!-- <item> |
|||
<bitmap |
|||
android:gravity="center" |
|||
android:src="@mipmap/launch_image" /> |
|||
</item> --> |
|||
</layer-list> |
|||
|
After Width: | Height: | Size: 544 B |
|
After Width: | Height: | Size: 442 B |
|
After Width: | Height: | Size: 721 B |
|
After Width: | Height: | Size: 1.0 KiB |
|
After Width: | Height: | Size: 1.4 KiB |
@ -0,0 +1,18 @@ |
|||
<?xml version="1.0" encoding="utf-8"?> |
|||
<resources> |
|||
<!-- Theme applied to the Android Window while the process is starting when the OS's Dark Mode setting is on --> |
|||
<style name="LaunchTheme" parent="@android:style/Theme.Black.NoTitleBar"> |
|||
<!-- Show a splash screen on the activity. Automatically removed when |
|||
the Flutter engine draws its first frame --> |
|||
<item name="android:windowBackground">@drawable/launch_background</item> |
|||
</style> |
|||
<!-- Theme applied to the Android Window as soon as the process has started. |
|||
This theme determines the color of the Android Window while your |
|||
Flutter UI initializes, as well as behind your Flutter UI while its |
|||
running. |
|||
|
|||
This Theme is only used starting with V2 of Flutter's Android embedding. --> |
|||
<style name="NormalTheme" parent="@android:style/Theme.Black.NoTitleBar"> |
|||
<item name="android:windowBackground">?android:colorBackground</item> |
|||
</style> |
|||
</resources> |
|||
@ -0,0 +1,18 @@ |
|||
<?xml version="1.0" encoding="utf-8"?> |
|||
<resources> |
|||
<!-- Theme applied to the Android Window while the process is starting when the OS's Dark Mode setting is off --> |
|||
<style name="LaunchTheme" parent="@android:style/Theme.Light.NoTitleBar"> |
|||
<!-- Show a splash screen on the activity. Automatically removed when |
|||
the Flutter engine draws its first frame --> |
|||
<item name="android:windowBackground">@drawable/launch_background</item> |
|||
</style> |
|||
<!-- Theme applied to the Android Window as soon as the process has started. |
|||
This theme determines the color of the Android Window while your |
|||
Flutter UI initializes, as well as behind your Flutter UI while its |
|||
running. |
|||
|
|||
This Theme is only used starting with V2 of Flutter's Android embedding. --> |
|||
<style name="NormalTheme" parent="@android:style/Theme.Light.NoTitleBar"> |
|||
<item name="android:windowBackground">?android:colorBackground</item> |
|||
</style> |
|||
</resources> |
|||
@ -0,0 +1,7 @@ |
|||
<manifest xmlns:android="http://schemas.android.com/apk/res/android"> |
|||
<!-- The INTERNET permission is required for development. Specifically, |
|||
the Flutter tool needs it to communicate with the running application |
|||
to allow setting breakpoints, to provide hot reload, etc. |
|||
--> |
|||
<uses-permission android:name="android.permission.INTERNET"/> |
|||
</manifest> |
|||
@ -0,0 +1,3 @@ |
|||
org.gradle.jvmargs=-Xmx8G -XX:MaxMetaspaceSize=4G -XX:ReservedCodeCacheSize=512m -XX:+HeapDumpOnOutOfMemoryError |
|||
android.useAndroidX=true |
|||
android.enableJetifier=true |
|||
@ -0,0 +1,5 @@ |
|||
distributionBase=GRADLE_USER_HOME |
|||
distributionPath=wrapper/dists |
|||
zipStoreBase=GRADLE_USER_HOME |
|||
zipStorePath=wrapper/dists |
|||
distributionUrl=https\://services.gradle.org/distributions/gradle-8.12-all.zip |
|||
@ -0,0 +1,478 @@ |
|||
package com.yunqiinnovation.azure_speech |
|||
|
|||
import android.content.Context |
|||
import android.util.Log |
|||
import okhttp3.* |
|||
import okio.ByteString |
|||
import okio.ByteString.Companion.toByteString |
|||
import org.json.JSONObject |
|||
import java.util.UUID |
|||
import java.util.concurrent.TimeUnit |
|||
import java.util.concurrent.atomic.AtomicBoolean |
|||
import kotlinx.coroutines.* |
|||
|
|||
/** |
|||
* 阿里云百炼（Bailian）实时音视频翻译助手 (qwen3-livetranslate-flash-realtime)。 |
|||
* 参考阿里云 "实时音视频翻译-通义千问" WebSocket API 实现。 |
|||
* 采用 Session-based 协议 (类似 OpenAI Realtime API)。 |
|||
*/ |
|||
class AliyunBailianE2EHelper( |
|||
private val context: Context |
|||
) { |
|||
private val TAG = "AliyunBailianHelper" |
|||
|
|||
interface Callback { |
|||
/** 会话开始 */ |
|||
fun onSessionStarted(sessionId: String) |
|||
|
|||
/** 增量文本到达（大模型回复的文本/翻译结果） */ |
|||
fun onPartialText(sessionId: String, text: String) |
|||
|
|||
/** 增量源文本到达（用户语音的实时识别结果） */ |
|||
fun onPartialSourceText(sessionId: String, text: String) |
|||
|
|||
/** 源文本最终结果到达（用户一句话识别完成） */ |
|||
fun onFinalSourceText(sessionId: String, finalText: String) |
|||
|
|||
/** 翻译/回复最终结果到达（大模型回复完成） */ |
|||
fun onFinalTranslatedText(sessionId: String, finalText: String) |
|||
|
|||
/** 增量音频片段到达（服务端返回的 TTS 音频） */ |
|||
fun onPartialAudio(sessionId: String, data: ByteArray) |
|||
|
|||
/** 会话结束，返回完整文本与完整合成音频 */ |
|||
fun onSessionFinished(sessionId: String, finalText: String, finalAudio: ByteArray) |
|||
|
|||
/** 会话失败或取消 */ |
|||
fun onSessionError(sessionId: String, code: Int, message: String) |
|||
} |
|||
|
|||
data class Config( |
|||
// 阿里云百炼 WebSocket URL |
|||
// 实时音视频翻译通常使用: wss://dashscope.aliyuncs.com/api-ws/v1/realtime |
|||
val wsUrl: String = "wss://dashscope.aliyuncs.com/api-ws/v1/realtime", |
|||
|
|||
// 阿里云 API Key |
|||
val apiKey: String = "", |
|||
|
|||
// 模型名称，例如 "qwen3-livetranslate-flash-realtime" |
|||
val appId: String = "qwen3-livetranslate-flash-realtime", |
|||
|
|||
// 采样率，默认 16000 |
|||
val sampleRate: Int = 16000, |
|||
|
|||
// 源语言 (e.g. "zh", "en", "ja") |
|||
val sourceLanguage: String = "zh", |
|||
|
|||
// 目标语言 (e.g. "en", "zh", "ja") |
|||
val targetLanguage: String = "en", |
|||
|
|||
// 音色 (e.g. "Cherry", "Kiki") |
|||
val voice: String = "" |
|||
) |
|||
|
|||
private var conf: Config = Config() |
|||
private var callback: Callback? = null |
|||
|
|||
private var client: OkHttpClient? = null |
|||
private var webSocket: WebSocket? = null |
|||
private var sessionId: String = "" |
|||
|
|||
// 用于累积文本，以便在结束时返回完整内容 |
|||
private val fullTextBuffer = StringBuilder() |
|||
// 用于累积音频 |
|||
private val fullAudioBuffer = java.io.ByteArrayOutputStream() |
|||
|
|||
// 当前句子的缓冲 |
|||
private val recvTextBuffer = StringBuilder() |
|||
private val recvText = mutableListOf<String>() |
|||
|
|||
// 音频分块缓冲 |
|||
private var audioChunkBuffer = ByteArray(0) |
|||
|
|||
private val isStarted = AtomicBoolean(false) |
|||
private val scope = CoroutineScope(Dispatchers.IO + SupervisorJob()) |
|||
|
|||
/** |
|||
* 初始化助手，设置配置与回调。 |
|||
* |
|||
* @param config 配置对象 |
|||
* @param cb 回调接口 |
|||
* @return 是否初始化成功 |
|||
*/ |
|||
fun initialize(config: Config, cb: Callback): Boolean { |
|||
|
|||
// 注意：通义千问 LiveTranslate 通常使用 2 位语言代码，但部分方言可能不同。 |
|||
// 这里做一个简单的处理，具体需参考文档支持的语种列表。 |
|||
|
|||
|
|||
conf = config.copy(sourceLanguage = config.sourceLanguage, targetLanguage = config.targetLanguage) |
|||
callback = cb |
|||
Log.d(TAG, "initialize: wsUrl=${conf.wsUrl}, model=${conf.appId}, src=${conf.sourceLanguage}, tgt=${conf.targetLanguage}") |
|||
|
|||
client = OkHttpClient.Builder() |
|||
.pingInterval(30, TimeUnit.SECONDS) |
|||
.readTimeout(0, TimeUnit.SECONDS) |
|||
.build() |
|||
|
|||
startContinuousConversation() |
|||
return true |
|||
} |
|||
|
|||
|
|||
/** |
|||
* 启动会话，建立 WebSocket 连接。 |
|||
*/ |
|||
fun startContinuousConversation(): Boolean { |
|||
if (client == null) return false |
|||
if (isStarted.get()) return true |
|||
|
|||
sessionId = UUID.randomUUID().toString() |
|||
Log.d(TAG, "startContinuousConversation: sessionId=${sessionId}") |
|||
|
|||
recvTextBuffer.setLength(0) |
|||
fullTextBuffer.setLength(0) |
|||
fullAudioBuffer.reset() |
|||
|
|||
// 构造 URL，必须包含 model 参数 |
|||
// wss://dashscope.aliyuncs.com/api-ws/v1/realtime?model=qwen3-livetranslate-flash-realtime |
|||
val urlWithModel = if (conf.wsUrl.contains("?")) { |
|||
// 如果已经包含了参数，假设用户自己拼接了 model 或者其他参数 |
|||
conf.wsUrl |
|||
} else { |
|||
// 强制拼接 model=qwen3-livetranslate-flash-realtime |
|||
"${conf.wsUrl}?model=qwen3-livetranslate-flash-realtime" |
|||
} |
|||
|
|||
val request = Request.Builder() |
|||
.url(urlWithModel) |
|||
.header("Authorization", "Bearer ${conf.apiKey}") |
|||
.build() |
|||
|
|||
val listener = object : WebSocketListener() { |
|||
override fun onOpen(ws: WebSocket, response: Response) { |
|||
Log.d(TAG, "onOpen: code=${response.code}") |
|||
webSocket = ws |
|||
isStarted.set(true) |
|||
|
|||
// 发送 session.update 配置会话 |
|||
sendSessionUpdate(ws) |
|||
|
|||
callback?.onSessionStarted(sessionId) |
|||
} |
|||
|
|||
override fun onMessage(ws: WebSocket, text: String) { |
|||
handleJsonMessage(text) |
|||
} |
|||
|
|||
override fun onMessage(ws: WebSocket, bytes: ByteString) { |
|||
// 通常 LiveTranslate 协议使用 JSON 传输所有内容（包括 Base64 音频） |
|||
// 但如果协议支持二进制帧，可以在此处理 |
|||
Log.w(TAG, "Received unexpected binary message: size=${bytes.size}") |
|||
} |
|||
|
|||
override fun onFailure(ws: WebSocket, t: Throwable, response: Response?) { |
|||
val msg = t.message ?: "unknown" |
|||
val code = response?.code ?: -1 |
|||
Log.e(TAG, "onFailure: ${msg} code=${code}") |
|||
|
|||
// // 检查是否需要自动重连 (例如网络异常) |
|||
// // 注意：鉴权失败(401)不应重试，避免死循环或账号封禁 |
|||
// if (isStarted.get() && code != 401 && code != 403) { |
|||
// Log.i(TAG, "onFailure: detected failure (code=$code), triggering auto-restart") |
|||
// restartSession() |
|||
// return |
|||
// } |
|||
|
|||
callback?.onSessionError(sessionId, 1011, msg) |
|||
isStarted.set(false) |
|||
} |
|||
|
|||
override fun onClosed(ws: WebSocket, code: Int, reason: String) { |
|||
Log.d(TAG, "onClosed: code=${code} reason=${reason}") |
|||
isStarted.set(false) |
|||
callback?.onSessionFinished(sessionId, fullTextBuffer.toString(), fullAudioBuffer.toByteArray()) |
|||
} |
|||
} |
|||
|
|||
Log.d(TAG, "connecting to $urlWithModel") |
|||
client!!.newWebSocket(request, listener) |
|||
return true |
|||
} |
|||
|
|||
/** |
|||
* 重启会话（用于异常自动重连） |
|||
*/ |
|||
private fun restartSession() { |
|||
Log.i(TAG, "restartSession: performing auto-restart...") |
|||
try { webSocket?.close(1000, "restarting") } catch (_: Exception) {} |
|||
webSocket = null |
|||
isStarted.set(false) |
|||
|
|||
try { |
|||
java.util.Timer().schedule(object : java.util.TimerTask() { |
|||
override fun run() { |
|||
Log.i(TAG, "restartSession: timer task running") |
|||
if (!isStarted.get()) { |
|||
Log.i(TAG, "restartSession: calling startContinuousConversation") |
|||
val result = startContinuousConversation() |
|||
Log.i(TAG, "restartSession: startContinuousConversation result=$result") |
|||
} else { |
|||
Log.i(TAG, "restartSession: already started, skipping") |
|||
} |
|||
} |
|||
}, 200) |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "restartSession: timer failed", e) |
|||
} |
|||
} |
|||
|
|||
private fun sendSessionUpdate(ws: WebSocket) { |
|||
try { |
|||
val json = JSONObject() |
|||
json.put("type", "session.update") |
|||
|
|||
val session = JSONObject() |
|||
|
|||
// 配置输入音频转写（ASR） |
|||
val transcription = JSONObject() |
|||
transcription.put("model", "qwen3-asr-flash-realtime") // 配合 LiveTranslate 使用的推荐 ASR 模型 |
|||
transcription.put("language", conf.sourceLanguage) // 源语言 |
|||
session.put("input_audio_transcription", transcription) |
|||
|
|||
// 配置翻译 |
|||
val translation = JSONObject() |
|||
translation.put("language", conf.targetLanguage) // 目标语言 |
|||
session.put("translation", translation) |
|||
|
|||
if (conf.voice.isNotEmpty()) { |
|||
session.put("voice", conf.voice) |
|||
} else { |
|||
if(conf.targetLanguage != "yue"){ |
|||
session.put("voice", "Cherry")//音色 |
|||
}else{ |
|||
session.put("voice", "Kiki")//音色 |
|||
} |
|||
} |
|||
|
|||
session.put("input_audio_format", "pcm16")//输入音频格式 |
|||
session.put("output_audio_format", "pcm24")//输出音频格式 |
|||
// 配置模态 (文本 + 音频) |
|||
val modalities = org.json.JSONArray() |
|||
modalities.put("text") |
|||
modalities.put("audio") |
|||
session.put("modalities", modalities) |
|||
|
|||
json.put("session", session) |
|||
|
|||
ws.send(json.toString()) |
|||
Log.d(TAG, "Sent session.update: $json") |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "Error sending session.update", e) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 将 24kHz PCM 音频重采样为 16kHz |
|||
*/ |
|||
private fun resample24kTo16k(input: ByteArray): ByteArray { |
|||
// 16 bit PCM |
|||
val inputShorts = ShortArray(input.size / 2) |
|||
java.nio.ByteBuffer.wrap(input).order(java.nio.ByteOrder.LITTLE_ENDIAN).asShortBuffer().get(inputShorts) |
|||
|
|||
// 24k -> 16k (ratio 1.5) |
|||
// Output size = Input size / 1.5 = Input size * 2 / 3 |
|||
val outputSize = (inputShorts.size * 2) / 3 |
|||
val outputShorts = ShortArray(outputSize) |
|||
|
|||
for (i in 0 until outputSize) { |
|||
val inputIndex = (i * 1.5).toFloat() |
|||
val index = inputIndex.toInt() |
|||
val frac = inputIndex - index |
|||
|
|||
if (index + 1 < inputShorts.size) { |
|||
val val1 = inputShorts[index] |
|||
val val2 = inputShorts[index + 1] |
|||
outputShorts[i] = (val1 * (1 - frac) + val2 * frac).toInt().toShort() |
|||
} else if (index < inputShorts.size) { |
|||
outputShorts[i] = inputShorts[index] |
|||
} |
|||
} |
|||
|
|||
val outputBytes = ByteArray(outputShorts.size * 2) |
|||
java.nio.ByteBuffer.wrap(outputBytes).order(java.nio.ByteOrder.LITTLE_ENDIAN).asShortBuffer().put(outputShorts) |
|||
|
|||
return outputBytes |
|||
} |
|||
|
|||
private fun handleJsonMessage(text: String) { |
|||
try { |
|||
val json = JSONObject(text) |
|||
val type = json.optString("type") |
|||
|
|||
// Log.d(TAG, "Received event: $type") |
|||
|
|||
when (type) { |
|||
"error" -> { |
|||
val error = json.optJSONObject("error") |
|||
val msg = error?.optString("message") ?: "Unknown error" |
|||
val code = error?.optString("code") ?: "" |
|||
Log.e(TAG, "Error from server: $code - $msg") |
|||
callback?.onSessionError(sessionId, 1012, msg) |
|||
} |
|||
"session.created" -> { |
|||
Log.d(TAG, "Session created") |
|||
} |
|||
"session.updated" -> { |
|||
Log.d(TAG, "Session updated") |
|||
} |
|||
// 源语言识别结果 (Partial) |
|||
"conversation.item.input_audio_transcription.text" -> { |
|||
val txt = json.optString("text", "") |
|||
if (txt.isNotEmpty()) { |
|||
Log.d(TAG, "onPartialSourceText: $txt") |
|||
callback?.onPartialSourceText(sessionId, txt) |
|||
} |
|||
} |
|||
// 源语言识别结果 (Final) |
|||
"conversation.item.input_audio_transcription.completed" -> { |
|||
val item = json.optJSONObject("item") |
|||
val content = item?.optJSONArray("content") |
|||
// 提取 text |
|||
var finalTxt = "" |
|||
if (content != null && content.length() > 0) { |
|||
for (i in 0 until content.length()) { |
|||
val part = content.optJSONObject(i) |
|||
if (part?.optString("type") == "input_audio") { |
|||
finalTxt = part.optString("transcript", "") |
|||
break |
|||
} |
|||
} |
|||
} |
|||
if (finalTxt.isEmpty()) { |
|||
// 备用：有些事件格式可能直接在 payload 里 |
|||
finalTxt = json.optString("transcript", "") |
|||
} |
|||
|
|||
if (finalTxt.isNotEmpty()) { |
|||
Log.d(TAG, "sessionId: ${sessionId}, onFinalSourceText: $finalTxt") |
|||
callback?.onFinalSourceText(sessionId, finalTxt) |
|||
} |
|||
} |
|||
|
|||
// 翻译/生成结果 (Partial Text) |
|||
"response.audio_transcript.delta" -> { |
|||
val delta = json.optString("delta", "") |
|||
if (delta.isNotEmpty()) { |
|||
recvTextBuffer.append(delta) |
|||
Log.d(TAG, "onPartialText: $delta") |
|||
callback?.onPartialText(sessionId, delta) |
|||
} |
|||
} |
|||
|
|||
// 翻译/生成结果 (Final Text) |
|||
"response.audio_transcript.done" -> { |
|||
val transcript = json.optString("transcript", "") |
|||
val finalText = if (transcript.isNotEmpty()) transcript else recvTextBuffer.toString() |
|||
Log.d(TAG, "sessionId: ${sessionId}, onFinalTranslatedText: $finalText") |
|||
callback?.onFinalTranslatedText(sessionId, finalText) |
|||
|
|||
if (fullTextBuffer.isNotEmpty()) fullTextBuffer.append(" ") |
|||
fullTextBuffer.append(finalText) |
|||
recvTextBuffer.setLength(0) |
|||
} |
|||
|
|||
// 翻译/生成音频 (Delta) |
|||
"response.audio.delta" -> { |
|||
val base64Audio = json.optString("delta", "") |
|||
if (base64Audio.isNotEmpty()) { |
|||
// Log.d(TAG, "event_id: ${UUID.randomUUID().toString()}, onPartialAudio: $base64Audio") |
|||
try { |
|||
val audioBytes = android.util.Base64.decode(base64Audio, android.util.Base64.DEFAULT) |
|||
if (audioBytes != null && audioBytes.isNotEmpty()) { |
|||
// 默认输出 24k，需重采样到 16k |
|||
val resampled = resample24kTo16k(audioBytes) |
|||
|
|||
try { |
|||
fullAudioBuffer.write(resampled) |
|||
} catch (_: Exception) {} |
|||
|
|||
Log.d(TAG, "onPartialAudio: rawSize=${audioBytes.size} resampledSize=${resampled.size}") |
|||
// 使用 processAudioChunk 进行分块回调 |
|||
processAudioChunk(resampled) |
|||
|
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "Base64 decode error", e) |
|||
} |
|||
} |
|||
} |
|||
|
|||
"response.done" -> { |
|||
// 一次响应结束 |
|||
Log.d(TAG, "Response done") |
|||
} |
|||
|
|||
else -> { |
|||
// Log.d(TAG, "Unhandled event: $type") |
|||
} |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "handleJsonMessage error: ${e.message}") |
|||
} |
|||
} |
|||
|
|||
private fun processAudioChunk( |
|||
incoming: ByteArray, |
|||
multiple: Int = 1280 |
|||
) { |
|||
if (incoming.isNotEmpty()) { |
|||
audioChunkBuffer += incoming |
|||
} |
|||
val sendLen = (audioChunkBuffer.size / multiple) * multiple |
|||
if (sendLen > 0) { |
|||
val chunk = audioChunkBuffer.copyOfRange(0, sendLen) |
|||
callback?.onPartialAudio(sessionId, chunk) |
|||
audioChunkBuffer = audioChunkBuffer.copyOfRange(sendLen, audioChunkBuffer.size) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 推送音频数据。 |
|||
* LiveTranslate 协议要求发送 input_audio_buffer.append 事件， |
|||
* 音频数据需 Base64 编码。 |
|||
*/ |
|||
fun pushAudioData(data: ByteArray): Boolean { |
|||
val ws = webSocket ?: return false |
|||
if (!isStarted.get()) return false |
|||
|
|||
try { |
|||
// 将 PCM 数据转为 Base64 |
|||
val base64Audio = data.toByteString().base64() |
|||
|
|||
val json = JSONObject() |
|||
json.put("type", "input_audio_buffer.append") |
|||
json.put("audio", base64Audio) |
|||
ws.send(json.toString()) |
|||
return true |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "pushAudioData error", e) |
|||
return false |
|||
} |
|||
} |
|||
|
|||
fun stopContinuousConversation(): Boolean { |
|||
// LiveTranslate 没有明确的 "stop task" 指令，通常直接 close 连接即可 |
|||
// 或者发送 commit 强制模型生成（如果处于等待状态） |
|||
// 这里简单处理为关闭连接 |
|||
webSocket?.close(1000, "User stopped") |
|||
return true |
|||
} |
|||
|
|||
fun dispose() { |
|||
try { webSocket?.close(1000, "dispose") } catch (_: Exception) {} |
|||
webSocket = null |
|||
isStarted.set(false) |
|||
scope.cancel() |
|||
} |
|||
} |
|||
@ -0,0 +1,758 @@ |
|||
package com.yunqiinnovation.azure_speech |
|||
|
|||
import com.yunqiinnovation.azure_speech.utils.FileLogger |
|||
import com.example.astclient.DoubaoE2ETranslateHelper |
|||
|
|||
/** |
|||
* Interface for sending events back to Flutter/Native |
|||
*/ |
|||
fun interface AstEventSender { |
|||
fun send(event: Map<String, Any>) |
|||
} |
|||
|
|||
/** |
|||
* Interface for writing audio data (PCM) |
|||
*/ |
|||
fun interface AudioWriter { |
|||
fun write(data: ByteArray) |
|||
} |
|||
|
|||
/** |
|||
* Azure AST Callback Implementation |
|||
*/ |
|||
class AzureAstCallback( |
|||
private val serviceId: String, |
|||
private val direction: String, |
|||
private val eventSender: AstEventSender, |
|||
private val audioWriter: AudioWriter? = null |
|||
) : IntegratedSpeechTranslationService.ServiceEventCallback { |
|||
private val tag = "AzureAstCallback" |
|||
|
|||
override fun onServiceInitialized() { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "serviceInitialized", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onRecognizing( |
|||
utteranceId: String, |
|||
text: String, |
|||
language: String, |
|||
confidence: Float |
|||
) { |
|||
FileLogger.d( |
|||
tag, |
|||
"[$serviceId/$direction] 识别中: $text, 语言: $language, 置信度: $confidence" |
|||
) |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "recognizing", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text, |
|||
"language" to language, |
|||
"confidence" to confidence |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onRecognized( |
|||
utteranceId: String, |
|||
text: String, |
|||
language: String, |
|||
confidence: Float |
|||
) { |
|||
FileLogger.d( |
|||
tag, |
|||
"[$serviceId/$direction] 识别到文本: $text, 语言: $language, 置信度: $confidence" |
|||
) |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "recognized", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text, |
|||
"language" to language, |
|||
"confidence" to confidence |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onTranslated( |
|||
utteranceId: String, |
|||
originalText: String, |
|||
translatedText: String, |
|||
targetLanguage: String |
|||
) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translated", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"originalText" to originalText, |
|||
"translatedText" to translatedText, |
|||
"targetLanguage" to targetLanguage |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onInterimTranslated( |
|||
utteranceId: String, |
|||
originalText: String, |
|||
translatedText: String, |
|||
targetLanguage: String |
|||
) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translatedInterim", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"originalText" to originalText, |
|||
"translatedText" to translatedText, |
|||
"targetLanguage" to targetLanguage |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onTranslationStarted(utteranceId: String, text: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translationStarted", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onTranslationFailed(utteranceId: String, text: String, error: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translationFailed", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text, |
|||
"error" to error |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onSynthesisStarted(utteranceId: String, text: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "synthesisStarted", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onSynthesisCompleted(utteranceId: String, text: String) { |
|||
FileLogger.d(tag, "[$serviceId/$direction] 语音合成完成,文本=${text}") |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "synthesisCompleted", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onSynthesisAudioGenerated( |
|||
utteranceId: String, |
|||
text: String, |
|||
audioData: ByteArray |
|||
) { |
|||
FileLogger.d( |
|||
tag, |
|||
"[$serviceId/$direction] 合成音频生成,文本=${text},音频=${audioData.size}" |
|||
) |
|||
// val pcmData = extractPcmFromWav(audioData) // 原逻辑已注释，直接使用 audioData |
|||
val pcmData = audioData |
|||
if (pcmData.isNotEmpty()) { |
|||
audioWriter?.write(pcmData) |
|||
} else { |
|||
FileLogger.e(tag, "[$serviceId/$direction] 音频数据为空") |
|||
} |
|||
} |
|||
|
|||
override fun onSynthesisFailed(utteranceId: String, text: String, error: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "synthesisFailed", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text, |
|||
"error" to error |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onSynthesisProgress( |
|||
utteranceId: String, |
|||
text: String, |
|||
progress: Float |
|||
) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "synthesisProgress", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text, |
|||
"progress" to progress |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onRecognitionStarted() { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "recognitionStarted", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onRecognitionStopped() { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "recognitionStopped", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onStateChanged(component: String, isActive: Boolean) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "stateChanged", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"component" to component, |
|||
"isActive" to isActive |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onError(component: String, error: String) { |
|||
FileLogger.e(tag, "[$serviceId/$direction] 错误,组件=${component},错误=${error}") |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "error", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"component" to component, |
|||
"error" to error |
|||
) |
|||
) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* Iflytek AST Callback Implementation |
|||
*/ |
|||
class IflytekAstCallback( |
|||
private val serviceId: String, |
|||
private val direction: String, |
|||
private val eventSender: AstEventSender, |
|||
private val audioWriter: AudioWriter? = null |
|||
) : IflytekIntegratedSpeechService.ServiceEventCallback { |
|||
private val tag = "IflytekAstCallback" |
|||
|
|||
override fun onServiceInitialized() { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "serviceInitialized", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onRecognizing( |
|||
utteranceId: String, |
|||
text: String, |
|||
language: String, |
|||
confidence: Float |
|||
) { |
|||
FileLogger.d( |
|||
tag, |
|||
"[$serviceId/$direction] 识别中: $text, 语言: $language, 置信度: $confidence" |
|||
) |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "recognizing", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text, |
|||
"language" to language, |
|||
"confidence" to confidence |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onRecognized( |
|||
utteranceId: String, |
|||
text: String, |
|||
language: String, |
|||
confidence: Float |
|||
) { |
|||
FileLogger.d( |
|||
tag, |
|||
"[$serviceId/$direction] 识别到文本: $text, 语言: $language, 置信度: $confidence" |
|||
) |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "recognized", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text, |
|||
"language" to language, |
|||
"confidence" to confidence |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onTranslated( |
|||
utteranceId: String, |
|||
originalText: String, |
|||
translatedText: String, |
|||
targetLanguage: String |
|||
) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translated", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"originalText" to originalText, |
|||
"translatedText" to translatedText, |
|||
"targetLanguage" to targetLanguage |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onInterimTranslated( |
|||
utteranceId: String, |
|||
originalText: String, |
|||
translatedText: String, |
|||
targetLanguage: String |
|||
) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translatedInterim", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"originalText" to originalText, |
|||
"translatedText" to translatedText, |
|||
"targetLanguage" to targetLanguage |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onTranslationStarted(utteranceId: String, text: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translationStarted", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onTranslationFailed(utteranceId: String, text: String, error: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translationFailed", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text, |
|||
"error" to error |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onSynthesisStarted(utteranceId: String, text: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "synthesisStarted", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onSynthesisCompleted(utteranceId: String, text: String) { |
|||
FileLogger.d(tag, "[$serviceId/$direction] 语音合成完成,文本=${text}") |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "synthesisCompleted", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onSynthesisAudioGenerated( |
|||
utteranceId: String, |
|||
text: String, |
|||
audioData: ByteArray |
|||
) { |
|||
FileLogger.d( |
|||
tag, |
|||
"[$serviceId/$direction] 合成音频生成,文本=${text},音频=${audioData.size}" |
|||
) |
|||
val pcmData = audioData |
|||
if (pcmData.isNotEmpty()) { |
|||
audioWriter?.write(pcmData) |
|||
} else { |
|||
FileLogger.e(tag, "[$serviceId/$direction] 音频数据为空或不是WAV格式") |
|||
} |
|||
} |
|||
|
|||
override fun onSynthesisFailed(utteranceId: String, text: String, error: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "synthesisFailed", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text, |
|||
"error" to error |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onSynthesisProgress( |
|||
utteranceId: String, |
|||
text: String, |
|||
progress: Float |
|||
) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "synthesisProgress", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to utteranceId, |
|||
"text" to text, |
|||
"progress" to progress |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onRecognitionStarted() { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "recognitionStarted", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onRecognitionStopped() { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "recognitionStopped", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onStateChanged(component: String, isActive: Boolean) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "stateChanged", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"component" to component, |
|||
"isActive" to isActive |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onError(component: String, error: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "error", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"component" to component, |
|||
"error" to error |
|||
) |
|||
) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* Doubao AST Callback Implementation |
|||
*/ |
|||
class DoubaoAstCallback( |
|||
private val serviceId: String, |
|||
private val direction: String, |
|||
private val targetLanguage: String, |
|||
private val eventSender: AstEventSender, |
|||
private val audioWriter: AudioWriter? = null |
|||
) : DoubaoE2ETranslateHelper.Callback { |
|||
private val tag = "DoubaoAstCallback" |
|||
private val doubaoFinalSourceTextCache: MutableMap<String, String> = mutableMapOf() |
|||
|
|||
override fun onSessionStarted(sessionId: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "serviceInitialized", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"sessionId" to sessionId |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onPartialSourceText(sessionId: String, text: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "recognizing", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to sessionId, |
|||
"text" to text, |
|||
"language" to direction.split("->").firstOrNull().orEmpty() |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onFinalSourceText(sessionId: String, finalText: String) { |
|||
val key = "$serviceId:$sessionId" |
|||
doubaoFinalSourceTextCache[key] = finalText |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "recognized", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to sessionId, |
|||
"text" to finalText, |
|||
"language" to direction.split("->").firstOrNull().orEmpty() |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onPartialText(sessionId: String, text: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translatedInterim", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to sessionId, |
|||
"originalText" to "", |
|||
"translatedText" to text, |
|||
"targetLanguage" to targetLanguage |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onPartialAudio(sessionId: String, data: ByteArray) { |
|||
audioWriter?.write(data) |
|||
} |
|||
|
|||
override fun onSessionFinished( |
|||
sessionId: String, |
|||
finalText: String, |
|||
finalAudio: ByteArray |
|||
) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translated", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to sessionId, |
|||
"originalText" to (doubaoFinalSourceTextCache.remove("$serviceId:$sessionId") ?: ""), |
|||
"translatedText" to finalText, |
|||
"targetLanguage" to targetLanguage |
|||
) |
|||
) |
|||
|
|||
// if (finalAudio.isNotEmpty()) { |
|||
// audioWriter?.write(finalAudio) |
|||
// } |
|||
} |
|||
|
|||
override fun onSessionError(sessionId: String, code: Int, message: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "error", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"component" to "doubaoAst", |
|||
"error" to message, |
|||
"code" to code |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onFinalTranslatedText(sessionId: String, finalText: String) { |
|||
val key = "$serviceId:$sessionId" |
|||
val original = doubaoFinalSourceTextCache.remove(key) ?: "" |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translated", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to sessionId, |
|||
"originalText" to original, |
|||
"translatedText" to finalText, |
|||
"targetLanguage" to targetLanguage |
|||
) |
|||
) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* Aliyun Bailian AST Callback Implementation |
|||
*/ |
|||
class AliyunAstCallback( |
|||
private val serviceId: String, |
|||
private val direction: String, |
|||
private val targetLanguage: String, |
|||
private val eventSender: AstEventSender, |
|||
private val audioWriter: AudioWriter? = null |
|||
) : AliyunBailianE2EHelper.Callback { |
|||
private val tag = "AliyunAstCallback" |
|||
private val sourceTextCache: MutableMap<String, String> = mutableMapOf() |
|||
|
|||
override fun onSessionStarted(sessionId: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "serviceInitialized", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"sessionId" to sessionId |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onPartialSourceText(sessionId: String, text: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "recognizing", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to sessionId, |
|||
"text" to text, |
|||
"language" to direction.split("->").firstOrNull().orEmpty() |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onFinalSourceText(sessionId: String, finalText: String) { |
|||
val key = "$serviceId:$sessionId" |
|||
sourceTextCache[key] = finalText |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "recognized", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to sessionId, |
|||
"text" to finalText, |
|||
"language" to direction.split("->").firstOrNull().orEmpty() |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onPartialText(sessionId: String, text: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translatedInterim", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to sessionId, |
|||
"originalText" to "", |
|||
"translatedText" to text, |
|||
"targetLanguage" to targetLanguage |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onPartialAudio(sessionId: String, data: ByteArray) { |
|||
audioWriter?.write(data) |
|||
} |
|||
|
|||
override fun onSessionFinished( |
|||
sessionId: String, |
|||
finalText: String, |
|||
finalAudio: ByteArray |
|||
) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translated", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to sessionId, |
|||
"originalText" to (sourceTextCache.remove("$serviceId:$sessionId") ?: ""), |
|||
"translatedText" to finalText, |
|||
"targetLanguage" to targetLanguage |
|||
) |
|||
) |
|||
|
|||
// if (finalAudio.isNotEmpty()) { |
|||
// audioWriter?.write(finalAudio) |
|||
// } |
|||
} |
|||
|
|||
override fun onSessionError(sessionId: String, code: Int, message: String) { |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "error", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"component" to "aliyunAst", |
|||
"error" to message, |
|||
"code" to code |
|||
) |
|||
) |
|||
} |
|||
|
|||
override fun onFinalTranslatedText(sessionId: String, finalText: String) { |
|||
val key = "$serviceId:$sessionId" |
|||
val original = sourceTextCache.remove(key) ?: "" |
|||
eventSender.send( |
|||
mapOf( |
|||
"type" to "translated", |
|||
"serviceId" to serviceId, |
|||
"direction" to direction, |
|||
"utteranceId" to sessionId, |
|||
"originalText" to original, |
|||
"translatedText" to finalText, |
|||
"targetLanguage" to targetLanguage |
|||
) |
|||
) |
|||
} |
|||
} |
|||
@ -0,0 +1,473 @@ |
|||
package com.example.astclient |
|||
|
|||
import android.content.Context |
|||
import android.net.Uri |
|||
import android.util.Log |
|||
import okhttp3.* |
|||
import okio.ByteString |
|||
import java.io.ByteArrayOutputStream |
|||
import java.io.File |
|||
import java.io.InputStream |
|||
import java.util.UUID |
|||
import java.util.concurrent.TimeUnit |
|||
import java.util.concurrent.atomic.AtomicBoolean |
|||
import kotlinx.coroutines.* |
|||
import data.speech.ast.TranslateRequest |
|||
import data.speech.ast.TranslateResponse |
|||
import data.speech.ast.ReqParams |
|||
import data.speech.understanding.User |
|||
import data.speech.understanding.Audio |
|||
import data.speech.common.RequestMeta |
|||
import data.speech.event.Type |
|||
|
|||
/** |
|||
* AST 流式翻译助手，提供与 `IntegratedSpeechTranslationService` 类似的接口， |
|||
* 以便在插件中通过 `pushAudioData` 输入音频并在会话结束时获取合成音频。 |
|||
*/ |
|||
class DoubaoE2ETranslateHelper( |
|||
private val context: Context |
|||
) { |
|||
private val TAG = "DoubaoE2ETranslateHelper" |
|||
interface Callback { |
|||
/** 会话开始 */ |
|||
fun onSessionStarted(sessionId: String) |
|||
|
|||
/** 增量文本到达 */ |
|||
fun onPartialText(sessionId: String, text: String) |
|||
|
|||
/** 增量源文本到达（实时识别的原文字幕） */ |
|||
fun onPartialSourceText(sessionId: String, text: String) |
|||
|
|||
/** 源文本最终结果到达（原文识别完成） */ |
|||
fun onFinalSourceText(sessionId: String, finalText: String) |
|||
|
|||
/** 翻译最终结果到达（译文生成完成） */ |
|||
fun onFinalTranslatedText(sessionId: String, finalText: String) |
|||
|
|||
/** 增量音频片段到达（服务端返回的目标音频） */ |
|||
fun onPartialAudio(sessionId: String, data: ByteArray) |
|||
|
|||
/** 会话结束，返回完整文本与完整合成音频 */ |
|||
fun onSessionFinished(sessionId: String, finalText: String, finalAudio: ByteArray) |
|||
|
|||
/** 会话失败或取消 */ |
|||
fun onSessionError(sessionId: String, code: Int, message: String) |
|||
} |
|||
|
|||
private var conf: Config = Config() |
|||
private var callback: Callback? = null |
|||
|
|||
private var client: OkHttpClient? = null |
|||
private var webSocket: WebSocket? = null |
|||
private var sessionId: String = "" |
|||
private val sessionLock = Any() |
|||
|
|||
private val recvAudio = ByteArrayOutputStream() |
|||
private val recvText = mutableListOf<String>() |
|||
private val recvSourceText = mutableListOf<String>() |
|||
|
|||
private var audioChunkBuffer = ByteArray(0) |
|||
|
|||
private val isStarted = AtomicBoolean(false) |
|||
private val scope = CoroutineScope(Dispatchers.IO + SupervisorJob()) |
|||
|
|||
/** |
|||
* 初始化助手，设置配置与回调。 |
|||
*/ |
|||
fun initialize(config: Config, cb: Callback): Boolean { |
|||
Log.d(TAG, "initialize: wsUrl=${config.wsUrl}, resourceId=${config.resourceId}") |
|||
conf = config |
|||
callback = cb |
|||
client = OkHttpClient.Builder() |
|||
.pingInterval(30, TimeUnit.SECONDS) |
|||
.readTimeout(0, TimeUnit.SECONDS) |
|||
.build() |
|||
Log.d(TAG, "initialize: client created") |
|||
startContinuousTranslation() |
|||
return true |
|||
} |
|||
|
|||
/** |
|||
* 重启会话（用于超时自动重连） |
|||
*/ |
|||
private fun restartSession() { |
|||
Log.i(TAG, "restartSession: performing auto-restart...") |
|||
try { webSocket?.close(1000, "restarting") } catch (_: Exception) {} |
|||
webSocket = null |
|||
isStarted.set(false) |
|||
|
|||
try { |
|||
java.util.Timer().schedule(object : java.util.TimerTask() { |
|||
override fun run() { |
|||
Log.i(TAG, "restartSession: timer task running") |
|||
if (!isStarted.get()) { |
|||
Log.i(TAG, "restartSession: calling startContinuousTranslation") |
|||
val result = startContinuousTranslation() |
|||
Log.i(TAG, "restartSession: startContinuousTranslation result=$result") |
|||
} else { |
|||
Log.i(TAG, "restartSession: already started, skipping") |
|||
} |
|||
} |
|||
}, 200) |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "restartSession: timer failed", e) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 启动会话，建立 WebSocket 并发送 StartSession。 |
|||
*/ |
|||
fun startContinuousTranslation(): Boolean { |
|||
Log.d(TAG, "startContinuousTranslation: isStarted=${isStarted.get()} clientIsNull=${client==null}") |
|||
if (client == null) return false |
|||
if (isStarted.get()) return true |
|||
|
|||
synchronized(sessionLock) { |
|||
sessionId = UUID.randomUUID().toString() |
|||
Log.d(TAG, "startContinuousTranslation: sessionId=${sessionId}") |
|||
recvAudio.reset() |
|||
recvText.clear() |
|||
} |
|||
|
|||
val connId = UUID.randomUUID().toString() |
|||
val request = Request.Builder() |
|||
.url(conf.wsUrl) |
|||
.header("X-Api-App-Key", conf.appKey) |
|||
.header("X-Api-Access-Key", conf.accessKey) |
|||
.header("X-Api-Resource-Id", conf.resourceId) |
|||
.header("X-Api-Connect-Id", connId) |
|||
.build() |
|||
|
|||
val listener = object : WebSocketListener() { |
|||
/** |
|||
* WebSocket 连接建立时回调。 |
|||
* 设置内部会话状态，构造并发送 `StartSession` 请求， |
|||
* 然后通过 `callback.onSessionStarted` 通知上层会话已开始。 |
|||
* |
|||
* @param ws 当前的 WebSocket 连接 |
|||
* @param response 握手响应，包含状态码与头信息 |
|||
*/ |
|||
override fun onOpen(ws: WebSocket, response: Response) { |
|||
Log.d(TAG, "onOpen: code=${response.code} logid=${response.header("X-Tt-Logid")}") |
|||
webSocket = ws |
|||
isStarted.set(true) |
|||
val startReq = makeStartRequest(sessionId) |
|||
ws.send(ByteString.of(*startReq.toByteArray())) |
|||
Log.d(TAG, "onOpen: StartSession sent") |
|||
callback?.onSessionStarted(sessionId) |
|||
} |
|||
|
|||
/** |
|||
* 接收服务端消息的回调。 |
|||
* 解析为 `TranslateResponse` 后依据 `event` 分类处理： |
|||
* - `UsageResponse`：仅记录日志后返回； |
|||
* - `SessionFailed`/`SessionCanceled`：触发错误回调并关闭连接； |
|||
* - `SessionFinished`：汇总累计的文本与音频，触发完成回调并正常关闭； |
|||
* - 其他：追加增量音频与文本，并通过 `callback` 将增量结果向上层反馈。 |
|||
* |
|||
* @param ws 当前的 WebSocket 连接 |
|||
* @param bytes 服务端下发的二进制消息 |
|||
*/ |
|||
override fun onMessage(ws: WebSocket, bytes: ByteString) { |
|||
Log.d(TAG, "onMessage: ${bytes.size} bytes") |
|||
val resp = TranslateResponse.parseFrom(bytes.toByteArray()) |
|||
val event = resp.event |
|||
val text = resp.text |
|||
val data = resp.data.toByteArray() |
|||
Log.d(TAG, "onMessage: event=${event} textLen=${text?.length ?: 0} audioLen=${data.size}") |
|||
|
|||
if (event == Type.UsageResponse) { |
|||
Log.d(TAG, "onMessage: UsageResponse") |
|||
return |
|||
} |
|||
|
|||
if (event == Type.SessionFailed || event == Type.SessionCanceled) { |
|||
val msg = if (resp.hasResponseMeta()) resp.responseMeta.message else "" |
|||
Log.e(TAG, "onMessage: session error event=${event} msg=${msg}") |
|||
|
|||
if (msg.contains("Timeout waiting next packet", ignoreCase = true)) { |
|||
Log.i(TAG, "onMessage: detected timeout error, triggering auto-restart") |
|||
restartSession() |
|||
return |
|||
} |
|||
|
|||
callback?.onSessionError(sessionId, 1011, msg) |
|||
ws.close(1011, msg) |
|||
return |
|||
} |
|||
|
|||
if (event == Type.SessionFinished) { |
|||
Log.d(TAG, "onMessage: SessionFinished") |
|||
val finalText = recvText.joinToString(" ") |
|||
val finalAudio = recvAudio.toByteArray() |
|||
callback?.onSessionFinished(sessionId, finalText, finalAudio) |
|||
ws.close(1000, "Completed") |
|||
return |
|||
} |
|||
|
|||
if (data.isNotEmpty()) { |
|||
recvAudio.write(data) |
|||
Log.d(TAG, "onMessage: partial audio appended size=${data.size}") |
|||
processAudioChunk(data) |
|||
} |
|||
if (!text.isNullOrBlank()) { |
|||
when (event) { |
|||
Type.SourceSubtitleStart -> { |
|||
Log.d(TAG, "onMessage: SourceSubtitleStart") |
|||
recvSourceText.clear() |
|||
} |
|||
Type.SourceSubtitleResponse -> { |
|||
Log.d(TAG, "onMessage: SourceSubtitleResponse text='${text}'") |
|||
recvSourceText.add(text) |
|||
callback?.onPartialSourceText(sessionId, text) |
|||
} |
|||
Type.SourceSubtitleEnd -> { |
|||
Log.d(TAG, "onMessage: SourceSubtitleEnd") |
|||
val finalSrc = recvSourceText.joinToString(" ") |
|||
callback?.onFinalSourceText(sessionId, finalSrc) |
|||
} |
|||
Type.TranslationSubtitleStart -> { |
|||
Log.d(TAG, "onMessage: TranslationSubtitleStart") |
|||
recvText.clear() |
|||
} |
|||
Type.TranslationSubtitleResponse -> { |
|||
Log.d(TAG, "onMessage: TranslationSubtitleResponse text='${text}'") |
|||
recvText.add(text) |
|||
callback?.onPartialText(sessionId, text) |
|||
} |
|||
Type.TranslationSubtitleEnd -> { |
|||
Log.d(TAG, "onMessage: TranslationSubtitleEnd") |
|||
val finalTgt = recvText.joinToString(" ") |
|||
callback?.onFinalTranslatedText(sessionId, finalTgt) |
|||
// 一句话结束后，重新生成 sessionId 并在同一连接上发送新的 StartSession |
|||
startNewSubSession(ws) |
|||
} |
|||
else -> { |
|||
// 其他文本事件按译文处理以保持兼容 |
|||
// recvText.add(text) |
|||
Log.d(TAG, "onMessage: partial text appended text='${text}' (compat)") |
|||
//callback?.onPartialText(sessionId, text) |
|||
} |
|||
} |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* WebSocket 连接发生错误时的回调。 |
|||
* 读取异常信息并通过 `callback.onSessionError` 向上层报告， |
|||
* 不在此处执行重试逻辑，由上层决定后续策略。 |
|||
* |
|||
* @param ws 当前的 WebSocket 连接 |
|||
* @param t 抛出的异常 |
|||
* @param response 可选的响应对象（可能为 null） |
|||
*/ |
|||
override fun onFailure(ws: WebSocket, t: Throwable, response: Response?) { |
|||
val msg = t.message ?: "unknown" |
|||
Log.e(TAG, "onFailure: ${msg} code=${response?.code}") |
|||
callback?.onSessionError(sessionId, 1011, msg) |
|||
} |
|||
|
|||
/** |
|||
* WebSocket 连接关闭时的回调（正常或异常关闭）。 |
|||
* 释放音频缓冲并重置会话标志，便于下一次会话重新开始。 |
|||
* |
|||
* @param ws 当前的 WebSocket 连接 |
|||
* @param code 关闭码 |
|||
* @param reason 关闭原因描述 |
|||
*/ |
|||
override fun onClosed(ws: WebSocket, code: Int, reason: String) { |
|||
Log.d(TAG, "onClosed: code=${code} reason=${reason}") |
|||
try { recvAudio.close() } catch (_: Exception) {} |
|||
isStarted.set(false) |
|||
} |
|||
} |
|||
|
|||
Log.d(TAG, "startContinuousTranslation: connecting wsUrl=${conf.wsUrl}") |
|||
client!!.newWebSocket(request, listener) |
|||
Log.d(TAG, "startContinuousTranslation: newWebSocket invoked") |
|||
return true |
|||
} |
|||
|
|||
/** |
|||
* 准备下一句话的接收缓冲区。 |
|||
* 不再重新生成 SessionID，保持整个 WebSocket 连接使用同一个 SessionID。 |
|||
* |
|||
* @param ws 当前的 WebSocket 连接 |
|||
*/ |
|||
private fun startNewSubSession(ws: WebSocket) { |
|||
synchronized(sessionLock) { |
|||
// sessionId 保持不变，避免服务端报错 "unexpected client event: current state=Started" |
|||
// sessionId = UUID.randomUUID().toString() |
|||
Log.d(TAG, "startNewSubSession: keeping sessionId=${sessionId} clearing buffers") |
|||
recvText.clear() |
|||
recvSourceText.clear() |
|||
} |
|||
} |
|||
|
|||
private fun processAudioChunk( |
|||
incoming: ByteArray, |
|||
multiple: Int = 1280 |
|||
) { |
|||
if (incoming.isNotEmpty()) { |
|||
audioChunkBuffer += incoming |
|||
} |
|||
val sendLen = (audioChunkBuffer.size / multiple) * multiple |
|||
if (sendLen > 0) { |
|||
val chunk = audioChunkBuffer.copyOfRange(0, sendLen) |
|||
callback?.onPartialAudio(sessionId, chunk) |
|||
audioChunkBuffer = audioChunkBuffer.copyOfRange(sendLen, audioChunkBuffer.size) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 推送一段 PCM/WAV 音频数据到服务端。 |
|||
*/ |
|||
fun pushAudioData(data: ByteArray): Boolean { |
|||
// Log.d(TAG, "pushAudioData: size=${data.size} isStarted=${isStarted.get()} wsIsNull=${webSocket==null}") |
|||
val ws = webSocket ?: return false |
|||
if (!isStarted.get()) return false |
|||
synchronized(sessionLock) { |
|||
val req = makeChunkRequest(sessionId, data) |
|||
val ok = ws.send(ByteString.of(*req.toByteArray())) |
|||
// Log.d(TAG, "pushAudioData: sent=${ok}") |
|||
return ok |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 停止会话，发送 FinishSession，并等待服务端返回。 |
|||
*/ |
|||
fun stopContinuousTranslation(): Boolean { |
|||
Log.d(TAG, "stopContinuousTranslation: wsIsNull=${webSocket==null}") |
|||
val ws = webSocket ?: return false |
|||
synchronized(sessionLock) { |
|||
val finishReq = makeFinishRequest(sessionId) |
|||
val ok = ws.send(ByteString.of(*finishReq.toByteArray())) |
|||
Log.d(TAG, "stopContinuousTranslation: sent=${ok}") |
|||
return ok |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 释放资源，关闭会话。 |
|||
*/ |
|||
fun dispose() { |
|||
Log.d(TAG, "dispose: begin") |
|||
try { webSocket?.close(1000, "dispose") } catch (_: Exception) {} |
|||
webSocket = null |
|||
isStarted.set(false) |
|||
try { recvAudio.close() } catch (_: Exception) {} |
|||
scope.cancel() |
|||
Log.d(TAG, "dispose: done") |
|||
} |
|||
|
|||
/** |
|||
* 构造 StartSession 请求 |
|||
*/ |
|||
private fun makeStartRequest(sessionId: String): TranslateRequest { |
|||
Log.d(TAG, "makeStartRequest: sessionId=${sessionId}") |
|||
val user = User.newBuilder() |
|||
.setUid("ast_android_client") |
|||
.setDid("ast_android_client") |
|||
.setPlatform("android") |
|||
.build() |
|||
|
|||
val sourceAudio = Audio.newBuilder() |
|||
.setFormat("wav") |
|||
.setRate(16000) |
|||
.setBits(16) |
|||
.setChannel(1) |
|||
.build() |
|||
|
|||
val targetAudio = Audio.newBuilder() |
|||
.setFormat("pcm") |
|||
.setRate(16000) |
|||
.build() |
|||
|
|||
val params = ReqParams.newBuilder() |
|||
.setMode("s2s") |
|||
.setSourceLanguage(conf.sourceLanguage) |
|||
.setTargetLanguage(conf.targetLanguage) |
|||
.build() |
|||
|
|||
val meta = RequestMeta.newBuilder() |
|||
.setSessionID(sessionId) |
|||
.build() |
|||
|
|||
return TranslateRequest.newBuilder() |
|||
.setRequestMeta(meta) |
|||
.setEvent(Type.StartSession) |
|||
.setUser(user) |
|||
.setSourceAudio(sourceAudio) |
|||
.setTargetAudio(targetAudio) |
|||
.setRequest(params) |
|||
.build() |
|||
} |
|||
|
|||
|
|||
|
|||
/** |
|||
* 构造音频片段请求 |
|||
*/ |
|||
private fun makeChunkRequest(sessionId: String, chunk: ByteArray): TranslateRequest { |
|||
//Log.d(TAG, "makeChunkRequest: sessionId=${sessionId} chunkSize=${chunk.size}") |
|||
val sourceAudio = Audio.newBuilder() |
|||
.setFormat("wav") |
|||
.setRate(16000) |
|||
.setBits(16) |
|||
.setChannel(1) |
|||
.setBinaryData(com.google.protobuf.ByteString.copyFrom(chunk)) |
|||
.build() |
|||
|
|||
val meta = RequestMeta.newBuilder() |
|||
.setSessionID(sessionId) |
|||
.build() |
|||
|
|||
return TranslateRequest.newBuilder() |
|||
.setRequestMeta(meta) |
|||
.setEvent(Type.TaskRequest) |
|||
.setUser( |
|||
User.newBuilder() |
|||
.setUid("ast_android_client") |
|||
.setDid("ast_android_client") |
|||
.build() |
|||
) |
|||
.setSourceAudio(sourceAudio) |
|||
.build() |
|||
} |
|||
|
|||
/** |
|||
* 构造结束请求 |
|||
*/ |
|||
private fun makeFinishRequest(sessionId: String): TranslateRequest { |
|||
Log.d(TAG, "makeFinishRequest: sessionId=${sessionId}") |
|||
val meta = RequestMeta.newBuilder() |
|||
.setSessionID(sessionId) |
|||
.build() |
|||
|
|||
return TranslateRequest.newBuilder() |
|||
.setRequestMeta(meta) |
|||
.setEvent(Type.FinishSession) |
|||
.setUser( |
|||
User.newBuilder() |
|||
.setUid("ast_android_client") |
|||
.setDid("ast_android_client") |
|||
.build() |
|||
) |
|||
.setSourceAudio(Audio.newBuilder().build()) |
|||
.build() |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 客户端配置，包含服务端地址与鉴权信息。 |
|||
*/ |
|||
data class Config( |
|||
val wsUrl: String = "wss://openspeech.bytedance.com/api/v4/ast/v2/translate", |
|||
val appKey: String = "", |
|||
val accessKey: String = "", |
|||
val resourceId: String = "volc.service_type.10053", |
|||
val sourceLanguage: String = "zh", |
|||
val targetLanguage: String = "en", |
|||
val clientWaitMs: Long = 60_000 // 等待会话结束的最大时间，可按需调整 |
|||
) |
|||
@ -0,0 +1,376 @@ |
|||
package com.yunqiinnovation.azure_speech |
|||
|
|||
import android.content.Context |
|||
import android.os.Handler |
|||
import android.os.Looper |
|||
import android.util.Base64 |
|||
import android.util.Log |
|||
import okhttp3.* |
|||
import okio.ByteString.Companion.toByteString |
|||
import org.json.JSONObject |
|||
import java.net.URLEncoder |
|||
import java.nio.charset.StandardCharsets |
|||
import java.security.MessageDigest |
|||
import java.util.UUID |
|||
import javax.crypto.Mac |
|||
import javax.crypto.spec.SecretKeySpec |
|||
|
|||
// 讯飞语音识别助手 |
|||
//调用讯飞开放平台的实时语音转写大模型接口 |
|||
// 参考文档：https://www.xfyun.cn/doc/spark/asr_llm/rtasr_llm.html |
|||
|
|||
/** |
|||
* 讯飞语音识别助手，封装与讯飞 AST 实时转写接口的交互。 |
|||
* 负责建立 WebSocket 连接、发送音频数据并解析识别结果， |
|||
* 并通过回调向上层报告会话开始/识别中/最终结果/错误等事件。 |
|||
*/ |
|||
class IflytekAsrHelper(private val context: Context) { |
|||
private val tag = "IflytekAsrHelper" |
|||
|
|||
private var appId: String = "" |
|||
private var accessKeyId: String = "" |
|||
private var accessKeySecret: String = "" |
|||
|
|||
private var webSocket: WebSocket? = null |
|||
private val client = OkHttpClient() |
|||
|
|||
private var currentSessionId = "" |
|||
private var sessionId = "" |
|||
private var lastIntermediateResult = "" |
|||
|
|||
private var currentLanguage = "zh-CN" |
|||
private var isAutoDetectLanguage = false |
|||
|
|||
// 是否第去除一个文字的标点符号 |
|||
private var removeFirstPunctuation = true |
|||
|
|||
/** |
|||
* 使用指定配置构造识别助手。 |
|||
* @param context Android 上下文 |
|||
* @param appId 讯飞应用 ID |
|||
* @param accessKeyId 访问密钥 ID |
|||
* @param accessKeySecret 访问密钥 Secret |
|||
*/ |
|||
constructor( |
|||
context: Context, |
|||
appId: String, |
|||
accessKeyId: String, |
|||
accessKeySecret: String |
|||
) : this(context) { |
|||
updateConfig(appId = appId, accessKeyId = accessKeyId, accessKeySecret = accessKeySecret) |
|||
} |
|||
|
|||
/** |
|||
* 更新讯飞接口鉴权配置。 |
|||
* @param appId 讯飞应用 ID |
|||
* @param accessKeyId 访问密钥 ID |
|||
* @param accessKeySecret 访问密钥 Secret |
|||
*/ |
|||
fun updateConfig(appId: String, accessKeyId: String, accessKeySecret: String) { |
|||
this.appId = appId |
|||
this.accessKeyId = accessKeyId |
|||
this.accessKeySecret = accessKeySecret |
|||
} |
|||
|
|||
/** |
|||
* 校验当前配置是否完整可用。 |
|||
* @return 若 `appId`、`accessKeyId`、`accessKeySecret` 均不为空返回 `true`，否则 `false` |
|||
*/ |
|||
private fun isConfigValid(): Boolean { |
|||
return appId.isNotBlank() && accessKeyId.isNotBlank() && accessKeySecret.isNotBlank() |
|||
} |
|||
|
|||
/** |
|||
* 启动实时识别会话并建立 WebSocket 连接。 |
|||
* 会在连接建立后触发 `onSessionStarted`，在收到中间结果触发 `onRecognizing`， |
|||
* 收到最终结果触发 `onResult` 并重置会话 ID，连接关闭时触发 `onSessionStopped`。 |
|||
* 若配置缺失或连接异常，将触发 `onError`。 |
|||
* @param callback 识别回调接口 |
|||
* @param language 指定语言代码（如 `zh-CN`、`en-US`），当 `isAutoDetect` 为 `false` 时生效 |
|||
* @param isAutoDetect 是否自动检测中英文（根据结果文本是否包含中文字符判断） |
|||
* @param isRemoveFirstPunctuation 是否去除结果首字符的标点符号 |
|||
*/ |
|||
fun start(callback: AzureAsrHelper.ContinuousRecognizeCallback, language: String, isAutoDetect: Boolean, isRemoveFirstPunctuation: Boolean) { |
|||
try { |
|||
webSocket?.cancel() |
|||
webSocket = null |
|||
} catch (e: Exception) { |
|||
Log.e(tag, "Clear previous websocket failed: ${e.message}") |
|||
} |
|||
|
|||
currentSessionId = UUID.randomUUID().toString() |
|||
lastIntermediateResult = "" |
|||
|
|||
currentLanguage = language |
|||
isAutoDetectLanguage = isAutoDetect |
|||
removeFirstPunctuation = isRemoveFirstPunctuation |
|||
|
|||
if (!isConfigValid()) { |
|||
callback.onError(currentSessionId, 1007, "Iflytek config missing") |
|||
return |
|||
} |
|||
|
|||
val url = getWebSocketUrl() |
|||
val request = Request.Builder().url(url).build() |
|||
|
|||
webSocket = client.newWebSocket(request, object : WebSocketListener() { |
|||
override fun onOpen(webSocket: WebSocket, response: Response) { |
|||
Log.d(tag, "WebSocket Opened") |
|||
callback.onSessionStarted(currentSessionId) |
|||
} |
|||
|
|||
override fun onMessage(webSocket: WebSocket, text: String) { |
|||
renderResult(text, callback) |
|||
} |
|||
|
|||
override fun onClosing(webSocket: WebSocket, code: Int, reason: String) { |
|||
Log.d(tag, "WebSocket Closing: $code / $reason") |
|||
webSocket.close(1000, null) |
|||
} |
|||
|
|||
override fun onClosed(webSocket: WebSocket, code: Int, reason: String) { |
|||
Log.d(tag, "WebSocket Closed: $code / $reason") |
|||
if (currentSessionId.isNotEmpty() && lastIntermediateResult.isNotEmpty()) { |
|||
val detectedLanguage = if (isAutoDetectLanguage) { |
|||
if (lastIntermediateResult.contains(Regex("[\\u4e00-\\u9fa5]"))) { |
|||
"zh-CN" |
|||
} else { |
|||
"en-US" |
|||
} |
|||
} else { |
|||
currentLanguage |
|||
} |
|||
callback.onResult(currentSessionId, lastIntermediateResult, detectedLanguage) |
|||
} |
|||
callback.onSessionStopped(currentSessionId) |
|||
} |
|||
|
|||
override fun onFailure(webSocket: WebSocket, t: Throwable, response: Response?) { |
|||
Log.e(tag, "WebSocket Error", t) |
|||
callback.onError(currentSessionId, 1004, "Iflytek connection failed: ${t.message}") |
|||
} |
|||
}) |
|||
} |
|||
|
|||
/** |
|||
* 发送一帧原始音频数据到服务端。 |
|||
* @param frameBuffer 音频字节数组（例如 16kHz、16bit PCM） |
|||
*/ |
|||
fun sendAudio(frameBuffer: ByteArray) { |
|||
webSocket?.let { ws: WebSocket -> |
|||
ws.send(frameBuffer.toByteString(0, frameBuffer.size)) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 主动结束当前识别会话。 |
|||
* 向服务端发送结束标记并在 1 秒后关闭连接。 |
|||
* 若关闭失败会记录错误日志。 |
|||
*/ |
|||
fun stop() { |
|||
try { |
|||
webSocket?.let { ws: WebSocket -> |
|||
ws.send("{\"end\": true, \"sessionId\": \"$sessionId\"}") |
|||
Handler(Looper.getMainLooper()).postDelayed({ |
|||
try { |
|||
ws.close(1000, "User stopped") |
|||
} catch (e: Exception) { |
|||
Log.e(tag, "Close Iflytek failed: ${e.message}") |
|||
} |
|||
}, 1000) |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(tag, "Stop Iflytek failed: ${e.message}") |
|||
} finally { |
|||
webSocket = null |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 解析服务端文本消息并根据类型触发相应回调。 |
|||
* 支持 `action`（握手）、`result`（识别结果）、`error`（错误）三类消息。 |
|||
* 在 `result` 中会根据 `st.type` 区分中间结果与最终结果，并可进行语言自动检测。 |
|||
* @param resultData 服务端返回的 JSON 字符串 |
|||
* @param callback 识别回调接口 |
|||
*/ |
|||
private fun renderResult( |
|||
resultData: String, |
|||
callback: AzureAsrHelper.ContinuousRecognizeCallback |
|||
) { |
|||
try { |
|||
val jsonData = JSONObject(resultData) |
|||
Log.d(tag, "-------------------$jsonData") |
|||
val msgType = jsonData.optString("msg_type") |
|||
|
|||
when (msgType) { |
|||
"action" -> { |
|||
val sid = jsonData.optString("sessionId") |
|||
if (!sid.isNullOrEmpty()) { |
|||
sessionId = sid |
|||
} |
|||
Log.d(tag, "Handshake success") |
|||
} |
|||
|
|||
"result" -> { |
|||
val dataStr = jsonData.optString("data") |
|||
val data = JSONObject(dataStr) |
|||
|
|||
val ls = data.optBoolean("ls") |
|||
val cn = data.optJSONObject("cn") ?: return |
|||
val st = cn.optJSONObject("st") ?: return |
|||
val rt = st.optJSONArray("rt") ?: return |
|||
|
|||
var resultTextTemp = "" |
|||
var punctuationCount = 0 |
|||
var wordCount = 0 |
|||
|
|||
for (i in 0 until rt.length()) { |
|||
val j = rt.getJSONObject(i) |
|||
val ws = j.optJSONArray("ws") ?: continue |
|||
for (k in 0 until ws.length()) { |
|||
val kObj = ws.getJSONObject(k) |
|||
val cw = kObj.optJSONArray("cw") ?: continue |
|||
if(removeFirstPunctuation) { |
|||
// 如果resultTextTemp中第一个文字是标点符号，则去除第一个文字的标点符号 |
|||
for (l in 0 until cw.length()) { |
|||
val lObj = cw.getJSONObject(l) |
|||
val w = lObj.optString("w") |
|||
val wp = lObj.optString("wp") |
|||
|
|||
if (wp == "p") { |
|||
punctuationCount++ |
|||
if (resultTextTemp.isEmpty()) { |
|||
continue |
|||
} |
|||
} else { |
|||
wordCount++ |
|||
} |
|||
resultTextTemp += w |
|||
} |
|||
} else { |
|||
// 不去除resultTextTemp第一个文字的标点符号 |
|||
for (l in 0 until cw.length()) { |
|||
val lObj = cw.getJSONObject(l) |
|||
resultTextTemp += lObj.optString("w") |
|||
if (lObj.optString("wp") == "p") { |
|||
punctuationCount++ |
|||
} else { |
|||
wordCount++ |
|||
} |
|||
} |
|||
} |
|||
} |
|||
} |
|||
|
|||
if (ls && wordCount == 0 && punctuationCount == 1) { |
|||
callback.onSessionStopped(currentSessionId) |
|||
return |
|||
} |
|||
|
|||
val detectedLanguage = if (isAutoDetectLanguage) { |
|||
if (resultTextTemp.contains(Regex("[\\u4e00-\\u9fa5]"))) { |
|||
"zh-CN" |
|||
} else { |
|||
"en-US" |
|||
} |
|||
} else { |
|||
currentLanguage |
|||
} |
|||
|
|||
if (st.optInt("type") == 0) { |
|||
callback.onResult(currentSessionId, resultTextTemp, detectedLanguage) |
|||
currentSessionId = UUID.randomUUID().toString() |
|||
lastIntermediateResult = "" |
|||
} else { |
|||
lastIntermediateResult = resultTextTemp |
|||
callback.onRecognizing(currentSessionId, resultTextTemp, detectedLanguage) |
|||
} |
|||
} |
|||
|
|||
"error" -> { |
|||
callback.onError(currentSessionId, 1005, "Iflytek error: $resultData") |
|||
Log.e(tag, "Error: $resultData") |
|||
} |
|||
|
|||
else -> { |
|||
callback.onError(currentSessionId, 1006, "Iflytek unknown msg_type: $msgType") |
|||
Log.w(tag, "Unknown msg_type: $msgType") |
|||
} |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(tag, "Parse error", e) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 生成用于连接讯飞 AST 的 WebSocket URL。 |
|||
* 包含音频编码、采样率、语言、鉴权参数等，并对参数进行 HMAC-SHA1 签名。 |
|||
* @return 可直接用于发起 WebSocket 请求的完整 URL |
|||
*/ |
|||
private fun getWebSocketUrl(): String { |
|||
val baseWsUrl = "wss://office-api-ast-dx.iflyaisol.com/ast/communicate/v1" |
|||
|
|||
val params = java.util.TreeMap<String, String>() |
|||
params["audio_encode"] = "pcm_s16le" |
|||
params["lang"] = "autodialect" |
|||
params["samplerate"] = "16000" |
|||
params["accessKeyId"] = accessKeyId |
|||
params["appId"] = appId |
|||
params["utc"] = getUtcTime() |
|||
|
|||
val signature = calculateSignature(params) |
|||
params["signature"] = signature |
|||
|
|||
val sb = StringBuilder() |
|||
var first = true |
|||
for ((key, value) in params) { |
|||
if (!first) { |
|||
sb.append("&") |
|||
} |
|||
sb.append(URLEncoder.encode(key, "UTF-8")).append("=").append(URLEncoder.encode(value, "UTF-8")) |
|||
first = false |
|||
} |
|||
|
|||
return "$baseWsUrl?$sb" |
|||
} |
|||
|
|||
/** |
|||
* 获取当前时间的字符串表示。 |
|||
* 注意此处使用 `GMT+8` 时区格式化为 `yyyy-MM-dd'T'HH:mm:ssZ`。 |
|||
* @return 当前时间字符串 |
|||
*/ |
|||
private fun getUtcTime(): String { |
|||
val sdf = java.text.SimpleDateFormat("yyyy-MM-dd'T'HH:mm:ssZ") |
|||
sdf.timeZone = java.util.TimeZone.getTimeZone("GMT+8") |
|||
return sdf.format(java.util.Date()) |
|||
} |
|||
|
|||
/** |
|||
* 计算请求参数的签名值。 |
|||
* 使用 HMAC-SHA1 对未包含 `signature` 且非空的参数进行签名，并返回 Base64 编码的结果。 |
|||
* @param params 待签名的参数映射 |
|||
* @return Base64 编码的签名字符串 |
|||
*/ |
|||
private fun calculateSignature(params: Map<String, String>): String { |
|||
val baseStr = StringBuilder() |
|||
var first = true |
|||
for ((key, value) in params) { |
|||
if ("signature" == key) continue |
|||
if (value.isEmpty()) continue |
|||
|
|||
if (!first) { |
|||
baseStr.append("&") |
|||
} |
|||
baseStr.append(URLEncoder.encode(key, "UTF-8")).append("=").append(URLEncoder.encode(value, "UTF-8")) |
|||
first = false |
|||
} |
|||
|
|||
val mac = Mac.getInstance("HmacSHA1") |
|||
val keySpec = SecretKeySpec(accessKeySecret.toByteArray(StandardCharsets.UTF_8), "HmacSHA1") |
|||
mac.init(keySpec) |
|||
val signBytes = mac.doFinal(baseStr.toString().toByteArray(StandardCharsets.UTF_8)) |
|||
|
|||
return Base64.encodeToString(signBytes, Base64.NO_WRAP) |
|||
} |
|||
} |
|||
|
|||
@ -0,0 +1,659 @@ |
|||
package com.yunqiinnovation.azure_speech |
|||
|
|||
import android.content.Context |
|||
import android.util.Log |
|||
import kotlinx.coroutines.* |
|||
import java.util.concurrent.atomic.AtomicBoolean |
|||
import java.util.concurrent.LinkedBlockingQueue |
|||
import kotlin.coroutines.CoroutineContext |
|||
import org.json.JSONObject |
|||
|
|||
/** |
|||
* 讯飞识别 + 讯飞翻译 + 讯飞TTS 的整合服务 |
|||
* 音频输入 -> 讯飞ASR -> 讯飞翻译 -> 讯飞TTS -> 音频输出/回调 |
|||
*/ |
|||
class IflytekIntegratedSpeechService( |
|||
private val context: Context |
|||
) : CoroutineScope { |
|||
|
|||
companion object { |
|||
private const val TAG = "IflytekIntegratedSpeechService" |
|||
private const val SAMPLE_RATE = 16000 |
|||
private const val CHANNELS = 1 |
|||
private const val BITS_PER_SAMPLE = 16 |
|||
} |
|||
|
|||
private var job = SupervisorJob() |
|||
override val coroutineContext: CoroutineContext |
|||
get() = Dispatchers.Main + job |
|||
|
|||
// 讯飞 TTS 客户端 |
|||
private var ttsClient: IflytekTtsWs? = null |
|||
|
|||
// 讯飞 ASR / 翻译 |
|||
private var asrHelper: IflytekAsrHelper? = null |
|||
private var translationService: IntegratedSpeechTranslationService.TranslationServiceInterface? = null |
|||
|
|||
// 配置与状态 |
|||
private var serviceConfig = ServiceConfiguration() |
|||
private val serviceState = ServiceState() |
|||
private var eventCallback: ServiceEventCallback? = null |
|||
|
|||
// 识别中翻译任务管理 |
|||
private var recognizingTranslationJob: Job? = null |
|||
|
|||
// 录音/推送输入（用于保存/中转） |
|||
private var audioQueue = LinkedBlockingQueue<ByteArray>() |
|||
private val isAudioThreadRunning = AtomicBoolean(false) |
|||
private var audioThread: Thread? = null |
|||
|
|||
private var ttsdata: ByteArray = ByteArray(0) |
|||
private var currentSynthesisId: String? = null |
|||
private var currentSynthesisText: String? = null |
|||
|
|||
/** |
|||
* 服务事件回调接口 |
|||
*/ |
|||
interface ServiceEventCallback { |
|||
fun onServiceInitialized() |
|||
fun onRecognizing(utteranceId: String, text: String, language: String, confidence: Float) |
|||
fun onRecognized(utteranceId: String, text: String, language: String, confidence: Float) |
|||
fun onTranslated(utteranceId: String, originalText: String, translatedText: String, targetLanguage: String) |
|||
fun onInterimTranslated(utteranceId: String, originalText: String, translatedText: String, targetLanguage: String) |
|||
fun onTranslationStarted(utteranceId: String, text: String) |
|||
fun onTranslationFailed(utteranceId: String, text: String, error: String) |
|||
fun onSynthesisStarted(utteranceId: String, text: String) |
|||
fun onSynthesisCompleted(utteranceId: String, text: String) |
|||
fun onSynthesisFailed(utteranceId: String, text: String, error: String) |
|||
fun onSynthesisProgress(utteranceId: String, text: String, progress: Float) |
|||
fun onRecognitionStarted() |
|||
fun onRecognitionStopped() |
|||
fun onStateChanged(component: String, isActive: Boolean) |
|||
fun onError(component: String, error: String) |
|||
fun onSynthesisAudioGenerated(utteranceId: String, text: String, audioData: ByteArray) |
|||
} |
|||
|
|||
/** |
|||
* 服务配置 |
|||
*/ |
|||
data class ServiceConfiguration( |
|||
var sourceLanguage: String = "zh-CN", |
|||
var targetLanguage: String = "en-US", |
|||
var translationSourceLanguage: String = "", |
|||
var translationTargetLanguage: String = "", |
|||
var currentVoice: String = "x4_yezi", |
|||
var speechRate: String = "0%", |
|||
var speechPitch: String = "0%", |
|||
var speechVolume: String = "100%", |
|||
var enableAudioPlayback: Boolean = false, |
|||
var enableAutoLanguageDetection: Boolean = false, |
|||
var translationTimeout: Long = 10000L, |
|||
var removeFirstPunctuation: Boolean = true, |
|||
// 讯飞鉴权 |
|||
var xfyunAppId: String = "", |
|||
var xfyunAccessKeyId: String = "", |
|||
var xfyunAccessKeySecret: String = "" |
|||
) |
|||
|
|||
/** |
|||
* 服务状态 |
|||
*/ |
|||
data class ServiceState( |
|||
val isInitialized: AtomicBoolean = AtomicBoolean(false), |
|||
val isRecognizing: AtomicBoolean = AtomicBoolean(false), |
|||
val isTranslating: AtomicBoolean = AtomicBoolean(false), |
|||
val isSynthesizing: AtomicBoolean = AtomicBoolean(false) |
|||
) |
|||
|
|||
/** |
|||
* 微软 TTS 配置（保留入参兼容，不再使用） |
|||
*/ |
|||
data class AzureConfiguration( |
|||
val subscriptionKey: String, |
|||
val region: String |
|||
) |
|||
|
|||
/** |
|||
* 讯飞翻译配置 |
|||
*/ |
|||
data class IflytekTranslationConfiguration( |
|||
val appId: String, |
|||
val apiKey: String, |
|||
val apiSecret: String, |
|||
val host: String = "ntrans.xfyun.cn", |
|||
val timeoutMs: Long = 8000 |
|||
) { |
|||
fun toConfigMap(): Map<String, String> = mapOf( |
|||
"appId" to appId, |
|||
"apiKey" to apiKey, |
|||
"apiSecret" to apiSecret, |
|||
"host" to host, |
|||
"timeout" to timeoutMs.toString() |
|||
) |
|||
} |
|||
|
|||
/** |
|||
* 初始化整合服务 |
|||
*/ |
|||
suspend fun initialize( |
|||
azureConfig: AzureConfiguration, |
|||
translationConfig: IflytekTranslationConfiguration, |
|||
serviceConfig: ServiceConfiguration? = null, |
|||
callback: ServiceEventCallback |
|||
): Boolean = withContext(Dispatchers.IO) { |
|||
try { |
|||
Log.i(TAG, "开始初始化整合服务") |
|||
dispose() |
|||
this@IflytekIntegratedSpeechService.eventCallback = callback |
|||
serviceConfig?.let { this@IflytekIntegratedSpeechService.serviceConfig = it } |
|||
|
|||
// 不再初始化微软TTS，改用讯飞TTS WebSocket |
|||
|
|||
translationService = IflytekTranslationServiceImpl().apply { |
|||
initialize(translationConfig.toConfigMap()) |
|||
} |
|||
Log.i(TAG, "讯飞翻译服务初始化完成") |
|||
|
|||
asrHelper = IflytekAsrHelper( |
|||
context, |
|||
this@IflytekIntegratedSpeechService.serviceConfig.xfyunAppId, |
|||
this@IflytekIntegratedSpeechService.serviceConfig.xfyunAccessKeyId, |
|||
this@IflytekIntegratedSpeechService.serviceConfig.xfyunAccessKeySecret |
|||
) |
|||
Log.i(TAG, "讯飞ASR助手创建完成") |
|||
|
|||
startAudioThread() |
|||
setupTtsClient() |
|||
Log.i(TAG, "音频线程与TTS合成器已启动") |
|||
|
|||
serviceState.isInitialized.set(true) |
|||
withContext(Dispatchers.Main) { callback.onServiceInitialized() } |
|||
Log.i(TAG, "整合服务初始化成功") |
|||
startContinuousTranslation() |
|||
true |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "初始化失败", e) |
|||
withContext(Dispatchers.Main) { callback.onError("Initialization", "初始化异常: ${e.message}") } |
|||
false |
|||
} |
|||
} |
|||
|
|||
|
|||
/** |
|||
* 启动连续识别与翻译 |
|||
*/ |
|||
fun startContinuousTranslation(): Boolean { |
|||
if (!serviceState.isInitialized.get()) { |
|||
eventCallback?.onError("Service", "服务未初始化") |
|||
return false |
|||
} |
|||
return try { |
|||
Log.i(TAG, "启动连续识别与翻译: autoDetect=${serviceConfig.enableAutoLanguageDetection}, src=${serviceConfig.sourceLanguage}, tgt=${serviceConfig.targetLanguage}") |
|||
asrHelper?.start(object : AzureAsrHelper.ContinuousRecognizeCallback { |
|||
override fun onResult(sessiond: String, text: String, detectedLanguage: String) { |
|||
val confidence = 0.9f |
|||
Log.d(TAG, "最终识别结果(${detectedLanguage}): ${text.take(100)}") |
|||
eventCallback?.onRecognized(sessiond, text, detectedLanguage, confidence) |
|||
launch { processTranslationAndSynthesis(sessiond, text) } |
|||
} |
|||
override fun onRecognizing(sessiond: String, recognizing: String, detectedLanguage: String) { |
|||
val confidence = 0.6f |
|||
Log.d(TAG, "识别中(${detectedLanguage}): ${recognizing.take(80)}") |
|||
eventCallback?.onRecognizing(sessiond, recognizing, detectedLanguage, confidence) |
|||
recognizingTranslationJob?.cancel() |
|||
recognizingTranslationJob = launch { processTranslationOnly(sessiond, recognizing) } |
|||
} |
|||
override fun onSessionStarted(sessiond: String) { |
|||
serviceState.isRecognizing.set(true) |
|||
Log.i(TAG, "识别会话开始: ${sessiond}") |
|||
eventCallback?.onRecognitionStarted() |
|||
eventCallback?.onStateChanged("Recognition", true) |
|||
} |
|||
override fun onSessionStopped(sessiond: String) { |
|||
serviceState.isRecognizing.set(false) |
|||
Log.i(TAG, "识别会话结束: ${sessiond}") |
|||
eventCallback?.onRecognitionStopped() |
|||
eventCallback?.onStateChanged("Recognition", false) |
|||
} |
|||
override fun onCanceled(sessiond: String, reason: String, errorDetails: String) { |
|||
Log.w(TAG, "识别取消: ${reason}, ${errorDetails}") |
|||
eventCallback?.onError("Recognition", "识别取消: $reason, $errorDetails") |
|||
} |
|||
override fun onError(sessiond: String, code: Int, error: String) { |
|||
Log.e(TAG, "讯飞识别错误(${code}): ${error}") |
|||
eventCallback?.onError("Recognition", "讯飞识别错误($code): $error") |
|||
} |
|||
}, serviceConfig.sourceLanguage, serviceConfig.enableAutoLanguageDetection, serviceConfig.removeFirstPunctuation) |
|||
true |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "启动识别失败", e) |
|||
eventCallback?.onError("Service", "启动失败: ${e.message}") |
|||
false |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 停止连续识别与翻译 |
|||
*/ |
|||
fun stopContinuousTranslation() { |
|||
try { |
|||
Log.i(TAG, "停止连续识别与翻译") |
|||
serviceState.isRecognizing.set(false) |
|||
asrHelper?.stop() |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "停止失败", e) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 推送外部音频（转发到讯飞ASR） |
|||
*/ |
|||
fun pushAudioData(audioData: ByteArray) { |
|||
// Log.d(TAG, "推送音频数据: ${audioData.size} 字节") |
|||
audioQueue.offer(audioData) |
|||
asrHelper?.sendAudio(audioData) |
|||
} |
|||
|
|||
/** |
|||
* 仅翻译识别中的文本 |
|||
*/ |
|||
/** |
|||
* 仅翻译识别中的文本（不触发合成） |
|||
*/ |
|||
private suspend fun processTranslationOnly(utteranceId: String, text: String) { |
|||
if (!serviceState.isRecognizing.get()) return |
|||
if (serviceState.isTranslating.get()) return |
|||
|
|||
serviceState.isTranslating.set(true) |
|||
Log.d(TAG, "开始临时翻译: ${text.take(80)}") |
|||
eventCallback?.onTranslationStarted(utteranceId, text) |
|||
eventCallback?.onStateChanged("Translation", true) |
|||
try { |
|||
val (srcLang, tgtLang) = resolveTranslationLanguages() |
|||
val res = withTimeout(serviceConfig.translationTimeout) { |
|||
translationService?.translateText(text, srcLang, tgtLang) |
|||
} |
|||
when { |
|||
res?.success == true && !res.translatedText.isNullOrEmpty() -> { |
|||
Log.d(TAG, "临时翻译结果: ${res.translatedText!!.take(80)}") |
|||
val dst = extractDstFromTranslated(res.translatedText!!) |
|||
if (dst != null && dst.isNotEmpty()) { |
|||
Log.i(TAG, "使用 dst 进行合成: ${dst.take(120)}") |
|||
eventCallback?.onInterimTranslated(utteranceId, text, dst, tgtLang) |
|||
} else { |
|||
Log.w(TAG, "未能解析 dst，回退使用原始翻译文本进行合成") |
|||
|
|||
} |
|||
|
|||
} |
|||
res?.error != null -> { |
|||
Log.w(TAG, "临时翻译错误: ${res.error}") |
|||
eventCallback?.onTranslationFailed(utteranceId, text, res.error!!) |
|||
} |
|||
else -> { |
|||
Log.w(TAG, "临时翻译返回空结果") |
|||
eventCallback?.onTranslationFailed(utteranceId, text, "翻译服务返回空结果") |
|||
} |
|||
} |
|||
} catch (e: TimeoutCancellationException) { |
|||
Log.w(TAG, "临时翻译超时") |
|||
eventCallback?.onTranslationFailed(utteranceId, text, "翻译超时") |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "临时翻译异常: ${e.message}") |
|||
eventCallback?.onTranslationFailed(utteranceId, text, "翻译异常: ${e.message}") |
|||
} finally { |
|||
serviceState.isTranslating.set(false) |
|||
eventCallback?.onStateChanged("Translation", false) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 翻译并合成最终文本 |
|||
*/ |
|||
/** |
|||
* 翻译并合成最终文本 |
|||
*/ |
|||
private suspend fun processTranslationAndSynthesis(utteranceId: String, text: String) { |
|||
if (!serviceState.isRecognizing.get()) return |
|||
if (serviceState.isTranslating.get()) return |
|||
|
|||
serviceState.isTranslating.set(true) |
|||
Log.d(TAG, "开始最终翻译与合成: ${text.take(120)}") |
|||
eventCallback?.onTranslationStarted(utteranceId, text) |
|||
eventCallback?.onStateChanged("Translation", true) |
|||
try { |
|||
val (srcLang, tgtLang) = resolveTranslationLanguages() |
|||
val res = withTimeout(serviceConfig.translationTimeout) { |
|||
translationService?.translateText(text, srcLang, tgtLang) |
|||
} |
|||
serviceState.isTranslating.set(false) |
|||
eventCallback?.onStateChanged("Translation", false) |
|||
when { |
|||
res?.success == true && !res.translatedText.isNullOrEmpty() -> { |
|||
Log.i(TAG, "翻译成功，触发合成: ${res.translatedText!!.take(120)}") |
|||
|
|||
val dst = extractDstFromTranslated(res.translatedText!!) |
|||
if (dst != null && dst.isNotEmpty()) { |
|||
Log.i(TAG, "使用 dst 进行合成: ${dst.take(120)}") |
|||
eventCallback?.onTranslated(utteranceId, text, dst, tgtLang) |
|||
synthesizeText(utteranceId, dst) |
|||
} else { |
|||
Log.w(TAG, "未能解析 dst，回退使用原始翻译文本进行合成") |
|||
synthesizeText(utteranceId, res.translatedText!!) |
|||
} |
|||
} |
|||
res?.error != null -> eventCallback?.onTranslationFailed(utteranceId, text, res.error!!) |
|||
else -> eventCallback?.onTranslationFailed(utteranceId, text, "翻译服务返回空结果") |
|||
} |
|||
} catch (e: TimeoutCancellationException) { |
|||
serviceState.isTranslating.set(false) |
|||
eventCallback?.onStateChanged("Translation", false) |
|||
Log.w(TAG, "最终翻译超时") |
|||
eventCallback?.onTranslationFailed(utteranceId, text, "翻译超时") |
|||
} catch (e: Exception) { |
|||
serviceState.isTranslating.set(false) |
|||
eventCallback?.onStateChanged("Translation", false) |
|||
Log.e(TAG, "最终翻译异常: ${e.message}") |
|||
eventCallback?.onTranslationFailed(utteranceId, text, "翻译异常: ${e.message}") |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 解析翻译所用的源/目标语言 |
|||
* 若配置中提供 `translationSourceLanguage`/`translationTargetLanguage`，优先使用；否则回退到会话的 `sourceLanguage`/`targetLanguage` |
|||
*/ |
|||
private fun resolveTranslationLanguages(): Pair<String, String> { |
|||
val src = if (serviceConfig.translationSourceLanguage.isNotEmpty()) serviceConfig.translationSourceLanguage else serviceConfig.sourceLanguage |
|||
val tgt = if (serviceConfig.translationTargetLanguage.isNotEmpty()) serviceConfig.translationTargetLanguage else serviceConfig.targetLanguage |
|||
return Pair(src, tgt) |
|||
} |
|||
|
|||
/** |
|||
* 初始化成功后示例合成一句目标语言话语 |
|||
* 功能：将固定中文短句翻译为目标语言并进行语音合成，便于确认语音配置与管线是否正常 |
|||
*/ |
|||
fun speakDemoSentence() { |
|||
if (!serviceState.isInitialized.get()) { |
|||
return |
|||
} |
|||
|
|||
val srcLang = serviceConfig.sourceLanguage |
|||
val tgtLang = serviceConfig.targetLanguage |
|||
|
|||
try { |
|||
val sourceText = "初始化成功" |
|||
val res = runBlocking { translationService?.translateText(sourceText, srcLang, tgtLang) } |
|||
val textToSpeak = res?.translatedText ?: run { |
|||
when { |
|||
tgtLang.lowercase().startsWith("en") -> "Initialization successful" |
|||
tgtLang.lowercase().startsWith("zh") -> "初始化成功" |
|||
tgtLang.lowercase().startsWith("ja") -> "初期化が完了しました" |
|||
tgtLang.lowercase().startsWith("ko") -> "초기화가 완료되었습니다" |
|||
tgtLang.lowercase().startsWith("fr") -> "Initialisation réussie" |
|||
tgtLang.lowercase().startsWith("es") -> "Inicialización completada" |
|||
else -> "Initialization completed" |
|||
} |
|||
} |
|||
|
|||
val uttId = "utt-demo-${System.currentTimeMillis()}" |
|||
synthesizeText(uttId, textToSpeak) |
|||
} catch (e: Exception) { |
|||
Log.w(TAG, "演示合成失败: ${e.message}") |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 从翻译服务返回的 JSON 文本中提取 `trans_result.dst` |
|||
* |
|||
* 参数： |
|||
* - translatedJson: 翻译服务返回的字符串（形如 {"trans_result":{"dst":"..."}}） |
|||
* 返回： |
|||
* - 若解析成功返回 dst 字段，否则返回 null |
|||
*/ |
|||
private fun extractDstFromTranslated(translatedJson: String): String? { |
|||
return try { |
|||
val root = JSONObject(translatedJson) |
|||
val trans = root.optJSONObject("trans_result") |
|||
when { |
|||
trans != null -> trans.optString("dst", null) |
|||
else -> root.optString("dst", null) |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.w(TAG, "解析翻译结果JSON失败: ${e.message}") |
|||
null |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 合成文本到音频 |
|||
*/ |
|||
/** |
|||
* 触发语音合成（异步） |
|||
* |
|||
* 参数： |
|||
* - utteranceId: 当前句子的唯一标识 |
|||
* - text: 需要合成的目标文本 |
|||
* 行为： |
|||
* - 仅设置当前合成上下文并调用异步合成；事件回调在合成器监听中统一分发。 |
|||
*/ |
|||
private fun synthesizeText(utteranceId: String, text: String) { |
|||
try { |
|||
currentSynthesisId = utteranceId |
|||
currentSynthesisText = text |
|||
|
|||
val vcn = serviceConfig.currentVoice |
|||
val pitch = parsePercentToScale(serviceConfig.speechPitch) |
|||
val speed = parsePercentToScale(serviceConfig.speechRate) |
|||
|
|||
ttsClient?.synthesize( |
|||
appId = serviceConfig.xfyunAppId, |
|||
apiKey = serviceConfig.xfyunAccessKeyId, |
|||
apiSecret = serviceConfig.xfyunAccessKeySecret, |
|||
text = text, |
|||
vcn = vcn, |
|||
tte = "UTF8", |
|||
pitch = pitch, |
|||
speed = speed, |
|||
aue = "raw", |
|||
callback = object : IflytekTtsWs.Callback { |
|||
override fun onStarted() { |
|||
ttsdata = ByteArray(0) |
|||
serviceState.isSynthesizing.set(true) |
|||
val uttId = currentSynthesisId ?: "" |
|||
val t = currentSynthesisText ?: "" |
|||
Log.i(TAG, "TTS合成开始: ${uttId}") |
|||
eventCallback?.onSynthesisStarted(uttId, t) |
|||
eventCallback?.onStateChanged("Synthesis", true) |
|||
} |
|||
|
|||
override fun onAudioChunk(data: ByteArray) { |
|||
val uttId = currentSynthesisId ?: "" |
|||
val t = currentSynthesisText ?: "" |
|||
Log.i(TAG, "TTS合成数据: ${uttId} 数据长度=${data.size}") |
|||
if (data.isNotEmpty()) { |
|||
processSynthesisAudioChunk(uttId, t, data, 1280) |
|||
eventCallback?.onSynthesisProgress(uttId, t, 0.5f) |
|||
} |
|||
} |
|||
|
|||
override fun onCompleted(sid: String?) { |
|||
val uttId = currentSynthesisId ?: "" |
|||
val t = currentSynthesisText ?: "" |
|||
Log.i(TAG, "TTS合成完成: ${uttId} sid=${sid ?: ""}") |
|||
eventCallback?.onSynthesisCompleted(uttId, t) |
|||
// 发送剩余缓存 |
|||
if (ttsdata.isNotEmpty()) { |
|||
eventCallback?.onSynthesisAudioGenerated(uttId, t, ttsdata) |
|||
ttsdata = ByteArray(0) |
|||
} |
|||
currentSynthesisId = null |
|||
currentSynthesisText = null |
|||
serviceState.isSynthesizing.set(false) |
|||
eventCallback?.onStateChanged("Synthesis", false) |
|||
} |
|||
|
|||
override fun onError(code: Int?, message: String) { |
|||
val uttId = currentSynthesisId ?: "" |
|||
val t = currentSynthesisText ?: "" |
|||
Log.w(TAG, "TTS合成错误: code=${code ?: -1} msg=$message") |
|||
eventCallback?.onSynthesisFailed(uttId, t, "语音合成错误: ${message}") |
|||
currentSynthesisId = null |
|||
currentSynthesisText = null |
|||
serviceState.isSynthesizing.set(false) |
|||
eventCallback?.onStateChanged("Synthesis", false) |
|||
} |
|||
} |
|||
) |
|||
} catch (e: Exception) { |
|||
eventCallback?.onSynthesisFailed(utteranceId, text, "合成异常: ${e.message}") |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 生成基本 SSML |
|||
*/ |
|||
/** |
|||
* 生成优化的 SSML 文本 |
|||
* 清理与转义输入文本，并应用速率/音高/音量与当前语音 |
|||
*/ |
|||
private fun generateSsml(text: String): String { |
|||
val cleanText = cleanTextForSynthesis(text) |
|||
val escaped = escapeXmlText(cleanText) |
|||
return """ |
|||
<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xml:lang="${serviceConfig.targetLanguage}"> |
|||
<voice name="${serviceConfig.currentVoice}"> |
|||
<prosody rate="${serviceConfig.speechRate}" pitch="${serviceConfig.speechPitch}" volume="${serviceConfig.speechVolume}"> |
|||
$escaped |
|||
</prosody> |
|||
</voice> |
|||
</speak> |
|||
""".trimIndent() |
|||
} |
|||
|
|||
/** |
|||
* 清理文本用于语音合成 |
|||
* 去除 URL、表情符号并规范空白字符 |
|||
*/ |
|||
private fun cleanTextForSynthesis(text: String): String { |
|||
return text |
|||
.replace(Regex("https?://[^\\s]+"), "") |
|||
.replace(Regex("[\\p{So}\\p{Sk}]"), "") |
|||
.replace(Regex("\\s+"), " ") |
|||
.trim() |
|||
} |
|||
|
|||
/** |
|||
* 将文本进行 XML 转义 |
|||
*/ |
|||
private fun escapeXmlText(text: String): String { |
|||
return text |
|||
.replace("&", "&") |
|||
.replace("<", "<") |
|||
.replace(">", ">") |
|||
.replace("\"", """) |
|||
.replace("'", "'") |
|||
} |
|||
|
|||
/** |
|||
* 设置微软 TTS 合成器 |
|||
*/ |
|||
/** |
|||
* 设置微软 TTS 合成器（事件驱动） |
|||
* |
|||
* 行为： |
|||
* - 播放启用：输出到系统扬声器 |
|||
* - 播放禁用：创建推流输出以保持有效的 AudioConfig（丢弃数据），音频通过事件返回 |
|||
* - 监听合成开始/进行中/完成/取消事件，分发状态与音频数据 |
|||
*/ |
|||
private fun setupTtsClient() { |
|||
ttsClient?.close() |
|||
ttsClient = IflytekTtsWs() |
|||
} |
|||
|
|||
/** |
|||
* 将合成中的音频按 1280 字节对齐分片并回调 |
|||
* |
|||
* 参数： |
|||
* - utteranceId: 当前句子的唯一标识 |
|||
* - text: 合成文本内容 |
|||
* - incoming: 新收到的音频数据 |
|||
* - multiple: 分片对齐字节数（默认 1280） |
|||
* 行为： |
|||
* - 追加到缓存，仅当满足 multiple 的整数倍时发送；余量保留等待补齐。 |
|||
*/ |
|||
private fun processSynthesisAudioChunk( |
|||
utteranceId: String, |
|||
text: String, |
|||
incoming: ByteArray, |
|||
multiple: Int = 1280 |
|||
) { |
|||
if (incoming.isNotEmpty()) { |
|||
ttsdata = ttsdata + incoming |
|||
} |
|||
val sendLen = (ttsdata.size / multiple) * multiple |
|||
if (sendLen > 0) { |
|||
val chunk = ttsdata.copyOfRange(0, sendLen) |
|||
Log.d(TAG, "TTS分片回调: 发送 ${sendLen} 字节，剩余 ${ttsdata.size - sendLen} 字节") |
|||
eventCallback?.onSynthesisAudioGenerated(utteranceId, text, chunk) |
|||
ttsdata = ttsdata.copyOfRange(sendLen, ttsdata.size) |
|||
} |
|||
} |
|||
|
|||
private fun parsePercentToScale(percent: String): Int { |
|||
val v = percent.trim().removeSuffix("%") |
|||
val num = v.toIntOrNull() ?: 0 |
|||
val scaled = 50 + num // 0% -> 50, -50% -> 0, +50% -> 100 |
|||
return scaled.coerceIn(0, 100) |
|||
} |
|||
|
|||
/** |
|||
* 启动音频线程（保存/中转用途） |
|||
*/ |
|||
private fun startAudioThread() { |
|||
if (isAudioThreadRunning.get()) return |
|||
isAudioThreadRunning.set(true) |
|||
audioThread = Thread { |
|||
while (isAudioThreadRunning.get()) { |
|||
try { |
|||
val data = audioQueue.poll(100, java.util.concurrent.TimeUnit.MILLISECONDS) |
|||
if (data != null) { |
|||
// 此处可扩展：保存到文件或做能量检测 |
|||
// Log.v(TAG, "音频队列取出: ${data.size} 字节") |
|||
} |
|||
} catch (_: InterruptedException) { |
|||
Thread.currentThread().interrupt() |
|||
break |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "音频线程异常", e) |
|||
} |
|||
} |
|||
}.apply { name = "IflytekAudioThread"; Log.i(TAG, "启动音频线程"); start() } |
|||
} |
|||
|
|||
/** |
|||
* 释放资源 |
|||
*/ |
|||
fun dispose() { |
|||
try { |
|||
Log.i(TAG, "释放整合服务资源") |
|||
isAudioThreadRunning.set(false) |
|||
audioThread?.interrupt() |
|||
audioThread = null |
|||
|
|||
recognizingTranslationJob?.cancel() |
|||
|
|||
asrHelper?.stop() |
|||
asrHelper = null |
|||
|
|||
ttsClient?.close() |
|||
ttsClient = null |
|||
|
|||
serviceState.isInitialized.set(false) |
|||
serviceState.isRecognizing.set(false) |
|||
serviceState.isTranslating.set(false) |
|||
serviceState.isSynthesizing.set(false) |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "释放资源失败", e) |
|||
} |
|||
} |
|||
} |
|||
@ -0,0 +1,290 @@ |
|||
package com.yunqiinnovation.azure_speech |
|||
|
|||
import android.util.Log |
|||
import kotlinx.coroutines.Dispatchers |
|||
import kotlinx.coroutines.withContext |
|||
import okhttp3.MediaType.Companion.toMediaType |
|||
import okhttp3.OkHttpClient |
|||
import okhttp3.Request |
|||
import okhttp3.RequestBody |
|||
import okhttp3.RequestBody.Companion.toRequestBody |
|||
import org.json.JSONObject |
|||
import java.text.SimpleDateFormat |
|||
import java.util.Date |
|||
import java.util.Locale |
|||
import java.util.TimeZone |
|||
import java.util.concurrent.TimeUnit |
|||
import javax.crypto.Mac |
|||
import javax.crypto.spec.SecretKeySpec |
|||
import android.util.Base64 |
|||
|
|||
/** |
|||
* 讯飞机器翻译 OTS 服务实现 |
|||
* 兼容 IntegratedSpeechTranslationService 的 TranslationServiceInterface |
|||
*/ |
|||
class IflytekTranslationServiceImpl : |
|||
IntegratedSpeechTranslationService.TranslationServiceInterface { |
|||
|
|||
companion object { |
|||
private const val TAG = "IflytekTranslationService" |
|||
private const val DEFAULT_HOST = "ntrans.xfyun.cn" |
|||
private const val REQUEST_URI = "/v2/ots" |
|||
private const val HTTP_METHOD = "POST" |
|||
private const val HTTP_PROTO = "HTTP/1.1" |
|||
private const val ALG = "hmac-sha256" |
|||
} |
|||
|
|||
private var isInitialized = false |
|||
private var appId: String = "" |
|||
private var apiKey: String = "" |
|||
private var apiSecret: String = "" |
|||
private var host: String = DEFAULT_HOST |
|||
private var timeoutMs: Long = 8000 |
|||
|
|||
private val httpClient = OkHttpClient.Builder() |
|||
.connectTimeout(10, TimeUnit.SECONDS) |
|||
.readTimeout(30, TimeUnit.SECONDS) |
|||
.writeTimeout(30, TimeUnit.SECONDS) |
|||
.build() |
|||
|
|||
/** |
|||
* 初始化服务配置 |
|||
*/ |
|||
override suspend fun initialize(config: Map<String, String>): Boolean { |
|||
return try { |
|||
appId = (config["appId"]?.trim() ?: "") |
|||
apiKey = (config["apiKey"]?.trim() ?: "") |
|||
apiSecret = (config["apiSecret"]?.trim() ?: "") |
|||
val cfgHost = config["host"]?.trim() |
|||
host = if (cfgHost.isNullOrEmpty()) DEFAULT_HOST else cfgHost |
|||
timeoutMs = config["timeout"]?.toLongOrNull() ?: 8000 |
|||
Log.d(TAG, "初始化配置: appId=" + appId + ", apiKey=" + apiKey + ", apiSecret=" + apiSecret) |
|||
if (appId.isEmpty() || apiKey.isEmpty() || apiSecret.isEmpty()) { |
|||
Log.e(TAG, "缺少必要的鉴权参数") |
|||
false |
|||
} else { |
|||
isInitialized = true |
|||
Log.d(TAG, "讯飞 OTS 翻译服务初始化成功") |
|||
Log.d(TAG, "初始化配置: host=" + host + ", timeoutMs=" + timeoutMs) |
|||
true |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "初始化失败", e) |
|||
false |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 文本翻译 |
|||
*/ |
|||
override suspend fun translateText( |
|||
text: String, |
|||
sourceLanguage: String, |
|||
targetLanguage: String |
|||
): IntegratedSpeechTranslationService.TranslationResult { |
|||
if (!isInitialized) { |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "翻译服务未初始化" |
|||
) |
|||
} |
|||
|
|||
return withContext(Dispatchers.IO) { |
|||
try { |
|||
Log.d(TAG, "开始翻译: src=" + sourceLanguage + ", dst=" + targetLanguage + ", 文本长度=" + text.length) |
|||
|
|||
if (host.isBlank()) { |
|||
Log.w(TAG, "host为空，使用默认host: " + DEFAULT_HOST) |
|||
host = DEFAULT_HOST |
|||
} |
|||
|
|||
|
|||
val bodyJson = buildBody(text, sourceLanguage, targetLanguage) |
|||
val digest = buildDigest(bodyJson) |
|||
val date = httpDate(Date()) |
|||
val authorization = buildAuthorization(digest, date) |
|||
val headers = buildHeaders(digest, authorization, date) |
|||
Log.d(TAG, "请求体长度=" + bodyJson.length) |
|||
Log.d(TAG, "Digest长度=" + digest.length) |
|||
|
|||
val mediaType = "application/json".toMediaType() |
|||
val reqBody: RequestBody = bodyJson.toRequestBody(mediaType) |
|||
val url = "https://$host$REQUEST_URI" |
|||
Log.d(TAG, "发起请求: url=" + url) |
|||
|
|||
val request = Request.Builder() |
|||
.url(url) |
|||
.post(reqBody) |
|||
.apply { |
|||
headers.forEach { (k, v) -> addHeader(k, v) } |
|||
} |
|||
.build() |
|||
|
|||
val startNanos = System.nanoTime() |
|||
val response = httpClient.newCall(request).execute() |
|||
val durationMs = java.util.concurrent.TimeUnit.NANOSECONDS.toMillis(System.nanoTime() - startNanos) |
|||
val code = response.code |
|||
val respStr = response.body?.string() |
|||
Log.d(TAG, "HTTP响应: code=" + code + ", bodyLen=" + (respStr?.length ?: -1) + ", 耗时=" + durationMs + "ms") |
|||
if (respStr == null) { |
|||
Log.w(TAG, "HTTP错误: code=" + code + ", 空响应体") |
|||
return@withContext IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "HTTP错误: $code" |
|||
) |
|||
} |
|||
|
|||
if (code != 200) { |
|||
val snippet = if (respStr.length > 256) respStr.substring(0, 256) else respStr |
|||
Log.w(TAG, "HTTP错误: code=" + code + ", bodySnippet=" + snippet) |
|||
try { |
|||
val errJson = JSONObject(respStr) |
|||
val msg = errJson.optString("message") |
|||
val c = errJson.optInt("code", code) |
|||
val friendly = interpretError(c, msg) |
|||
return@withContext IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = friendly |
|||
) |
|||
} catch (_: Exception) { |
|||
return@withContext IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "HTTP错误: $code" |
|||
) |
|||
} |
|||
} |
|||
|
|||
val json = JSONObject(respStr) |
|||
val errCode = json.optInt("code", -1) |
|||
Log.d(TAG, "服务返回码: " + errCode) |
|||
if (errCode == 0) { |
|||
val data = json.optJSONObject("data") |
|||
val result = when { |
|||
data == null -> null |
|||
data.has("result") -> data.optString("result") |
|||
data.has("translation") -> data.optString("translation") |
|||
data.has("trans_result") -> data.optString("trans_result") |
|||
else -> null |
|||
} |
|||
if (!result.isNullOrEmpty()) { |
|||
Log.d(TAG, "翻译成功: 结果长度=" + result.length) |
|||
return@withContext IntegratedSpeechTranslationService.TranslationResult( |
|||
success = true, |
|||
translatedText = result, |
|||
confidence = 0.9f |
|||
) |
|||
} |
|||
} |
|||
|
|||
val message = json.optString("message", "翻译失败") |
|||
val friendly = interpretError(errCode, message) |
|||
Log.w(TAG, "翻译失败: message=" + message) |
|||
return@withContext IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = friendly |
|||
) |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "翻译请求异常", e) |
|||
IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "请求异常: ${e.message}" |
|||
) |
|||
} |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 释放资源 |
|||
*/ |
|||
override fun dispose() { |
|||
isInitialized = false |
|||
try { |
|||
httpClient.dispatcher.executorService.shutdown() |
|||
httpClient.connectionPool.evictAll() |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "清理资源失败", e) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 构建请求体 JSON 字符串 |
|||
*/ |
|||
private fun buildBody(text: String, from: String, to: String): String { |
|||
val contentB64 = Base64.encodeToString(text.toByteArray(Charsets.UTF_8), Base64.NO_WRAP) |
|||
Log.v(TAG, "构建请求体: from=" + from + ", to=" + to + ", textB64Len=" + contentB64.length) |
|||
val root = JSONObject() |
|||
root.put("common", JSONObject().put("app_id", appId)) |
|||
root.put("business", JSONObject().put("from", from).put("to", to)) |
|||
root.put("data", JSONObject().put("text", contentB64)) |
|||
return root.toString() |
|||
} |
|||
|
|||
/** |
|||
* 构建 Digest 头（RFC 标准） |
|||
*/ |
|||
private fun buildDigest(bodyJson: String): String { |
|||
val md = java.security.MessageDigest.getInstance("SHA-256") |
|||
val hash = md.digest(bodyJson.toByteArray(Charsets.UTF_8)) |
|||
val b64 = Base64.encodeToString(hash, Base64.NO_WRAP) |
|||
return "SHA-256=$b64" |
|||
} |
|||
|
|||
/** |
|||
* 构建 Authorization 头 |
|||
*/ |
|||
private fun buildAuthorization(digest: String, date: String): String { |
|||
val signStr = buildSignString(host, date, digest) |
|||
val mac = Mac.getInstance("HmacSHA256") |
|||
mac.init(SecretKeySpec(apiSecret.toByteArray(Charsets.UTF_8), "HmacSHA256")) |
|||
val signatureB64 = Base64.encodeToString(mac.doFinal(signStr.toByteArray(Charsets.UTF_8)), Base64.NO_WRAP) |
|||
Log.d(TAG, "构建鉴权: date=" + date + ", 签名长度=" + signatureB64.length) |
|||
return "api_key=\"$apiKey\", algorithm=\"$ALG\", headers=\"host date request-line digest\", signature=\"$signatureB64\"" |
|||
} |
|||
|
|||
/** |
|||
* 构建请求头映射 |
|||
*/ |
|||
private fun buildHeaders(digest: String, authorization: String, date: String): Map<String, String> { |
|||
Log.d(TAG, "构建请求头: host=" + host + ", date=" + date) |
|||
return mapOf( |
|||
"Content-Type" to "application/json", |
|||
"Accept" to "application/json", |
|||
"Method" to HTTP_METHOD, |
|||
"Host" to host, |
|||
"Date" to date, |
|||
"Digest" to digest, |
|||
"Authorization" to authorization |
|||
) |
|||
} |
|||
|
|||
/** |
|||
* 统一错误码解释 |
|||
*/ |
|||
private fun interpretError(code: Int, message: String?): String { |
|||
return when (code) { |
|||
11200 -> "鉴权失败(11200): 请检查 APPID/APIKey/APISecret 是否正确，服务是否开通" |
|||
11201 -> "签名错误(11201): 请检查 Authorization/Date/Host/Digest 的签名串" |
|||
10105 -> "参数错误(10105): 请检查业务参数 from/to 与请求体" |
|||
else -> if (!message.isNullOrEmpty()) "HTTP错误: $code, $message" else "HTTP错误: $code" |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 生成签名原文字段 |
|||
*/ |
|||
private fun buildSignString(host: String, date: String, digest: String): String { |
|||
return "host: $host\n" + |
|||
"date: $date\n" + |
|||
"$HTTP_METHOD $REQUEST_URI $HTTP_PROTO\n" + |
|||
"digest: $digest" |
|||
} |
|||
|
|||
/** |
|||
* 生成 RFC 1123 GMT 时间字符串 |
|||
*/ |
|||
private fun httpDate(dt: Date): String { |
|||
val sdf = SimpleDateFormat("EEE, dd MMM yyyy HH:mm:ss 'GMT'", Locale.US) |
|||
sdf.timeZone = TimeZone.getTimeZone("GMT") |
|||
return sdf.format(dt) |
|||
} |
|||
} |
|||
@ -0,0 +1,271 @@ |
|||
package com.yunqiinnovation.azure_speech |
|||
|
|||
import android.util.Log |
|||
import okhttp3.* |
|||
import okhttp3.HttpUrl.Companion.toHttpUrl |
|||
import okio.ByteString |
|||
import java.net.URL |
|||
import java.nio.charset.StandardCharsets |
|||
import java.text.SimpleDateFormat |
|||
import java.util.Base64 |
|||
import java.util.Locale |
|||
import java.util.TimeZone |
|||
import javax.crypto.Mac |
|||
import javax.crypto.spec.SecretKeySpec |
|||
|
|||
/** |
|||
* 讯飞在线TTS WebSocket客户端 |
|||
* |
|||
* 负责生成鉴权URL并通过WebSocket进行流式语音合成，返回PCM原始数据 |
|||
*/ |
|||
class IflytekTtsWs( |
|||
private val hostUrl: String = "https://tts-api.xfyun.cn/v2/tts" |
|||
) { |
|||
private val tag = "IflytekTtsWs" |
|||
private val okHttp = OkHttpClient() |
|||
|
|||
private var webSocket: WebSocket? = null |
|||
private var isClosed = false |
|||
|
|||
/** |
|||
* 合成回调接口 |
|||
*/ |
|||
interface Callback { |
|||
fun onStarted() |
|||
fun onAudioChunk(data: ByteArray) |
|||
fun onCompleted(sid: String?) |
|||
fun onError(code: Int?, message: String) |
|||
} |
|||
|
|||
/** |
|||
* 进行一次文本合成 |
|||
* |
|||
* 参数说明： |
|||
* - appId: 讯飞应用 ID |
|||
* - apiKey: API Key |
|||
* - apiSecret: API Secret |
|||
* - text: 待合成文本（将按 `tte` 编码） |
|||
* - vcn: 发音人标识（如 `x4_yezi`），直接覆盖服务默认值 |
|||
* - tte: 文本编码（`UTF8`/`UNICODE`），默认 `UTF8` |
|||
* - pitch: 音高 0–100，默认 50 |
|||
* - speed: 语速 0–100，默认 50 |
|||
* - aue: 音频编码，`raw` 表示 PCM 原始流 |
|||
* - callback: 合成回调，流式返回音频分片并在完成时回调 |
|||
*/ |
|||
fun synthesize( |
|||
appId: String, |
|||
apiKey: String, |
|||
apiSecret: String, |
|||
text: String, |
|||
vcn: String = "x4_yezi", |
|||
tte: String = "UTF8", |
|||
pitch: Int = 50, |
|||
speed: Int = 50, |
|||
aue: String = "raw", |
|||
callback: Callback |
|||
) { |
|||
try { |
|||
val authUrl = buildAuthUrl(hostUrl, apiKey, apiSecret).replace("https://", "wss://") |
|||
val request = Request.Builder().url(authUrl).build() |
|||
|
|||
isClosed = false |
|||
webSocket = okHttp.newWebSocket(request, object : WebSocketListener() { |
|||
override fun onOpen(ws: WebSocket, response: Response) { |
|||
callback.onStarted() |
|||
val payload = buildRequestJson(appId, text, vcn, tte, pitch, speed, aue) |
|||
ws.send(payload) |
|||
} |
|||
|
|||
override fun onMessage(ws: WebSocket, textMsg: String) { |
|||
try { |
|||
val parse = JsonParse.fromJson(textMsg) |
|||
if (parse.code != 0) { |
|||
callback.onError(parse.code, "sid=${parse.sid}") |
|||
return |
|||
} |
|||
val data = parse.data |
|||
if (data != null) { |
|||
if (!data.audio.isNullOrEmpty()) { |
|||
val bytes = Base64.getDecoder().decode(data.audio) |
|||
callback.onAudioChunk(bytes) |
|||
} |
|||
if (data.status == 2) { |
|||
callback.onCompleted(parse.sid) |
|||
close() |
|||
} |
|||
} |
|||
} catch (e: Exception) { |
|||
callback.onError(null, "parse error: ${e.message}") |
|||
} |
|||
} |
|||
|
|||
override fun onMessage(ws: WebSocket, bytes: ByteString) { |
|||
// iFlytek服务使用文本消息承载；此分支通常不会触发 |
|||
} |
|||
|
|||
override fun onFailure(ws: WebSocket, t: Throwable, response: Response?) { |
|||
callback.onError(null, t.message ?: "unknown error") |
|||
close() |
|||
} |
|||
|
|||
override fun onClosing(ws: WebSocket, code: Int, reason: String) { |
|||
ws.close(code, reason) |
|||
} |
|||
|
|||
override fun onClosed(ws: WebSocket, code: Int, reason: String) { |
|||
isClosed = true |
|||
} |
|||
}) |
|||
} catch (e: Exception) { |
|||
callback.onError(null, e.message ?: "exception") |
|||
close() |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 关闭当前连接 |
|||
*/ |
|||
fun close() { |
|||
try { |
|||
if (!isClosed) { |
|||
webSocket?.close(1000, "normal") |
|||
} |
|||
} catch (_: Exception) { |
|||
} finally { |
|||
webSocket = null |
|||
isClosed = true |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 按语言进行文本合成 |
|||
* |
|||
* 用法说明： |
|||
* - iFlytek TTS 的语种由发音人 `vcn` 决定 |
|||
* - 可传入 `language`（如 `ja-JP`、`ko-KR`），并通过 `vcnOverride` 指定发音人 |
|||
* - 若未提供覆盖值，将尝试按语言解析默认发音人；未配置则抛出错误 |
|||
*/ |
|||
fun synthesizeByLanguage( |
|||
appId: String, |
|||
apiKey: String, |
|||
apiSecret: String, |
|||
text: String, |
|||
language: String, |
|||
callback: Callback, |
|||
vcnOverride: String? = null, |
|||
tte: String = "UTF8", |
|||
pitch: Int = 50, |
|||
speed: Int = 50, |
|||
aue: String = "raw" |
|||
) { |
|||
val vcn = resolveVcn(language, vcnOverride) |
|||
synthesize(appId, apiKey, apiSecret, text, vcn, tte, pitch, speed, aue, callback) |
|||
} |
|||
|
|||
/** |
|||
* 解析语言对应的默认发音人 |
|||
*/ |
|||
private fun resolveVcn(language: String, vcnOverride: String?): String { |
|||
if (!vcnOverride.isNullOrBlank()) return vcnOverride |
|||
return when (language.lowercase(Locale.ROOT)) { |
|||
"zh", "zh-cn" -> "x4_yezi" |
|||
// 以下语种需在控制台开启对应小语种发音人，并传入具体发音人ID |
|||
"ja", "ja-jp" -> throw IllegalArgumentException("未配置日语发音人vcn，请传入vcnOverride") |
|||
"ko", "ko-kr" -> throw IllegalArgumentException("未配置韩语发音人vcn，请传入vcnOverride") |
|||
else -> throw IllegalArgumentException("未配置语种(${language})发音人vcn，请传入vcnOverride") |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 构造业务请求JSON |
|||
*/ |
|||
private fun buildRequestJson( |
|||
appId: String, |
|||
text: String, |
|||
vcn: String, |
|||
tte: String, |
|||
pitch: Int, |
|||
speed: Int, |
|||
aue: String |
|||
): String { |
|||
val txt = when (tte.uppercase(Locale.ROOT)) { |
|||
"UNICODE" -> Base64.getEncoder().encodeToString(text.toByteArray(Charsets.UTF_16LE)) |
|||
else -> Base64.getEncoder().encodeToString(text.toByteArray(StandardCharsets.UTF_8)) |
|||
} |
|||
return """ |
|||
{ |
|||
"common": { "app_id": "$appId" }, |
|||
"business": { |
|||
"aue": "$aue", |
|||
"tte": "$tte", |
|||
"ent": "intp65", |
|||
"vcn": "$vcn", |
|||
"pitch": $pitch, |
|||
"speed": $speed |
|||
}, |
|||
"data": { |
|||
"status": 2, |
|||
"text": "$txt" |
|||
} |
|||
} |
|||
""".trimIndent() |
|||
} |
|||
|
|||
/** |
|||
* 生成带鉴权的URL |
|||
*/ |
|||
private fun buildAuthUrl(hostUrl: String, apiKey: String, apiSecret: String): String { |
|||
val url = URL(hostUrl) |
|||
val fmt = SimpleDateFormat("EEE, dd MMM yyyy HH:mm:ss z", Locale.US).apply { |
|||
timeZone = TimeZone.getTimeZone("GMT") |
|||
} |
|||
val date = fmt.format(java.util.Date()) |
|||
val preStr = "host: ${url.host}\n" + |
|||
"date: $date\n" + |
|||
"GET ${url.path} HTTP/1.1" |
|||
|
|||
val mac = Mac.getInstance("hmacsha256") |
|||
val spec = SecretKeySpec(apiSecret.toByteArray(StandardCharsets.UTF_8), "hmacsha256") |
|||
mac.init(spec) |
|||
val hex = mac.doFinal(preStr.toByteArray(StandardCharsets.UTF_8)) |
|||
val sha = Base64.getEncoder().encodeToString(hex) |
|||
val authorization = String.format( |
|||
"api_key=\"%s\", algorithm=\"%s\", headers=\"%s\", signature=\"%s\"", |
|||
apiKey, "hmac-sha256", "host date request-line", sha |
|||
) |
|||
|
|||
val httpUrl = "https://${url.host}${url.path}".toHttpUrl().newBuilder() |
|||
.addQueryParameter("authorization", Base64.getEncoder().encodeToString(authorization.toByteArray(StandardCharsets.UTF_8))) |
|||
.addQueryParameter("date", date) |
|||
.addQueryParameter("host", url.host) |
|||
.build() |
|||
return httpUrl.toString() |
|||
} |
|||
|
|||
/** |
|||
* 解析服务端JSON |
|||
*/ |
|||
private data class JsonParse(val code: Int, val sid: String?, val data: Data?) { |
|||
data class Data(val status: Int, val audio: String?) |
|||
|
|||
companion object { |
|||
fun fromJson(s: String): JsonParse { |
|||
// 简单解析以减少依赖：仅提取必须字段 |
|||
// 格式可能为 {"code":0, "sid":"...", "data":{"status":1,"audio":"..."}} |
|||
var code = -1 |
|||
var sid: String? = null |
|||
var status = -1 |
|||
var audio: String? = null |
|||
|
|||
fun extract(re: Regex): String? = re.find(s)?.groupValues?.getOrNull(1) |
|||
|
|||
extract(Regex("\"code\"\\s*:\\s*(\\d+)"))?.let { code = it.toIntOrNull() ?: -1 } |
|||
sid = extract(Regex("\"sid\"\\s*:\\s*\"([^\"]+)\"")) |
|||
extract(Regex("\"status\"\\s*:\\s*(\\d+)"))?.let { status = it.toIntOrNull() ?: -1 } |
|||
audio = extract(Regex("\"audio\"\\s*:\\s*\"([^\"]*)\"")) |
|||
|
|||
return JsonParse(code, sid, if (status >= 0) Data(status, audio) else null) |
|||
} |
|||
} |
|||
} |
|||
} |
|||
@ -0,0 +1,250 @@ |
|||
package com.yunqiinnovation.azure_speech |
|||
|
|||
import android.util.Log |
|||
import com.microsoft.cognitiveservices.speech.* |
|||
import com.microsoft.cognitiveservices.speech.audio.AudioConfig |
|||
import java.util.concurrent.atomic.AtomicBoolean |
|||
|
|||
/** |
|||
* Microsoft Azure ASR (Speech-to-Text) 服务实现 |
|||
*/ |
|||
class MicrosoftAsrServiceImpl { |
|||
|
|||
companion object { |
|||
private const val TAG = "MicrosoftAsrService" |
|||
} |
|||
|
|||
private var speechConfig: SpeechConfig? = null |
|||
private var recognizer: SpeechRecognizer? = null |
|||
private var audioConfig: AudioConfig? = null |
|||
private var audioProcessor: AudioProcessor? = null |
|||
|
|||
private var isRecognizing = AtomicBoolean(false) |
|||
private var callback: AsrCallback? = null |
|||
|
|||
/** |
|||
* ASR 回调接口 |
|||
*/ |
|||
interface AsrCallback { |
|||
fun onRecognizing(text: String, confidence: Float) |
|||
fun onRecognized(text: String, confidence: Float) |
|||
fun onSessionStarted() |
|||
fun onSessionStopped() |
|||
fun onCanceled(error: String) |
|||
fun onError(message: String) |
|||
} |
|||
|
|||
/** |
|||
* ASR 配置参数 |
|||
*/ |
|||
data class AsrConfig( |
|||
val subscriptionKey: String, |
|||
val region: String, |
|||
val language: String, |
|||
val enableAutoLanguageDetection: Boolean = false |
|||
) |
|||
|
|||
/** |
|||
* 初始化 ASR 服务 |
|||
*/ |
|||
fun initialize(config: AsrConfig, callback: AsrCallback): Boolean { |
|||
this.callback = callback |
|||
return try { |
|||
Log.d(TAG, "初始化 Azure ASR 服务: region=${config.region}, lang=${config.language}") |
|||
|
|||
// 1. 初始化 SpeechConfig |
|||
speechConfig = SpeechConfig.fromSubscription(config.subscriptionKey, config.region).apply { |
|||
speechRecognitionLanguage = config.language |
|||
if (config.enableAutoLanguageDetection) { |
|||
setProperty("SpeechServiceConnection_LanguageIdMode", "Continuous") |
|||
} |
|||
} |
|||
|
|||
// 2. 初始化 AudioProcessor (录音处理) |
|||
initializeAudioProcessor() |
|||
|
|||
// 3. 初始化 SpeechRecognizer |
|||
setupSpeechRecognizer(config.language) |
|||
|
|||
Log.d(TAG, "Azure ASR 服务初始化成功") |
|||
true |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "Azure ASR 服务初始化失败", e) |
|||
callback.onError("初始化失败: ${e.message}") |
|||
false |
|||
} |
|||
} |
|||
|
|||
private fun initializeAudioProcessor() { |
|||
audioProcessor = AudioProcessor().apply { |
|||
initialize() |
|||
} |
|||
// 使用 AudioProcessor 的流作为输入 |
|||
audioConfig = AudioConfig.fromStreamInput(audioProcessor?.getAudioStream()) |
|||
} |
|||
|
|||
private fun setupSpeechRecognizer(language: String) { |
|||
recognizer?.close() |
|||
recognizer = SpeechRecognizer(speechConfig, audioConfig).apply { |
|||
|
|||
// 识别中 |
|||
recognizing.addEventListener { _, event -> |
|||
if (event.result.text.isNotEmpty()) { |
|||
val confidence = extractConfidence(event.result) |
|||
callback?.onRecognizing(event.result.text, confidence) |
|||
} |
|||
} |
|||
|
|||
// 识别完成 |
|||
recognized.addEventListener { _, event -> |
|||
if (event.result.reason == ResultReason.RecognizedSpeech) { |
|||
if (event.result.text.isNotEmpty()) { |
|||
val confidence = extractConfidence(event.result) |
|||
callback?.onRecognized(event.result.text, confidence) |
|||
} |
|||
} else if (event.result.reason == ResultReason.NoMatch) { |
|||
Log.d(TAG, "未识别到语音") |
|||
} |
|||
} |
|||
|
|||
// 会话开始 |
|||
sessionStarted.addEventListener { _, _ -> |
|||
Log.d(TAG, "会话开始") |
|||
isRecognizing.set(true) |
|||
callback?.onSessionStarted() |
|||
} |
|||
|
|||
// 会话停止 |
|||
sessionStopped.addEventListener { _, _ -> |
|||
Log.d(TAG, "会话停止") |
|||
isRecognizing.set(false) |
|||
callback?.onSessionStopped() |
|||
} |
|||
|
|||
// 取消 |
|||
canceled.addEventListener { _, event -> |
|||
Log.d(TAG, "识别取消: ${event.errorDetails}") |
|||
isRecognizing.set(false) |
|||
callback?.onCanceled(event.errorDetails ?: "未知错误") |
|||
} |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 启动连续识别 |
|||
*/ |
|||
fun startContinuousRecognition(): Boolean { |
|||
if (isRecognizing.get()) { |
|||
Log.w(TAG, "识别已在进行中") |
|||
return true |
|||
} |
|||
return try { |
|||
audioProcessor?.startRecording() |
|||
recognizer?.startContinuousRecognitionAsync() |
|||
Log.d(TAG, "启动连续识别成功") |
|||
true |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "启动连续识别失败", e) |
|||
callback?.onError("启动识别失败: ${e.message}") |
|||
false |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 停止连续识别 |
|||
*/ |
|||
fun stopContinuousRecognition() { |
|||
Log.d(TAG, "停止连续识别") |
|||
try { |
|||
recognizer?.stopContinuousRecognitionAsync() |
|||
audioProcessor?.stopRecording() |
|||
isRecognizing.set(false) |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "停止识别失败", e) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 推送音频数据 |
|||
*/ |
|||
fun pushAudioData(data: ByteArray) { |
|||
audioProcessor?.pushAudioData(data) |
|||
} |
|||
|
|||
/** |
|||
* 释放资源 |
|||
*/ |
|||
fun dispose() { |
|||
try { |
|||
stopContinuousRecognition() |
|||
recognizer?.close() |
|||
speechConfig?.close() |
|||
audioConfig?.close() // AudioConfig close 可能会抛异常如果流已经关闭，视 SDK 版本而定，加 try-catch |
|||
recognizer = null |
|||
speechConfig = null |
|||
audioConfig = null |
|||
audioProcessor = null |
|||
callback = null |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "释放资源出错", e) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 提取置信度 (占位实现，与原逻辑一致) |
|||
*/ |
|||
private fun extractConfidence(result: SpeechRecognitionResult): Float { |
|||
return try { |
|||
// 这里可以添加具体的提取逻辑 |
|||
0.8f |
|||
} catch (e: Exception) { |
|||
0.5f |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 开启录音文件保存 (如果需要直接在这里控制) |
|||
* 原逻辑中 IntegratedSpeechTranslationService 控制了 RecordFile， |
|||
* 这里可以提供接口让外部获取 AudioProcessor 或直接传入 RecordFile， |
|||
* 但为了简化，暂时保持 AudioProcessor 内部管理或外部通过其他方式处理。 |
|||
* 根据原代码，RecordFile 似乎是独立的。 |
|||
*/ |
|||
} |
|||
|
|||
/** |
|||
* 音频处理器 |
|||
* 封装 Azure PushAudioInputStream，用于接收外部音频数据并提供给 ASR 服务 |
|||
*/ |
|||
class AudioProcessor { |
|||
private var pushStream: com.microsoft.cognitiveservices.speech.audio.PushAudioInputStream? = null |
|||
|
|||
fun initialize() { |
|||
// 使用默认的 PCM 格式: 16kHz, 16bit, 单声道 |
|||
val format = com.microsoft.cognitiveservices.speech.audio.AudioStreamFormat.getWaveFormatPCM(16000L, 16, 1) |
|||
pushStream = com.microsoft.cognitiveservices.speech.audio.AudioInputStream.createPushStream(format) |
|||
} |
|||
|
|||
fun getAudioStream(): com.microsoft.cognitiveservices.speech.audio.AudioInputStream? { |
|||
return pushStream |
|||
} |
|||
|
|||
fun startRecording() { |
|||
if (pushStream == null) { |
|||
initialize() |
|||
} |
|||
} |
|||
|
|||
fun stopRecording() { |
|||
// 可以在这里做一些清理 |
|||
} |
|||
|
|||
fun pushAudioData(data: ByteArray) { |
|||
pushStream?.write(data) |
|||
} |
|||
|
|||
fun close() { |
|||
pushStream?.close() |
|||
pushStream = null |
|||
} |
|||
} |
|||
|
|||
@ -0,0 +1,266 @@ |
|||
package com.yunqiinnovation.azure_speech |
|||
|
|||
import android.util.Log |
|||
import com.microsoft.cognitiveservices.speech.* |
|||
import com.microsoft.cognitiveservices.speech.audio.AudioConfig |
|||
import com.microsoft.cognitiveservices.speech.audio.AudioOutputStream |
|||
import com.microsoft.cognitiveservices.speech.audio.PushAudioOutputStream |
|||
import com.microsoft.cognitiveservices.speech.audio.PushAudioOutputStreamCallback |
|||
import java.util.concurrent.atomic.AtomicBoolean |
|||
|
|||
/** |
|||
* Microsoft Azure TTS (Text-to-Speech) 服务实现 |
|||
*/ |
|||
class MicrosoftTTSServiceImpl { |
|||
|
|||
companion object { |
|||
private const val TAG = "MicrosoftTTSService" |
|||
} |
|||
|
|||
private var speechConfig: SpeechConfig? = null |
|||
private var synthesizer: SpeechSynthesizer? = null |
|||
private var synthOutputStream: PushAudioOutputStream? = null |
|||
|
|||
private var isSynthesizing = AtomicBoolean(false) |
|||
private var callback: TtsCallback? = null |
|||
|
|||
// 内部音频缓冲，用于分片发送 |
|||
private var ttsDataBuffer: ByteArray = ByteArray(0) |
|||
|
|||
/** |
|||
* TTS 回调接口 |
|||
*/ |
|||
interface TtsCallback { |
|||
fun onSynthesisStarted(utteranceId: String, text: String) |
|||
fun onSynthesisProgress(utteranceId: String, text: String, progress: Float) |
|||
// 返回处理后的音频分片 |
|||
fun onAudioChunkGenerated(utteranceId: String, text: String, audioChunk: ByteArray) |
|||
fun onSynthesisCompleted(utteranceId: String, text: String, fullAudio: ByteArray?) |
|||
fun onSynthesisFailed(utteranceId: String, text: String, error: String) |
|||
fun onError(message: String) |
|||
} |
|||
|
|||
/** |
|||
* TTS 配置参数 |
|||
*/ |
|||
data class TtsConfig( |
|||
val subscriptionKey: String, |
|||
val region: String, |
|||
val targetLanguage: String, |
|||
val voiceName: String, |
|||
val speechRate: String = "0%", |
|||
val speechPitch: String = "0%", |
|||
val speechVolume: String = "100%", |
|||
val enableAudioPlayback: Boolean = false |
|||
) |
|||
|
|||
private var currentConfig: TtsConfig? = null |
|||
|
|||
/** |
|||
* 初始化 TTS 服务 |
|||
*/ |
|||
fun initialize(config: TtsConfig, callback: TtsCallback): Boolean { |
|||
this.callback = callback |
|||
this.currentConfig = config |
|||
return try { |
|||
Log.d(TAG, "初始化 Azure TTS 服务: region=${config.region}, voice=${config.voiceName}") |
|||
|
|||
speechConfig = SpeechConfig.fromSubscription(config.subscriptionKey, config.region).apply { |
|||
speechSynthesisLanguage = config.targetLanguage |
|||
speechSynthesisVoiceName = config.voiceName.ifEmpty { config.targetLanguage } |
|||
setSpeechSynthesisOutputFormat(SpeechSynthesisOutputFormat.Riff16Khz16BitMonoPcm) |
|||
} |
|||
|
|||
setupSpeechSynthesizer(config) |
|||
|
|||
Log.d(TAG, "Azure TTS 服务初始化成功") |
|||
true |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "Azure TTS 服务初始化失败", e) |
|||
callback.onError("初始化失败: ${e.message}") |
|||
false |
|||
} |
|||
} |
|||
|
|||
private fun setupSpeechSynthesizer(config: TtsConfig) { |
|||
synthesizer?.close() |
|||
|
|||
// 根据配置决定音频输出方式 |
|||
val audioConfig = if (config.enableAudioPlayback) { |
|||
AudioConfig.fromDefaultSpeakerOutput() |
|||
} else { |
|||
// 创建空消费的推流，避免使用 null AudioConfig |
|||
synthOutputStream?.close() |
|||
synthOutputStream = AudioOutputStream.createPushStream(object : PushAudioOutputStreamCallback() { |
|||
override fun write(dataBuffer: ByteArray): Int { |
|||
// 丢弃数据但保持流有效 (因为我们会通过事件获取数据) |
|||
return dataBuffer.size |
|||
} |
|||
override fun close() { /* no-op */ } |
|||
}) |
|||
AudioConfig.fromStreamOutput(synthOutputStream) |
|||
} |
|||
|
|||
synthesizer = SpeechSynthesizer(speechConfig, audioConfig).apply { |
|||
|
|||
// 合成开始 |
|||
SynthesisStarted.addEventListener { _, event -> |
|||
Log.d(TAG, "合成开始") |
|||
ttsDataBuffer = ByteArray(0) // 重置缓冲 |
|||
isSynthesizing.set(true) |
|||
// 这里我们无法直接获取 utteranceId 和 text，因为事件对象里没有 |
|||
// 它们需要由调用 synthesizeText 的地方管理，或者通过某种方式传递 |
|||
// 为了简化，我们假设 Service 层不持有 currentUtteranceId， |
|||
// 而是由 orchestrator 传递，但事件监听器这里是异步的... |
|||
// *注意*: 原代码是使用了类成员 currentSynthesisId 来追踪。 |
|||
// 我们这里也需要类似机制，或者简化回调只通知状态，不带 ID（但这不符合需求）。 |
|||
// 只能使用成员变量来暂存当前的 ID。 |
|||
} |
|||
|
|||
// 合成进行中 |
|||
Synthesizing.addEventListener { _, event -> |
|||
val incoming = event.result.audioData ?: ByteArray(0) |
|||
// 这里的 currentUtteranceId 依赖于 synthesizeText 调用时的设置 |
|||
// 由于 SpeechSynthesizer 是单实例串行工作的，这通常是安全的 |
|||
processSynthesisAudioChunk(incoming) |
|||
|
|||
// 进度回调 (模拟) |
|||
callback?.onSynthesisProgress(currentUtteranceId, currentText, 0.5f) |
|||
} |
|||
|
|||
// 合成完成 |
|||
SynthesisCompleted.addEventListener { _, event -> |
|||
Log.d(TAG, "合成完成") |
|||
isSynthesizing.set(false) |
|||
callback?.onSynthesisCompleted(currentUtteranceId, currentText, event.result.audioData) |
|||
} |
|||
|
|||
// 合成取消 |
|||
SynthesisCanceled.addEventListener { _, event -> |
|||
Log.d(TAG, "合成取消: ${event.result.reason}") |
|||
isSynthesizing.set(false) |
|||
callback?.onSynthesisFailed(currentUtteranceId, currentText, "取消: ${event.result.reason}") |
|||
} |
|||
} |
|||
} |
|||
|
|||
// 用于追踪当前合成任务的状态 |
|||
private var currentUtteranceId: String = "" |
|||
private var currentText: String = "" |
|||
|
|||
/** |
|||
* 执行语音合成 |
|||
*/ |
|||
fun synthesize(utteranceId: String, text: String): ByteArray? { |
|||
if (isSynthesizing.get()) { |
|||
Log.w(TAG, "正在合成中，忽略请求: $utteranceId") |
|||
return null |
|||
} |
|||
|
|||
currentUtteranceId = utteranceId |
|||
currentText = text |
|||
|
|||
callback?.onSynthesisStarted(utteranceId, text) |
|||
|
|||
try { |
|||
isSynthesizing.set(true) |
|||
val ssml = generateOptimizedSsml(text) |
|||
|
|||
// 使用 SpeakSsmlAsync 并等待结果 |
|||
val result = synthesizer?.SpeakSsmlAsync(ssml)?.get() |
|||
|
|||
if (result?.reason == ResultReason.SynthesizingAudioCompleted) { |
|||
// 发送剩余缓冲数据 |
|||
flushAudioBuffer() |
|||
return result.audioData |
|||
} else { |
|||
return null |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "合成异常", e) |
|||
callback?.onSynthesisFailed(utteranceId, text, e.message ?: "未知异常") |
|||
return null |
|||
} finally { |
|||
isSynthesizing.set(false) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 停止/取消合成 |
|||
*/ |
|||
fun stop() { |
|||
try { |
|||
synthesizer?.StopSpeakingAsync() |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "停止合成失败", e) |
|||
} |
|||
} |
|||
|
|||
fun dispose() { |
|||
try { |
|||
stop() |
|||
synthesizer?.close() |
|||
speechConfig?.close() |
|||
synthOutputStream?.close() |
|||
synthesizer = null |
|||
speechConfig = null |
|||
synthOutputStream = null |
|||
callback = null |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "释放资源失败", e) |
|||
} |
|||
} |
|||
|
|||
private fun processSynthesisAudioChunk(incoming: ByteArray, multiple: Int = 1280) { |
|||
if (incoming.isNotEmpty()) { |
|||
ttsDataBuffer += incoming |
|||
} |
|||
|
|||
val sendLen = (ttsDataBuffer.size / multiple) * multiple |
|||
if (sendLen > 0) { |
|||
val chunk = ttsDataBuffer.copyOfRange(0, sendLen) |
|||
callback?.onAudioChunkGenerated(currentUtteranceId, currentText, chunk) |
|||
ttsDataBuffer = ttsDataBuffer.copyOfRange(sendLen, ttsDataBuffer.size) |
|||
} |
|||
} |
|||
|
|||
private fun flushAudioBuffer() { |
|||
if (ttsDataBuffer.isNotEmpty()) { |
|||
callback?.onAudioChunkGenerated(currentUtteranceId, currentText, ttsDataBuffer) |
|||
ttsDataBuffer = ByteArray(0) |
|||
} |
|||
} |
|||
|
|||
private fun generateOptimizedSsml(text: String): String { |
|||
val config = currentConfig ?: return "" |
|||
val cleanText = cleanTextForSynthesis(text) |
|||
val escapedText = escapeXmlText(cleanText) |
|||
|
|||
return """ |
|||
<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xml:lang="${config.targetLanguage}"> |
|||
<voice name="${config.voiceName}"> |
|||
<prosody rate="${config.speechRate}" pitch="${config.speechPitch}" volume="${config.speechVolume}"> |
|||
$escapedText |
|||
</prosody> |
|||
</voice> |
|||
</speak> |
|||
""".trimIndent() |
|||
} |
|||
|
|||
private fun cleanTextForSynthesis(text: String): String { |
|||
return text |
|||
.replace(Regex("https?://[^\\s]+"), "") // 移除URL |
|||
.replace(Regex("[\\p{So}\\p{Sk}]"), "") // 移除表情符号 |
|||
.replace(Regex("\\s+"), " ") // 合并多个空格 |
|||
.trim() |
|||
} |
|||
|
|||
private fun escapeXmlText(text: String): String { |
|||
return text |
|||
.replace("&", "&") |
|||
.replace("<", "<") |
|||
.replace(">", ">") |
|||
.replace("\"", """) |
|||
.replace("'", "'") |
|||
} |
|||
} |
|||
@ -0,0 +1,189 @@ |
|||
package com.yunqiinnovation.azure_speech |
|||
|
|||
import android.util.Log |
|||
import kotlinx.coroutines.Dispatchers |
|||
import kotlinx.coroutines.withContext |
|||
import java.io.File |
|||
import java.io.FileOutputStream |
|||
import java.util.UUID |
|||
|
|||
/** |
|||
* 微软翻译与TTS综合服务 |
|||
* |
|||
* 功能： |
|||
* 1. 接收输入文本、源语言代码、目标语言代码 |
|||
* 2. 调用微软翻译服务将文本翻译为目标语言 |
|||
* 3. 调用微软TTS服务将翻译后的文本合成为语音 |
|||
* 4. 将合成的音频保存为 WAV 文件，文件名为 "{targetLanguage}.wav" |
|||
*/ |
|||
class MicrosoftTranslationAndTtsService { |
|||
|
|||
companion object { |
|||
private const val TAG = "MsTransAndTtsService" |
|||
} |
|||
|
|||
private val translationService = MicrosoftTranslationServiceImpl() |
|||
private val ttsService = MicrosoftTTSServiceImpl() |
|||
|
|||
private var serviceConfig: ServiceConfig? = null |
|||
|
|||
/** |
|||
* 服务配置 |
|||
* @param speechSubscriptionKey Azure 语音服务订阅密钥（用于 TTS） |
|||
* @param speechRegion Azure 语音服务区域 (例如 "eastus") |
|||
* @param translationSubscriptionKey Azure 翻译服务订阅密钥（用于文本翻译） |
|||
* @param translationRegion Azure 翻译服务区域 |
|||
*/ |
|||
data class ServiceConfig( |
|||
val speechSubscriptionKey: String, |
|||
val speechRegion: String, |
|||
val translationSubscriptionKey: String = speechSubscriptionKey, |
|||
val translationRegion: String = speechRegion |
|||
) |
|||
|
|||
/** |
|||
* 初始化服务 |
|||
* @param config 服务配置 |
|||
*/ |
|||
suspend fun initialize(config: ServiceConfig): Boolean { |
|||
this.serviceConfig = config |
|||
|
|||
// 初始化翻译服务 |
|||
// 翻译服务支持的配置键: subscriptionKey, region |
|||
val transConfig = mapOf( |
|||
"subscriptionKey" to config.translationSubscriptionKey, |
|||
"region" to config.translationRegion |
|||
) |
|||
|
|||
Log.d(TAG, "正在初始化翻译服务...") |
|||
val transInitSuccess = translationService.initialize(transConfig) |
|||
if (!transInitSuccess) { |
|||
Log.e(TAG, "翻译服务初始化失败") |
|||
return false |
|||
} |
|||
|
|||
Log.d(TAG, "服务初始化成功") |
|||
return true |
|||
} |
|||
|
|||
/** |
|||
* 执行 翻译 -> TTS -> 保存文件 流程 |
|||
* |
|||
* @param text 待翻译和合成的文本 |
|||
* @param sourceLanguage 源语言代码 (例如 "zh-CN") |
|||
* @param targetLanguage 翻译目标语言代码 (例如 "en-US") |
|||
* @param targetTTSLanguage 实际用于TTS的语言代码（可与翻译目标不同） |
|||
* @param outputDirectory 音频文件保存目录 |
|||
* @return 生成的音频文件绝对路径，如果失败则返回 null |
|||
*/ |
|||
suspend fun processAndSaveAudio( |
|||
text: String, |
|||
sourceLanguage: String, |
|||
targetLanguage: String, |
|||
targetTTSLanguage: String, |
|||
outputDirectory: File |
|||
): String? = withContext(Dispatchers.IO) { |
|||
val config = serviceConfig |
|||
if (config == null) { |
|||
Log.e(TAG, "服务尚未初始化，请先调用 initialize()") |
|||
return@withContext null |
|||
} |
|||
|
|||
// --- 步骤 1: 文本翻译 --- |
|||
Log.d(TAG, "开始翻译: text=$text, $sourceLanguage -> $targetLanguage") |
|||
val translationResult = translationService.translateText(text, sourceLanguage, targetLanguage) |
|||
|
|||
if (!translationResult.success || translationResult.translatedText.isNullOrEmpty()) { |
|||
Log.e(TAG, "翻译失败: ${translationResult.error}") |
|||
return@withContext null |
|||
} |
|||
|
|||
val translatedText = translationResult.translatedText!! |
|||
Log.i(TAG, "翻译完成: $translatedText") |
|||
|
|||
// --- 步骤 2: TTS 合成 --- |
|||
Log.d(TAG, "开始TTS合成...") |
|||
|
|||
// 根据上传的 TTS 语言代码优先配置 TTS，其次回退到翻译目标语言 |
|||
val ttsConfig = resolveTtsConfig(targetTTSLanguage, config) |
|||
|
|||
// 创建一个空的 Callback，因为我们主要依赖 synthesize 方法的返回值 |
|||
val ttsCallback = object : MicrosoftTTSServiceImpl.TtsCallback { |
|||
override fun onSynthesisStarted(utteranceId: String, text: String) {} |
|||
override fun onSynthesisProgress(utteranceId: String, text: String, progress: Float) {} |
|||
override fun onAudioChunkGenerated(utteranceId: String, text: String, audioChunk: ByteArray) {} |
|||
override fun onSynthesisCompleted(utteranceId: String, text: String, fullAudio: ByteArray?) {} |
|||
override fun onSynthesisFailed(utteranceId: String, text: String, error: String) { |
|||
Log.w(TAG, "TTS Callback 报告失败: $error") |
|||
} |
|||
override fun onError(message: String) { |
|||
Log.w(TAG, "TTS Callback 报告错误: $message") |
|||
} |
|||
} |
|||
|
|||
val ttsInitSuccess = ttsService.initialize(ttsConfig, ttsCallback) |
|||
if (!ttsInitSuccess) { |
|||
Log.e(TAG, "TTS 服务初始化失败") |
|||
return@withContext null |
|||
} |
|||
|
|||
val utteranceId = UUID.randomUUID().toString() |
|||
val audioData = ttsService.synthesize(utteranceId, translatedText) |
|||
|
|||
if (audioData == null || audioData.isEmpty()) { |
|||
Log.e(TAG, "TTS 合成返回空数据") |
|||
return@withContext null |
|||
} |
|||
Log.i(TAG, "TTS 合成成功，数据大小: ${audioData.size} bytes") |
|||
|
|||
// --- 步骤 3: 保存文件 --- |
|||
try { |
|||
if (!outputDirectory.exists()) { |
|||
outputDirectory.mkdirs() |
|||
} |
|||
|
|||
// 文件名为目标语言代码.wav |
|||
val fileName = "$targetLanguage.wav" |
|||
val outputFile = File(outputDirectory, fileName) |
|||
|
|||
FileOutputStream(outputFile).use { fos -> |
|||
fos.write(audioData) |
|||
fos.flush() |
|||
} |
|||
|
|||
Log.i(TAG, "音频文件已保存: ${outputFile.absolutePath}") |
|||
return@withContext outputFile.absolutePath |
|||
|
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "保存音频文件失败", e) |
|||
return@withContext null |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 根据目标语言构建 TTS 配置 |
|||
* @param targetTTSLanguage 翻译目标语言代码（如 en、en-US、zh-Hans） |
|||
* @param config 服务配置（包含语音服务 key/region） |
|||
*/ |
|||
private fun resolveTtsConfig( |
|||
targetTTSLanguage: String, |
|||
config: ServiceConfig |
|||
): MicrosoftTTSServiceImpl.TtsConfig { |
|||
|
|||
return MicrosoftTTSServiceImpl.TtsConfig( |
|||
subscriptionKey = config.speechSubscriptionKey, |
|||
region = config.speechRegion, |
|||
targetLanguage = targetTTSLanguage, |
|||
voiceName = targetTTSLanguage, |
|||
enableAudioPlayback = false |
|||
) |
|||
} |
|||
|
|||
/** |
|||
* 释放资源 |
|||
*/ |
|||
fun dispose() { |
|||
translationService.dispose() |
|||
ttsService.dispose() |
|||
} |
|||
} |
|||
@ -0,0 +1,268 @@ |
|||
package com.yunqiinnovation.azure_speech |
|||
|
|||
import android.util.Log |
|||
import kotlinx.coroutines.Dispatchers |
|||
import kotlinx.coroutines.delay |
|||
import kotlinx.coroutines.withContext |
|||
import okhttp3.MediaType.Companion.toMediaType |
|||
import okhttp3.OkHttpClient |
|||
import okhttp3.Request |
|||
import okhttp3.RequestBody |
|||
import org.json.JSONArray |
|||
import org.json.JSONObject |
|||
import okhttp3.HttpUrl.Companion.toHttpUrl |
|||
import java.util.UUID |
|||
import java.util.concurrent.TimeUnit |
|||
|
|||
/** |
|||
* 微软翻译服务实现 |
|||
* 实现 `IntegratedSpeechTranslationService.TranslationServiceInterface` 接口 |
|||
*/ |
|||
class MicrosoftTranslationServiceImpl : |
|||
IntegratedSpeechTranslationService.TranslationServiceInterface { |
|||
|
|||
companion object { |
|||
private const val TAG = "MicrosoftTranslationService" |
|||
private const val DEFAULT_ENDPOINT = "https://api.cognitive.microsofttranslator.com" |
|||
private const val API_PATH = "/translate" |
|||
private const val API_VERSION = "3.0" |
|||
} |
|||
|
|||
private var isInitialized = false |
|||
private var subscriptionKey: String = "" |
|||
private var region: String = "" |
|||
private var endpoint: String = DEFAULT_ENDPOINT |
|||
private var maxRetryAttempts: Int = 3 |
|||
private var timeoutMs: Long = 10000L |
|||
|
|||
private val httpClient = OkHttpClient.Builder() |
|||
.connectTimeout(10, TimeUnit.SECONDS) |
|||
.readTimeout(30, TimeUnit.SECONDS) |
|||
.writeTimeout(30, TimeUnit.SECONDS) |
|||
.build() |
|||
|
|||
|
|||
|
|||
/** |
|||
* 初始化翻译服务 |
|||
* @param config 配置映射，支持 `subscriptionKey`/`accessKey`、`region`/`location`、`endpoint`、`maxRetryAttempts`、`timeout` |
|||
* @return 是否初始化成功 |
|||
*/ |
|||
override suspend fun initialize(config: Map<String, String>): Boolean { |
|||
return try { |
|||
subscriptionKey = (config["subscriptionKey"] ?: config["accessKey"] ?: "").trim() |
|||
region = (config["region"] ?: config["location"] ?: "").trim() |
|||
endpoint = (config["endpoint"] ?: DEFAULT_ENDPOINT).trim() |
|||
maxRetryAttempts = config["maxRetryAttempts"]?.toIntOrNull() ?: 3 |
|||
timeoutMs = config["timeout"]?.toLongOrNull() ?: 10000L |
|||
|
|||
Log.d(TAG, "初始化参数: endpoint=$endpoint, region=$region, maxRetryAttempts=$maxRetryAttempts, timeoutMs=$timeoutMs, key=${maskKey(subscriptionKey)}") |
|||
|
|||
if (subscriptionKey.isEmpty()) { |
|||
Log.e(TAG, "缺少必要的订阅密钥(subscriptionKey)") |
|||
false |
|||
} else { |
|||
isInitialized = true |
|||
Log.d(TAG, "微软翻译服务初始化成功: endpoint=$endpoint, region=$region") |
|||
true |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "初始化失败", e) |
|||
false |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 文本翻译 |
|||
* @param text 待翻译文本 |
|||
* @param sourceLanguage 源语言代码（如 `zh-CN`） |
|||
* @param targetLanguage 目标语言代码（如 `en-US`） |
|||
* @return 翻译结果 |
|||
*/ |
|||
override suspend fun translateText( |
|||
text: String, |
|||
sourceLanguage: String, |
|||
targetLanguage: String |
|||
): IntegratedSpeechTranslationService.TranslationResult { |
|||
Log.d(TAG, "翻译请求: 文本=${text.take(20)}..., 源=$sourceLanguage, 目标=$targetLanguage") |
|||
if (!isInitialized) { |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "翻译服务未初始化" |
|||
) |
|||
} |
|||
|
|||
return withContext(Dispatchers.IO) { |
|||
Log.d(TAG, "翻译请求: 文本=${text.take(20)}..., 源=$sourceLanguage, 目标=$targetLanguage") |
|||
performTranslationWithRetry(text, sourceLanguage, targetLanguage) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 带重试的翻译执行 |
|||
*/ |
|||
private suspend fun performTranslationWithRetry( |
|||
text: String, |
|||
sourceLanguage: String, |
|||
targetLanguage: String |
|||
): IntegratedSpeechTranslationService.TranslationResult { |
|||
var lastError: String? = null |
|||
repeat(maxRetryAttempts) { attempt -> |
|||
Log.d(TAG, "翻译重试: 尝试=${attempt + 1}/$maxRetryAttempts, 文本长度=${text.length}, 源=$sourceLanguage, 目标=$targetLanguage") |
|||
val result = performTranslation(text, sourceLanguage, targetLanguage) |
|||
if (result.success) return result |
|||
lastError = result.error |
|||
if (attempt < maxRetryAttempts - 1) delay(500L * (attempt + 1)) |
|||
} |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = lastError ?: "翻译失败" |
|||
) |
|||
} |
|||
|
|||
/** |
|||
* 实际翻译执行 |
|||
*/ |
|||
private suspend fun performTranslation( |
|||
text: String, |
|||
sourceLanguage: String, |
|||
targetLanguage: String |
|||
): IntegratedSpeechTranslationService.TranslationResult { |
|||
try { |
|||
|
|||
Log.d(TAG, "语言映射: sourceLanguage=$sourceLanguage, targetLanguage=$targetLanguage") |
|||
if (sourceLanguage.isEmpty() || targetLanguage.isEmpty()) { |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "不支持的语言代码: $sourceLanguage -> $targetLanguage" |
|||
) |
|||
} |
|||
|
|||
val urlBuilder = endpoint.toHttpUrl().newBuilder() |
|||
.addEncodedPathSegments(API_PATH.trimStart('/')) |
|||
.addQueryParameter("api-version", API_VERSION) |
|||
.addQueryParameter("from", sourceLanguage) |
|||
.addQueryParameter("to", targetLanguage) |
|||
val url = urlBuilder.build() |
|||
Log.d(TAG, "请求URL: $url") |
|||
|
|||
val bodyArray = JSONArray().put(JSONObject().put("text", text)) |
|||
Log.d(TAG, "请求体长度: ${bodyArray.toString().length}") |
|||
val requestBody = RequestBody.create( |
|||
"application/json; charset=utf-8".toMediaType(), |
|||
bodyArray.toString() |
|||
) |
|||
|
|||
val headers = mapOf( |
|||
"Ocp-Apim-Subscription-Key" to subscriptionKey, |
|||
"Ocp-Apim-Subscription-Region" to region, |
|||
"Content-Type" to "application/json", |
|||
"X-ClientTraceId" to UUID.randomUUID().toString() |
|||
) |
|||
Log.d(TAG, "请求头: key=${maskKey(subscriptionKey)}, region=$region") |
|||
|
|||
val request = Request.Builder() |
|||
.url(url) |
|||
.post(requestBody) |
|||
.apply { headers.forEach { (k, v) -> addHeader(k, v) } } |
|||
.build() |
|||
|
|||
val startNs = System.nanoTime() |
|||
val response = httpClient.newCall(request).execute() |
|||
val durationMs = TimeUnit.NANOSECONDS.toMillis(System.nanoTime() - startNs) |
|||
val responseBody = response.body?.string() |
|||
Log.d(TAG, "HTTP响应: code=${response.code}, bodyLen=${responseBody?.length ?: -1}, 耗时=${durationMs}ms") |
|||
if (!response.isSuccessful || responseBody == null) { |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "HTTP错误: ${response.code}" |
|||
) |
|||
} |
|||
|
|||
val translation = extractTranslation(responseBody) |
|||
if (translation != null) { |
|||
Log.d(TAG, "解析翻译成功: 片段=${translation.take(64)}") |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = true, |
|||
translatedText = translation, |
|||
confidence = 0.9f |
|||
) |
|||
} |
|||
|
|||
val error = extractError(responseBody) |
|||
Log.w(TAG, "解析翻译失败: ${error ?: "未知错误"}") |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = error ?: "翻译结果为空" |
|||
) |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "翻译请求异常", e) |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "请求异常: ${e.message}" |
|||
) |
|||
} |
|||
} |
|||
|
|||
|
|||
/** |
|||
* 解析翻译结果 |
|||
*/ |
|||
private fun extractTranslation(responseBody: String): String? { |
|||
return try { |
|||
val root = JSONArray(responseBody) |
|||
if (root.length() == 0) return null |
|||
val item = root.getJSONObject(0) |
|||
if (item.has("translations")) { |
|||
val translations = item.getJSONArray("translations") |
|||
Log.d(TAG, "translations 数量: ${translations.length()}") |
|||
if (translations.length() > 0) { |
|||
val first = translations.getJSONObject(0) |
|||
return first.optString("text", null) |
|||
} |
|||
} |
|||
null |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "提取翻译结果失败", e) |
|||
null |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 解析错误信息 |
|||
*/ |
|||
private fun extractError(responseBody: String): String? { |
|||
return try { |
|||
// 错误响应通常为对象：{"error": {"code": ..., "message": ...}} |
|||
val obj = JSONObject(responseBody) |
|||
if (obj.has("error")) { |
|||
val err = obj.getJSONObject("error") |
|||
Log.w(TAG, "错误对象: code=${err.optString("code")}, message=${err.optString("message")}") |
|||
return err.optString("message", err.optString("code", "未知错误")) |
|||
} |
|||
null |
|||
} catch (_: Exception) { |
|||
null |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 释放资源 |
|||
*/ |
|||
override fun dispose() { |
|||
isInitialized = false |
|||
try { |
|||
Log.d(TAG, "释放HTTP资源") |
|||
httpClient.dispatcher.executorService.shutdown() |
|||
httpClient.connectionPool.evictAll() |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "清理资源失败", e) |
|||
} |
|||
} |
|||
|
|||
private fun maskKey(key: String): String { |
|||
if (key.isEmpty()) return "(empty)" |
|||
val shown = key.take(6) |
|||
return "$shown..." + "*".repeat(8) |
|||
} |
|||
} |
|||
@ -0,0 +1,357 @@ |
|||
package com.yunqiinnovation.azure_speech |
|||
|
|||
import android.util.Log |
|||
import kotlinx.coroutines.Dispatchers |
|||
import kotlinx.coroutines.delay |
|||
import kotlinx.coroutines.withContext |
|||
import okhttp3.MediaType.Companion.toMediaType |
|||
import okhttp3.OkHttpClient |
|||
import okhttp3.Request |
|||
import okhttp3.RequestBody |
|||
import org.json.JSONArray |
|||
import org.json.JSONObject |
|||
import okhttp3.HttpUrl.Companion.toHttpUrl |
|||
import java.net.URLEncoder |
|||
import java.security.MessageDigest |
|||
import java.text.SimpleDateFormat |
|||
import java.util.Date |
|||
import java.util.Locale |
|||
import java.util.TimeZone |
|||
import java.util.concurrent.TimeUnit |
|||
import javax.crypto.Mac |
|||
import javax.crypto.spec.SecretKeySpec |
|||
|
|||
/** |
|||
* 火山翻译服务实现 |
|||
* 实现 `IntegratedSpeechTranslationService.TranslationServiceInterface` 接口 |
|||
*/ |
|||
class VolcanoTranslationServiceImpl : |
|||
IntegratedSpeechTranslationService.TranslationServiceInterface { |
|||
|
|||
companion object { |
|||
private const val TAG = "VolcanoTranslationService" |
|||
private const val BASE_URL = "https://translate.volcengineapi.com" |
|||
private const val ENDPOINT = "/" |
|||
private const val SERVICE = "translate" |
|||
private const val VERSION = "2020-06-01" |
|||
private const val ACTION = "TranslateText" |
|||
private const val ALGORITHM = "HMAC-SHA256" |
|||
} |
|||
|
|||
private var isInitialized = false |
|||
private var accessKey: String = "" |
|||
private var secretKey: String = "" |
|||
private var region: String = "cn-north-1" |
|||
private var maxRetryAttempts: Int = 3 |
|||
private var timeout: Long = 10000L |
|||
|
|||
// HTTP客户端 |
|||
private val httpClient = OkHttpClient.Builder() |
|||
.connectTimeout(10, TimeUnit.SECONDS) |
|||
.readTimeout(30, TimeUnit.SECONDS) |
|||
.writeTimeout(30, TimeUnit.SECONDS) |
|||
.build() |
|||
|
|||
/** |
|||
* 初始化翻译服务 |
|||
* @param config 配置映射，包含 `accessKey`、`secretKey`、`region` 等 |
|||
* @return 是否初始化成功 |
|||
*/ |
|||
override suspend fun initialize(config: Map<String, String>): Boolean { |
|||
return try { |
|||
accessKey = config["accessKey"] ?: "" |
|||
secretKey = config["secretKey"] ?: "" |
|||
region = config["region"] ?: "cn-north-1" |
|||
maxRetryAttempts = config["maxRetryAttempts"]?.toIntOrNull() ?: 3 |
|||
timeout = config["timeout"]?.toLongOrNull() ?: 10000L |
|||
|
|||
if (accessKey.isEmpty() || secretKey.isEmpty()) { |
|||
Log.e(TAG, "缺少必要的API密钥") |
|||
false |
|||
} else { |
|||
isInitialized = true |
|||
Log.d(TAG, "火山翻译服务初始化成功") |
|||
true |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "初始化失败", e) |
|||
false |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 文本翻译 |
|||
* @param text 待翻译文本 |
|||
* @param sourceLanguage 源语言代码 |
|||
* @param targetLanguage 目标语言代码 |
|||
* @return 翻译结果 |
|||
*/ |
|||
override suspend fun translateText( |
|||
text: String, |
|||
sourceLanguage: String, |
|||
targetLanguage: String |
|||
): IntegratedSpeechTranslationService.TranslationResult { |
|||
if (!isInitialized) { |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "翻译服务未初始化" |
|||
) |
|||
} |
|||
Log.d(TAG, "火山翻译服务请求参数: text=$text, sourceLanguage=$sourceLanguage, targetLanguage=$targetLanguage") |
|||
return withContext(Dispatchers.IO) { |
|||
performTranslationWithRetry(text, sourceLanguage, targetLanguage) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 带重试的翻译执行 |
|||
*/ |
|||
private suspend fun performTranslationWithRetry( |
|||
text: String, |
|||
sourceLanguage: String, |
|||
targetLanguage: String |
|||
): IntegratedSpeechTranslationService.TranslationResult { |
|||
var lastException: Exception? = null |
|||
repeat(maxRetryAttempts) { attempt -> |
|||
try { |
|||
val result = performTranslation(text, sourceLanguage, targetLanguage) |
|||
if (result.success) return result |
|||
lastException = Exception(result.error) |
|||
} catch (e: Exception) { |
|||
lastException = e |
|||
Log.w(TAG, "翻译请求失败，尝试 ${attempt + 1}/$maxRetryAttempts", e) |
|||
if (attempt < maxRetryAttempts - 1) delay(500L * (attempt + 1)) |
|||
} |
|||
} |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "翻译失败: ${lastException?.message ?: "未知错误"}" |
|||
) |
|||
} |
|||
|
|||
/** |
|||
* 实际翻译执行 |
|||
*/ |
|||
private suspend fun performTranslation( |
|||
text: String, |
|||
sourceLanguage: String, |
|||
targetLanguage: String |
|||
): IntegratedSpeechTranslationService.TranslationResult { |
|||
try { |
|||
|
|||
if (sourceLanguage.isEmpty() || targetLanguage.isEmpty()) { |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "不支持的语言代码: $sourceLanguage -> $targetLanguage" |
|||
) |
|||
} |
|||
|
|||
val requestBody = JSONObject().apply { |
|||
put("SourceLanguage", sourceLanguage) |
|||
put("TargetLanguage", targetLanguage) |
|||
put("TextList", JSONArray().put(text)) |
|||
} |
|||
val queryParams = mapOf( |
|||
"Action" to ACTION, |
|||
"Version" to VERSION, |
|||
"Region" to region, |
|||
"Service" to SERVICE |
|||
) |
|||
|
|||
val headers = generateSignature("POST", requestBody, queryParams) |
|||
|
|||
val urlBuilder = BASE_URL.toHttpUrl().newBuilder() |
|||
queryParams.forEach { (key, value) -> urlBuilder.addQueryParameter(key, value) } |
|||
val url = urlBuilder.build() |
|||
|
|||
val requestBodyObj = RequestBody.create( |
|||
"application/json; charset=utf-8".toMediaType(), |
|||
requestBody.toString() |
|||
) |
|||
|
|||
val request = Request.Builder() |
|||
.url(url) |
|||
.post(requestBodyObj) |
|||
.apply { headers.forEach { (k, v) -> addHeader(k, v) } } |
|||
.build() |
|||
|
|||
val response = httpClient.newCall(request).execute() |
|||
if (response.isSuccessful) { |
|||
val responseBody = response.body?.string() |
|||
if (responseBody != null) { |
|||
val jsonResponse = JSONObject(responseBody) |
|||
val translation = extractTranslation(jsonResponse) |
|||
if (translation != null) { |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = true, |
|||
translatedText = translation, |
|||
confidence = 0.9f |
|||
) |
|||
} else { |
|||
val error = extractError(jsonResponse) |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = error ?: "翻译结果为空" |
|||
) |
|||
} |
|||
} |
|||
} |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "HTTP错误: ${response.code}" |
|||
) |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "翻译请求异常", e) |
|||
return IntegratedSpeechTranslationService.TranslationResult( |
|||
success = false, |
|||
error = "请求异常: ${e.message}" |
|||
) |
|||
} |
|||
} |
|||
|
|||
|
|||
/** |
|||
* 解析翻译结果 |
|||
*/ |
|||
private fun extractTranslation(jsonResponse: JSONObject): String? { |
|||
try { |
|||
if (jsonResponse.has("TranslationList")) { |
|||
val list = jsonResponse.getJSONArray("TranslationList") |
|||
if (list.length() > 0) { |
|||
val item = list.getJSONObject(0) |
|||
if (item.has("Translation")) return item.getString("Translation") |
|||
} |
|||
} |
|||
if (jsonResponse.has("Translation")) return jsonResponse.getString("Translation") |
|||
if (jsonResponse.has("Result")) { |
|||
val result = jsonResponse.getJSONObject("Result") |
|||
if (result.has("Translation")) return result.getString("Translation") |
|||
} |
|||
if (jsonResponse.has("Data")) { |
|||
val data = jsonResponse.getJSONObject("Data") |
|||
if (data.has("Translation")) return data.getString("Translation") |
|||
if (data.has("TranslationList")) { |
|||
val list = data.getJSONArray("TranslationList") |
|||
if (list.length() > 0) { |
|||
val item = list.getJSONObject(0) |
|||
if (item.has("Translation")) return item.getString("Translation") |
|||
} |
|||
} |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "提取翻译结果失败", e) |
|||
} |
|||
return null |
|||
} |
|||
|
|||
/** |
|||
* 解析错误信息 |
|||
*/ |
|||
private fun extractError(jsonResponse: JSONObject): String? { |
|||
try { |
|||
if (jsonResponse.has("ResponseMetadata")) { |
|||
val metadata = jsonResponse.getJSONObject("ResponseMetadata") |
|||
if (metadata.has("Error")) { |
|||
val error = metadata.getJSONObject("Error") |
|||
return error.optString("Message", error.optString("Code", "未知错误")) |
|||
} |
|||
} |
|||
if (jsonResponse.has("Error")) { |
|||
val error = jsonResponse.getJSONObject("Error") |
|||
return error.optString("Message", error.optString("Code", "未知错误")) |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "提取错误信息失败", e) |
|||
} |
|||
return null |
|||
} |
|||
|
|||
/** |
|||
* 生成签名头 |
|||
*/ |
|||
private fun generateSignature( |
|||
method: String, |
|||
requestBody: JSONObject, |
|||
queryParams: Map<String, String> |
|||
): Map<String, String> { |
|||
val now = Date() |
|||
val dateFormat = SimpleDateFormat("yyyyMMdd", Locale.US).apply { timeZone = TimeZone.getTimeZone("UTC") } |
|||
val timestampFormat = SimpleDateFormat("yyyyMMdd'T'HHmmss'Z'", Locale.US).apply { timeZone = TimeZone.getTimeZone("UTC") } |
|||
val date = dateFormat.format(now) |
|||
val timestamp = timestampFormat.format(now) |
|||
|
|||
val canonicalQueryString = queryParams.toSortedMap().map { (k, v) -> |
|||
"${URLEncoder.encode(k, "UTF-8")}=${URLEncoder.encode(v, "UTF-8")}" |
|||
}.joinToString("&") |
|||
|
|||
val contentType = "application/json" |
|||
val payloadHash = sha256(requestBody.toString()) |
|||
val host = "translate.volcengineapi.com" |
|||
|
|||
val canonicalHeaders = "host:$host\nx-date:$timestamp\n" |
|||
val signedHeaders = "host;x-date" |
|||
|
|||
val canonicalRequest = "$method\n$ENDPOINT\n$canonicalQueryString\n$canonicalHeaders\n$signedHeaders\n$payloadHash" |
|||
val credentialScope = "$date/$region/$SERVICE/request" |
|||
val stringToSign = "$ALGORITHM\n$timestamp\n$credentialScope\n${sha256(canonicalRequest)}" |
|||
|
|||
val kSecret = secretKey.toByteArray(Charsets.UTF_8) |
|||
val kDate = hmacSha256(kSecret, date) |
|||
val kRegion = hmacSha256(kDate, region) |
|||
val kService = hmacSha256(kRegion, SERVICE) |
|||
val kSigning = hmacSha256(kService, "request") |
|||
val signature = hmacSha256(kSigning, stringToSign).joinToString("") { "%02x".format(it) } |
|||
|
|||
val authorization = "$ALGORITHM Credential=$accessKey/$credentialScope, SignedHeaders=$signedHeaders, Signature=$signature" |
|||
return mapOf( |
|||
"Content-Type" to contentType, |
|||
"X-Date" to timestamp, |
|||
"Authorization" to authorization, |
|||
"Host" to host |
|||
) |
|||
} |
|||
|
|||
/** |
|||
* SHA-256 摘要 |
|||
*/ |
|||
private fun sha256(input: String): String { |
|||
val digest = MessageDigest.getInstance("SHA-256") |
|||
val hash = digest.digest(input.toByteArray(Charsets.UTF_8)) |
|||
return hash.joinToString("") { "%02x".format(it) } |
|||
} |
|||
|
|||
/** |
|||
* HMAC-SHA256 计算 |
|||
*/ |
|||
private fun hmacSha256(key: ByteArray, data: String): ByteArray { |
|||
val mac = Mac.getInstance("HmacSHA256") |
|||
val secretKeySpec = SecretKeySpec(key, "HmacSHA256") |
|||
mac.init(secretKeySpec) |
|||
return mac.doFinal(data.toByteArray(Charsets.UTF_8)) |
|||
} |
|||
|
|||
/** |
|||
* 释放资源 |
|||
*/ |
|||
override fun dispose() { |
|||
isInitialized = false |
|||
try { |
|||
val cleanup = Runnable { |
|||
try { |
|||
httpClient.dispatcher.executorService.shutdown() |
|||
httpClient.connectionPool.evictAll() |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "清理资源失败", e) |
|||
} |
|||
} |
|||
if (android.os.Looper.getMainLooper() == android.os.Looper.myLooper()) { |
|||
Thread(cleanup, "VolcanoDisposeThread").start() |
|||
} else { |
|||
cleanup.run() |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "清理资源失败", e) |
|||
} |
|||
} |
|||
} |
|||
|
|||
@ -0,0 +1,235 @@ |
|||
package com.yunqiinnovation.azure_speech.tools |
|||
|
|||
import android.app.Notification |
|||
import android.app.NotificationChannel |
|||
import android.app.NotificationManager |
|||
import android.app.PendingIntent |
|||
import android.app.Service |
|||
import android.content.Context |
|||
import android.content.Intent |
|||
import android.os.Build |
|||
import android.os.IBinder |
|||
import androidx.core.app.NotificationCompat |
|||
import android.util.Log |
|||
import android.media.projection.MediaProjectionManager |
|||
|
|||
/** |
|||
* 屏幕捕获前台服务 |
|||
* 以前台服务的形式执行 MediaProjection 屏幕捕获，满足 Android 14 的安全要求 |
|||
*/ |
|||
class ScreenCaptureForegroundService : Service() { |
|||
|
|||
companion object { |
|||
private const val TAG = "ScreenCaptureService" |
|||
private const val NOTIFICATION_ID = 2001 |
|||
private const val CHANNEL_ID = "screen_capture_channel" |
|||
private const val CHANNEL_NAME = "屏幕捕获服务" |
|||
|
|||
const val ACTION_START = "com.yunqiinnovation.azure_speech.tools.SCREEN_CAPTURE_START" |
|||
const val ACTION_STOP = "com.yunqiinnovation.azure_speech.tools.SCREEN_CAPTURE_STOP" |
|||
|
|||
const val EXTRA_RESULT_CODE = "result_code" |
|||
const val EXTRA_RESULT_DATA = "result_data" |
|||
const val EXTRA_WIDTH = "width" |
|||
const val EXTRA_HEIGHT = "height" |
|||
const val EXTRA_DPI = "dpi" |
|||
const val EXTRA_QUALITY = "quality" |
|||
|
|||
// 全局音频数据监听器，供 Plugin 设置 |
|||
var audioDataListener: ((ByteArray) -> Unit)? = null |
|||
|
|||
/** |
|||
* 启动屏幕捕获前台服务 |
|||
*/ |
|||
fun startService( |
|||
context: Context, |
|||
resultCode: Int, |
|||
data: Intent, |
|||
width: Int?, |
|||
height: Int?, |
|||
dpi: Int?, |
|||
quality: Int |
|||
) { |
|||
val intent = Intent(context, ScreenCaptureForegroundService::class.java).apply { |
|||
action = ACTION_START |
|||
putExtra(EXTRA_RESULT_CODE, resultCode) |
|||
putExtra(EXTRA_RESULT_DATA, data) |
|||
if (width != null) putExtra(EXTRA_WIDTH, width) |
|||
if (height != null) putExtra(EXTRA_HEIGHT, height) |
|||
if (dpi != null) putExtra(EXTRA_DPI, dpi) |
|||
putExtra(EXTRA_QUALITY, quality) |
|||
} |
|||
if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.O) { |
|||
context.startForegroundService(intent) |
|||
} else { |
|||
context.startService(intent) |
|||
} |
|||
Log.d(TAG, "屏幕捕获前台服务启动请求已发送") |
|||
} |
|||
|
|||
/** |
|||
* 停止屏幕捕获前台服务 |
|||
*/ |
|||
fun stopService(context: Context) { |
|||
val intent = Intent(context, ScreenCaptureForegroundService::class.java).apply { |
|||
action = ACTION_STOP |
|||
} |
|||
context.startService(intent) |
|||
Log.d(TAG, "发送停止屏幕捕获服务指令") |
|||
} |
|||
} |
|||
|
|||
private var screenCaptureManager: ScreenCaptureManager? = null |
|||
|
|||
override fun onCreate() { |
|||
super.onCreate() |
|||
createNotificationChannel() |
|||
} |
|||
|
|||
override fun onStartCommand(intent: Intent?, flags: Int, startId: Int): Int { |
|||
val action = intent?.action |
|||
Log.d(TAG, "onStartCommand action=$action") |
|||
|
|||
|
|||
// 如果是停止指令，直接停止服务，不需要 startForeground |
|||
if (action == ACTION_STOP) { |
|||
stopCapture() |
|||
stopForeground(true) |
|||
stopSelf() |
|||
return START_NOT_STICKY |
|||
} |
|||
|
|||
// 创建前台通知，必须尽快调用 |
|||
val notification = createNotification() |
|||
try { |
|||
// 检查权限，虽然 Manifest 声明了，但为了健壮性 |
|||
startForeground(NOTIFICATION_ID, notification) |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "Start foreground failed: ${e.message}") |
|||
stopSelf() |
|||
return START_NOT_STICKY |
|||
} |
|||
when (action) { |
|||
ACTION_START -> { |
|||
try { |
|||
val resultCode = intent.getIntExtra(EXTRA_RESULT_CODE, 0) |
|||
val data = intent.getParcelableExtra<Intent>(EXTRA_RESULT_DATA) |
|||
val width = if (intent.hasExtra(EXTRA_WIDTH)) intent.getIntExtra(EXTRA_WIDTH, 0) else null |
|||
val height = if (intent.hasExtra(EXTRA_HEIGHT)) intent.getIntExtra(EXTRA_HEIGHT, 0) else null |
|||
val dpi = if (intent.hasExtra(EXTRA_DPI)) intent.getIntExtra(EXTRA_DPI, 0) else null |
|||
val quality = intent.getIntExtra(EXTRA_QUALITY, 80) |
|||
|
|||
if (data == null) { |
|||
Log.e(TAG, "缺少授权数据，无法启动屏幕捕获") |
|||
stopSelf() |
|||
return START_NOT_STICKY |
|||
} |
|||
|
|||
screenCaptureManager = ScreenCaptureManager(this) |
|||
// 设置音频回调，将数据转发给全局监听器 |
|||
screenCaptureManager?.setAudioCallback(object : ScreenCaptureManager.AudioDataCallback { |
|||
override fun onAudio(data: ByteArray) { |
|||
// Log.d(TAG, "Service received audio data: ${data.size} bytes, listener: $audioDataListener") |
|||
if (audioDataListener != null) { |
|||
audioDataListener?.invoke(data) |
|||
} else { |
|||
Log.w(TAG, "audioDataListener is null") |
|||
} |
|||
} |
|||
}) |
|||
screenCaptureManager?.startCapture( |
|||
resultCode = resultCode, |
|||
data = data, |
|||
targetWidth = width, |
|||
targetHeight = height, |
|||
targetDpi = dpi, |
|||
jpegQuality = quality |
|||
) |
|||
|
|||
isCapturing = true |
|||
Log.i(TAG, "屏幕捕获已启动") |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "启动屏幕捕获失败: ${e.message}") |
|||
stopSelf() |
|||
return START_NOT_STICKY |
|||
} |
|||
} |
|||
ACTION_STOP -> { |
|||
// 已在顶部处理 |
|||
return START_NOT_STICKY |
|||
} |
|||
else -> { |
|||
// 无操作 |
|||
} |
|||
} |
|||
|
|||
return START_STICKY |
|||
} |
|||
|
|||
override fun onBind(intent: Intent?): IBinder? = null |
|||
|
|||
override fun onDestroy() { |
|||
super.onDestroy() |
|||
stopCapture() |
|||
stopForeground(true) |
|||
} |
|||
|
|||
/** |
|||
* 停止屏幕捕获并释放相关资源 |
|||
*/ |
|||
private fun stopCapture() { |
|||
try { |
|||
screenCaptureManager?.stopCapture() |
|||
|
|||
screenCaptureManager = null |
|||
isCapturing = false |
|||
Log.i(TAG, "屏幕捕获已停止") |
|||
} catch (_: Exception) { } |
|||
} |
|||
|
|||
/** |
|||
* 创建通知渠道 |
|||
*/ |
|||
private fun createNotificationChannel() { |
|||
if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.O) { |
|||
val channel = NotificationChannel( |
|||
CHANNEL_ID, |
|||
CHANNEL_NAME, |
|||
NotificationManager.IMPORTANCE_LOW |
|||
).apply { |
|||
description = "用于屏幕捕获时保持应用运行" |
|||
setShowBadge(false) |
|||
setSound(null, null) |
|||
enableVibration(false) |
|||
} |
|||
val nm = getSystemService(Context.NOTIFICATION_SERVICE) as NotificationManager |
|||
nm.createNotificationChannel(channel) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 创建前台服务通知 |
|||
*/ |
|||
private fun createNotification(): Notification { |
|||
val intent = packageManager.getLaunchIntentForPackage(packageName) |
|||
val pendingIntent = PendingIntent.getActivity( |
|||
this, |
|||
0, |
|||
intent, |
|||
PendingIntent.FLAG_UPDATE_CURRENT or PendingIntent.FLAG_IMMUTABLE |
|||
) |
|||
|
|||
return NotificationCompat.Builder(this, CHANNEL_ID) |
|||
.setContentTitle("语音助手正在运行") |
|||
.setContentText("正在进行屏幕捕获，点击返回应用") |
|||
.setSmallIcon(android.R.drawable.ic_menu_camera) |
|||
.setContentIntent(pendingIntent) |
|||
.setOngoing(true) |
|||
.setPriority(NotificationCompat.PRIORITY_LOW) |
|||
.setCategory(NotificationCompat.CATEGORY_SERVICE) |
|||
.setShowWhen(false) |
|||
.build() |
|||
} |
|||
} |
|||
@Volatile |
|||
var isCapturing: Boolean = false |
|||
@ -0,0 +1,318 @@ |
|||
package com.yunqiinnovation.azure_speech.tools |
|||
|
|||
import android.content.Context |
|||
import android.content.Intent |
|||
import android.graphics.Bitmap |
|||
import android.graphics.PixelFormat |
|||
import android.hardware.display.DisplayManager |
|||
import android.hardware.display.VirtualDisplay |
|||
import android.media.ImageReader |
|||
import android.media.AudioRecord |
|||
import android.media.AudioFormat |
|||
import android.media.AudioAttributes |
|||
import android.media.AudioPlaybackCaptureConfiguration |
|||
import android.os.Build |
|||
import android.os.Environment |
|||
import android.media.projection.MediaProjection |
|||
import android.media.projection.MediaProjectionManager |
|||
import android.os.Handler |
|||
import android.os.HandlerThread |
|||
import android.util.DisplayMetrics |
|||
import android.util.Log |
|||
import android.view.WindowManager |
|||
import java.io.ByteArrayOutputStream |
|||
|
|||
class ScreenCaptureManager(private val context: Context) { |
|||
|
|||
companion object { |
|||
private const val TAG = "ScreenCaptureManager" |
|||
} |
|||
|
|||
private var projectionManager: MediaProjectionManager? = null |
|||
private var mediaProjection: MediaProjection? = null |
|||
private var virtualDisplay: VirtualDisplay? = null |
|||
private var imageReader: ImageReader? = null |
|||
private var handlerThread: HandlerThread? = null |
|||
private var handler: Handler? = null |
|||
private var jpegQuality: Int = 80 |
|||
// 音频捕获 |
|||
private var audioRecord: AudioRecord? = null |
|||
private var audioThread: Thread? = null |
|||
private var isAudioCapturing: java.util.concurrent.atomic.AtomicBoolean = |
|||
java.util.concurrent.atomic.AtomicBoolean(false) |
|||
//private var audioRaf: java.io.RandomAccessFile? = null |
|||
//private var audioFile: java.io.File? = null |
|||
private var totalPcmBytes: Long = 0 |
|||
private val audioSampleRate = 16000 |
|||
private val audioChannels = 1 |
|||
private val audioBitsPerSample = 16 |
|||
private var audioCallback: AudioDataCallback? = null |
|||
|
|||
/** |
|||
* 音频数据回调接口 |
|||
*/ |
|||
interface AudioDataCallback { |
|||
fun onAudio(data: ByteArray) |
|||
} |
|||
|
|||
/** |
|||
* 设置音频数据回调 |
|||
*/ |
|||
fun setAudioCallback(callback: AudioDataCallback?) { |
|||
this.audioCallback = callback |
|||
} |
|||
|
|||
/** |
|||
* 开始屏幕捕获并将帧通过提供的传输器发送 |
|||
* @param resultCode 媒体投射授权结果码 |
|||
* @param data 媒体投射授权返回的Intent |
|||
* @param targetWidth 目标宽度（可选，不传则使用屏幕真实宽度） |
|||
* @param targetHeight 目标高度（可选，不传则使用屏幕真实高度） |
|||
* @param targetDpi 目标DPI（可选，不传则使用屏幕DPI） |
|||
* @param jpegQuality JPEG压缩质量(0-100)，默认80 |
|||
*/ |
|||
fun startCapture( |
|||
resultCode: Int, |
|||
data: Intent, |
|||
targetWidth: Int? = null, |
|||
targetHeight: Int? = null, |
|||
targetDpi: Int? = null, |
|||
jpegQuality: Int = 80 |
|||
) { |
|||
if (mediaProjection != null) return |
|||
|
|||
this.jpegQuality = jpegQuality.coerceIn(0, 100) |
|||
projectionManager = context.getSystemService(Context.MEDIA_PROJECTION_SERVICE) as MediaProjectionManager |
|||
mediaProjection = projectionManager?.getMediaProjection(resultCode, data) |
|||
|
|||
if (mediaProjection == null) { |
|||
Log.e(TAG, "MediaProjection 创建失败") |
|||
return |
|||
} |
|||
|
|||
val metrics = getRealDisplayMetrics() |
|||
val width = targetWidth ?: metrics.widthPixels |
|||
val height = targetHeight ?: metrics.heightPixels |
|||
val dpi = targetDpi ?: metrics.densityDpi |
|||
|
|||
handlerThread = HandlerThread("ScreenCaptureThread").also { it.start() } |
|||
handler = Handler(handlerThread!!.looper) |
|||
|
|||
// Android 14 要求在开始捕获前注册回调以管理资源 |
|||
mediaProjection?.registerCallback(object : MediaProjection.Callback() { |
|||
override fun onStop() { |
|||
Log.i(TAG, "MediaProjection 回调: 捕获停止，清理资源") |
|||
try { |
|||
stopCapture() |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "回调清理资源失败: ${e.message}") |
|||
} |
|||
} |
|||
}, handler) |
|||
|
|||
imageReader = ImageReader.newInstance(width, height, PixelFormat.RGBA_8888, 2) |
|||
imageReader?.setOnImageAvailableListener({ reader -> |
|||
val image = reader.acquireLatestImage() ?: return@setOnImageAvailableListener |
|||
try { |
|||
val w = image.width |
|||
val h = image.height |
|||
val plane = image.planes[0] |
|||
val buffer = plane.buffer |
|||
val pixelStride = plane.pixelStride |
|||
val rowStride = plane.rowStride |
|||
val rowPadding = rowStride - pixelStride * w |
|||
|
|||
val bitmapWidth = w + rowPadding / pixelStride |
|||
val tempBitmap = Bitmap.createBitmap(bitmapWidth, h, Bitmap.Config.ARGB_8888) |
|||
tempBitmap.copyPixelsFromBuffer(buffer) |
|||
|
|||
val finalBitmap = if (bitmapWidth != w) { |
|||
Bitmap.createBitmap(tempBitmap, 0, 0, w, h) |
|||
} else tempBitmap |
|||
|
|||
val baos = ByteArrayOutputStream() |
|||
finalBitmap.compress(Bitmap.CompressFormat.JPEG, this.jpegQuality, baos) |
|||
val bytes = baos.toByteArray() |
|||
baos.reset() |
|||
//屏幕帧 |
|||
//transmitter.sendFrame(bytes) |
|||
|
|||
if (finalBitmap !== tempBitmap) finalBitmap.recycle() |
|||
tempBitmap.recycle() |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "处理屏幕帧失败: ${e.message}") |
|||
} finally { |
|||
image.close() |
|||
} |
|||
}, handler) |
|||
|
|||
virtualDisplay = mediaProjection?.createVirtualDisplay( |
|||
"ScreenCapture", |
|||
width, |
|||
height, |
|||
dpi, |
|||
DisplayManager.VIRTUAL_DISPLAY_FLAG_AUTO_MIRROR, |
|||
imageReader!!.surface, |
|||
null, |
|||
handler |
|||
) |
|||
|
|||
// 启动系统播放音频捕获（Android 10+） |
|||
if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.Q) { |
|||
startAudioCapture() |
|||
} else { |
|||
Log.w(TAG, "当前系统版本不支持AudioPlaybackCapture（需要Android 10+）") |
|||
} |
|||
|
|||
Log.i(TAG, "屏幕捕获已启动 ${width}x${height}@${dpi}dpi") |
|||
} |
|||
|
|||
/** |
|||
* 启动系统播放音频捕获（MediaProjection + AudioPlaybackCaptureConfiguration） |
|||
*/ |
|||
private fun startAudioCapture() { |
|||
try { |
|||
val mp = mediaProjection ?: run { |
|||
Log.e(TAG, "MediaProjection 不可用，无法捕获系统音频") |
|||
return |
|||
} |
|||
|
|||
// 构建播放音频捕获配置：捕获媒体/游戏等播放流 |
|||
val config = AudioPlaybackCaptureConfiguration.Builder(mp) |
|||
.addMatchingUsage(AudioAttributes.USAGE_MEDIA) |
|||
.addMatchingUsage(AudioAttributes.USAGE_GAME) |
|||
.addMatchingUsage(AudioAttributes.USAGE_ASSISTANCE_SONIFICATION) |
|||
.build() |
|||
|
|||
val sampleRate = audioSampleRate |
|||
val channelMask = AudioFormat.CHANNEL_IN_MONO |
|||
val encoding = AudioFormat.ENCODING_PCM_16BIT |
|||
val audioFormat = AudioFormat.Builder() |
|||
.setSampleRate(sampleRate) |
|||
.setChannelMask(channelMask) |
|||
.setEncoding(encoding) |
|||
.build() |
|||
|
|||
audioRecord = AudioRecord.Builder() |
|||
.setAudioPlaybackCaptureConfig(config) |
|||
.setAudioFormat(audioFormat) |
|||
.build() |
|||
|
|||
val minBuf = AudioRecord.getMinBufferSize(sampleRate, channelMask, encoding) |
|||
if (minBuf <= 0) { |
|||
Log.e(TAG, "无法获取有效的缓冲区大小: $minBuf") |
|||
return |
|||
} |
|||
|
|||
// 准备WAV文件输出 |
|||
// prepareWavFile() |
|||
|
|||
audioRecord?.startRecording() |
|||
isAudioCapturing.set(true) |
|||
|
|||
audioThread = Thread { |
|||
Log.i(TAG, "系统播放音频捕获线程已启动") |
|||
val buffer = ByteArray(minBuf) |
|||
var totalBytes: Long = 0 |
|||
try { |
|||
while (isAudioCapturing.get()) { |
|||
val read = audioRecord?.read(buffer, 0, buffer.size) ?: 0 |
|||
// Log.d(TAG, "AudioRecord read result: $read") |
|||
if (read > 0) { |
|||
totalBytes += read |
|||
totalPcmBytes += read |
|||
// 写入PCM到WAV数据区 |
|||
// audioRaf?.write(buffer, 0, read) |
|||
|
|||
// 回调音频数据 |
|||
if (read == buffer.size) { |
|||
audioCallback?.onAudio(buffer.clone()) |
|||
} else { |
|||
val validData = ByteArray(read) |
|||
System.arraycopy(buffer, 0, validData, 0, read) |
|||
audioCallback?.onAudio(validData) |
|||
} |
|||
|
|||
|
|||
|
|||
// 调试输出：打印捕获到的字节数，避免过度日志 |
|||
if (totalBytes % (minBuf * 50L) == 0L) { |
|||
Log.i(TAG, "累计捕获系统音频字节: $totalBytes") |
|||
} |
|||
} |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "系统音频捕获异常: ${e.message}") |
|||
} finally { |
|||
Log.i(TAG, "系统播放音频捕获线程结束，总字节: $totalBytes") |
|||
} |
|||
}.apply { |
|||
name = "PlaybackAudioCapture" |
|||
start() |
|||
} |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "启动系统音频捕获失败: ${e.message}") |
|||
stopAudioCapture() |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 停止系统播放音频捕获并清理资源 |
|||
*/ |
|||
private fun stopAudioCapture() { |
|||
try { |
|||
isAudioCapturing.set(false) |
|||
audioRecord?.stop() |
|||
audioRecord?.release() |
|||
audioRecord = null |
|||
try { audioThread?.join(300) } catch (_: Exception) {} |
|||
audioThread = null |
|||
// finalizeWavFile() |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "停止系统音频捕获失败: ${e.message}") |
|||
} |
|||
} |
|||
|
|||
|
|||
/** |
|||
* 停止屏幕捕获并清理资源 |
|||
*/ |
|||
fun stopCapture() { |
|||
try { |
|||
virtualDisplay?.release() |
|||
virtualDisplay = null |
|||
|
|||
imageReader?.close() |
|||
imageReader = null |
|||
|
|||
mediaProjection?.stop() |
|||
mediaProjection = null |
|||
|
|||
|
|||
// 停止系统播放音频捕获 |
|||
stopAudioCapture() |
|||
|
|||
handlerThread?.quitSafely() |
|||
handlerThread = null |
|||
handler = null |
|||
} catch (e: Exception) { |
|||
Log.e(TAG, "停止捕获失败: ${e.message}") |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 是否处于捕获中 |
|||
* @return true 表示正在捕获，false 表示未捕获 |
|||
*/ |
|||
fun isCapturing(): Boolean { |
|||
return mediaProjection != null && virtualDisplay != null |
|||
} |
|||
|
|||
private fun getRealDisplayMetrics(): DisplayMetrics { |
|||
val metrics = DisplayMetrics() |
|||
val wm = context.getSystemService(Context.WINDOW_SERVICE) as WindowManager |
|||
@Suppress("DEPRECATION") |
|||
wm.defaultDisplay.getRealMetrics(metrics) |
|||
return metrics |
|||
} |
|||
} |
|||
@ -0,0 +1,10 @@ |
|||
# gRPC-Protobuf文件构建指引 |
|||
|
|||
## Go语言 |
|||
|
|||
1. 安装[protoc](https://github.com/protocolbuffers/protobuf/releases/download/v21.12/protoc-21.12-linux-x86_64.zip)和[proto-gen-go](https://github.com/protocolbuffers/protobuf-go/releases/download/v1.32.0/protoc-gen-go.v1.32.0.linux.amd64.tar.gz)命令行工具 |
|||
1. 假定你的Go项目代码仓库为`github.com/foo/bar`，并且已经初始化（若未初始化，请执行命令`go mod init github.com/foo/bar`命令完成初始化） |
|||
1. 将本目录(protos目录）复制到你的Go项目代码仓库中（务必放置在该仓库的根目录中，否则你需要自行修改`protos/build_go.sh`脚本） |
|||
3. 在仓库根目录中执行命令：`sh protos/build_go.sh` |
|||
4. 命令执行结束后，代码会生成到仓库根目录的`protogen`子目录中 |
|||
5. 在Go代码中引用相关package，例：`import "../kotlin/com/yunqiinnovation/azure_speech/protos/github.com/foo/bar/protogen/products/understanding/ast"` |
|||
@ -0,0 +1,38 @@ |
|||
#!/bin/bash |
|||
|
|||
# You must have `protoc` CMD tool installed. You can download compiled binary for Linux here: |
|||
# https://github.com/protocolbuffers/protobuf/releases/download/v21.12/protoc-21.12-linux-x86_64.zip |
|||
|
|||
# You must also have `proto-gen-go` installed, which you can find the compiled binary here: |
|||
# https://github.com/protocolbuffers/protobuf-go/releases/download/v1.32.0/protoc-gen-go.v1.32.0.linux.amd64.tar.gz |
|||
|
|||
set -ex |
|||
|
|||
proto_dir=$(dirname $0) |
|||
cd ${proto_dir}/.. |
|||
|
|||
import_prefix=$(go list -m)/protogen |
|||
|
|||
protoc \ |
|||
--proto_path=${proto_dir} \ |
|||
--go_out=. \ |
|||
--go_opt=Mcommon/events.proto=${import_prefix}/common/event \ |
|||
--go_opt=Mcommon/rpcmeta.proto=${import_prefix}/common/rpcmeta \ |
|||
--go_opt=Mproducts/understanding/base/au_base.proto=${import_prefix}/products/understanding/base \ |
|||
--go_opt=Mproducts/understanding/ast/ast_service.proto=${import_prefix}/products/understanding/ast \ |
|||
--go-grpc_out=. \ |
|||
--go-grpc_opt=Mcommon/events.proto=${import_prefix}/common/event \ |
|||
--go-grpc_opt=Mcommon/rpcmeta.proto=${import_prefix}/common/rpcmeta \ |
|||
--go-grpc_opt=Mproducts/understanding/base/au_base.proto=${import_prefix}/products/understanding/base \ |
|||
--go-grpc_opt=Mproducts/understanding/ast/ast_service.proto=${import_prefix}/products/understanding/ast \ |
|||
${proto_dir}/common/events.proto \ |
|||
${proto_dir}/common/rpcmeta.proto \ |
|||
${proto_dir}/products/understanding/base/au_base.proto \ |
|||
${proto_dir}/products/understanding/ast/ast_service.proto |
|||
|
|||
|
|||
# Copy generated files to the right place and remove redundant directories. |
|||
dest_dir=$(basename ${import_prefix}) |
|||
redundant_dir=${import_prefix%%/*} |
|||
rsync -a --remove-source-files ${import_prefix}/ ${dest_dir}/ |
|||
rm -rf ${redundant_dir} |
|||
@ -0,0 +1,133 @@ |
|||
syntax = "proto3"; |
|||
|
|||
package data.speech.event; |
|||
|
|||
option java_multiple_files = true; |
|||
option java_package = "data.speech.event"; |
|||
|
|||
enum Type { |
|||
option allow_alias = true; |
|||
/* |
|||
* All following event names do not conform to CONSTANT_CASE naming style |
|||
* because we intentionally make them consistant with the existing SAMI |
|||
* string-typed events so that these enum values can be easily converted to |
|||
* event strings by simply call their `String()` method (eg, |
|||
* `event.Type_StartTTS.String() == "StartTTS"`). |
|||
*/ |
|||
|
|||
// 默认事件，适用于不使用事件的方案或不需要传递事件的情况， |
|||
// 或者对于使用事件的方案，可以通过非0值来校验事件的合法性 |
|||
None = 0; |
|||
|
|||
/* |
|||
* 1 ~ 300 为与业务无关的通用事件 |
|||
*/ |
|||
|
|||
/*** 1 ～ 99 为Connection相关事件 ***/ |
|||
|
|||
// 1 ~ 49 为上行Connection事件 |
|||
StartConnection = 1; |
|||
StartTask = 1; // Alias of "StartConnection" |
|||
FinishConnection = 2; |
|||
FinishTask = 2; // Alias of "FinishConnection" |
|||
|
|||
// 50 ~ 99 为下行Connection事件 |
|||
// 成功建连 |
|||
ConnectionStarted = 50; |
|||
TaskStarted = 50; // Alias of "ConnectionStarted" |
|||
// 建连失败（可能是无法通过权限认证） |
|||
ConnectionFailed = 51; |
|||
TaskFailed = 51; // Alias of "ConnectionFailed" |
|||
// 连接结束 |
|||
ConnectionFinished = 52; |
|||
TaskFinished = 52; // Alias of "ConnectionFinished" |
|||
|
|||
/*** 100 ~ 199 为Session、用量等类型事件 ***/ |
|||
|
|||
// 100 ~ 149 为上行Session事件 |
|||
StartSession = 100; |
|||
CancelSession = 101; |
|||
FinishSession = 102; |
|||
|
|||
// 150 ～ 199 为下行Session事件 |
|||
SessionStarted = 150; |
|||
SessionCanceled = 151; |
|||
SessionFinished = 152; |
|||
SessionFailed = 153; |
|||
|
|||
// 用量事件 |
|||
UsageResponse = 154; |
|||
ChargeData = 154; // Alias of "UsageResponse" |
|||
|
|||
/*** 200 ～ 299 为不跟特定方案绑定的通用事件，会被解决方案及下游消费 ***/ |
|||
|
|||
// 200 ~ 249 为上行通用事件 |
|||
TaskRequest = 200; |
|||
UpdateConfig = 201; |
|||
|
|||
ImageRequest = 202; |
|||
|
|||
// 250 ~ 299 为下行通用事件 |
|||
AudioMuted = 250; |
|||
|
|||
/* |
|||
* 300及以上为业务相关的事件 |
|||
*/ |
|||
|
|||
/*** 300 ~ 399 为TTS相关事件 ***/ |
|||
|
|||
// 300 ~ 349 为上行TTS事件 |
|||
SayHello = 300; |
|||
|
|||
// 350 ~ 399 为下行TTS事件 |
|||
TTSSentenceStart = 350; |
|||
TTSSentenceEnd = 351; |
|||
TTSResponse = 352; |
|||
TTSEnded = 359; |
|||
PodcastRoundStart = 360; |
|||
PodcastRoundResponse = 361; |
|||
PodcastRoundEnd = 362; |
|||
|
|||
/*** 400 ~ 499 为ASR相关事件 ***/ |
|||
|
|||
// 400 ~ 449 为上行ASR事件 |
|||
|
|||
// 450 ~ 499 为下行ASR事件 |
|||
ASRInfo = 450; |
|||
ASRResponse = 451; |
|||
ASREnded = 459; |
|||
|
|||
/*** 500 ~ 599 为对话相关事件 ***/ |
|||
|
|||
// 500 ~ 549 为上行对话事件 |
|||
// (Ground-Truth-Alignment) text for speech synthesis |
|||
ChatTTSText = 500; |
|||
ChatTextQuery = 501; |
|||
|
|||
// 550 ~ 599 为下行对话事件 |
|||
ChatResponse = 550; |
|||
ChimeInStart = 551; |
|||
ChimeInEnd = 552; |
|||
ChatEnded = 559; |
|||
ThinkStart = 560; |
|||
ThinkResponse = 561; |
|||
ThinkEnd = 562; |
|||
ToolOutput = 563; |
|||
FCResponseStart = 564; |
|||
FCResponse = 565; |
|||
FCResponseEnd = 566; |
|||
|
|||
/*** 600 ~ 699 为Machine Translation相关事件 ***/ |
|||
|
|||
// 600 ~ 649 为上行对话事件 |
|||
|
|||
// 650 ~ 699 为下行对话事件 |
|||
// Events for source (original) language subtitle. |
|||
SourceSubtitleStart = 650; |
|||
SourceSubtitleResponse = 651; |
|||
SourceSubtitleEnd = 652; |
|||
// Events for target (translation) language subtitle. |
|||
TranslationSubtitleStart = 653; |
|||
TranslationSubtitleResponse = 654; |
|||
TranslationSubtitleEnd = 655; |
|||
} |
|||
@ -0,0 +1,100 @@ |
|||
syntax = "proto3"; |
|||
|
|||
package data.speech.common; |
|||
|
|||
option java_multiple_files = true; |
|||
option java_package = "data.speech.common"; |
|||
|
|||
// NOTE(lucas): We define the fields of RequestMeta and ResponseMeta using |
|||
// PascalCase, which does not comform to normal practice, so that they are |
|||
// mostly compatible with the original definition declared as Go structs in: |
|||
// code.byted.org/lab-speech/gocommon/v2/srvutil/rpcmeta |
|||
|
|||
message RequestMeta { |
|||
// Required. |
|||
// Backend endpoint name. Gateway will also use it to infer HTTP/WebSocket URL |
|||
// path. |
|||
string Endpoint = 1 [ json_name = "endpoint" ]; |
|||
// Required. |
|||
// For historical reasons, this can be either a volcanic appid or a sail |
|||
// platform appKey. If both are transmitted at the same time, you need to put |
|||
// the volcengine appid in the AppID |
|||
string AppKey = 2 [ json_name = "app_key" ]; |
|||
// Optional. |
|||
// appID in volcengine |
|||
string AppID = 3 [ json_name = "app_id" ]; |
|||
// - Required by gateway, you must provide a ResourceID on calling gateway. |
|||
// - Optional for backend services, gateway doesn't need to pass this to |
|||
// backend. |
|||
string ResourceID = 4 [ json_name = "resource_id" ]; |
|||
// Required for Websocket connection (SDK -> Gateway), optional otherwise. |
|||
// In websocket communication, a connection can be reused for multiple |
|||
// simultaneous sessions. Servers use this ID mainly for debugging purposes. |
|||
string ConnectionID = 5 [ json_name = "connection_id" ]; |
|||
// Required. Session is the minimum unit that performs a specific task. The |
|||
// request data in a session can be devided into a number data packets |
|||
// transmitted (but not necessarily processed by the server) in a sequential |
|||
// manner. Each data packet can be denoted by a "Sequence" number (although |
|||
// populating the "Sequence" field is not mandatory). In the simplest case, |
|||
// all request data in a session is sent as a lump (ie, only one packet). |
|||
string SessionID = 6 [ json_name = "session_id" ]; |
|||
// Optional, if not passed, default value 0 will be assumed. For streaming |
|||
// RPC, this field can be omitted (because the order of sending is preserved |
|||
// at the receiving side) unless the client needs the acknowledgement of |
|||
// packet receipt from the server side. |
|||
int32 Sequence = 7 [ json_name = "sequence" ]; |
|||
} |
|||
|
|||
message BillingItem { |
|||
// Required. |
|||
// Billing unit, eg: |
|||
// - minute (billed by the number of minutes used) |
|||
// - word (billed by the number of words submitted or generated) |
|||
// - call (billed by the number of RPC/HTTP/Websocket calls) |
|||
string Unit = 1 [ json_name = "unit" ]; |
|||
// Optional. |
|||
// The amount that the consumer consumed counted by `BillingItem.Unit`. We use |
|||
// a float number because sometimes we want precision better than whole |
|||
// numbers. |
|||
float Quantity = 2 [ json_name = "quantity" ]; |
|||
} |
|||
|
|||
message Billing { |
|||
// Optional. |
|||
// For new billing items, use this field. This is a list because there may be |
|||
// more than one billing items in one request/session. |
|||
repeated BillingItem Items = 1 [ json_name = "items" ]; |
|||
|
|||
/* |
|||
* The following are reserved for compatibility. For new billing items, use |
|||
* field `Items`. |
|||
* We do NOT recommend using `Items` and the fields below at the same time. |
|||
*/ |
|||
|
|||
// Optional. |
|||
// For commodities that are priced w.r.t (typically audio/video) duration in |
|||
// milliseconds. |
|||
int64 DurationMsec = 2 [ json_name = "duration_msec" ]; |
|||
// Optional. |
|||
// For commodities priced w.r.t text length (number of words/tokens). |
|||
int64 WordCount = 3 [ json_name = "word_count" ]; |
|||
} |
|||
|
|||
message ResponseMeta { |
|||
// Required. |
|||
// The same SessionID in RequestMeta. |
|||
string SessionID = 1 [ json_name = "session_id" ]; |
|||
// Optional. |
|||
// The same sequence in RequestMeta (except that server *may* turn a positive |
|||
// sequence to its negative counterpart). |
|||
int32 Sequence = 2 [ json_name = "sequence" ]; |
|||
// Optional. |
|||
// Response status code. |
|||
int32 StatusCode = 3 [ json_name = "status_code" ]; |
|||
// Optional. |
|||
// Detailed status information. |
|||
string Message = 4 [ json_name = "message" ]; |
|||
// Billing information in case only backend servers can collect this |
|||
// information. |
|||
optional Billing Billing = 5 [ json_name = "billing" ]; |
|||
} |
|||
@ -0,0 +1,46 @@ |
|||
syntax = "proto3"; |
|||
|
|||
import "common/events.proto"; |
|||
import "common/rpcmeta.proto"; |
|||
import "products/understanding/base/au_base.proto"; |
|||
|
|||
package data.speech.ast; |
|||
|
|||
option java_multiple_files = true; |
|||
option java_package = "data.speech.ast"; |
|||
|
|||
message ReqParams { |
|||
string mode = 1; // 可能是s2t , s2s 选一个, 控制是否需要语音 |
|||
string source_language = 2; // 源语言 |
|||
string target_language = 3; // 目标语言 |
|||
string speaker_id = 4; |
|||
|
|||
data.speech.understanding.Corpus corpus = 100; |
|||
} |
|||
|
|||
message TranslateRequest { |
|||
optional data.speech.common.RequestMeta request_meta = 1; |
|||
data.speech.event.Type event = 2; |
|||
data.speech.understanding.User user = 3; |
|||
// bytes data = 2; // request binary data |
|||
data.speech.understanding.Audio source_audio = 4; // 源音频信息 |
|||
// 目标音频信息，只需要传format（pcm/ogg）、rate、bits、channel这些 |
|||
data.speech.understanding.Audio target_audio = 5; |
|||
ReqParams request = 6; // 请求参数 |
|||
optional bool denoise = 7; // 是否开启降噪 |
|||
} |
|||
|
|||
message TranslateResponse { |
|||
optional data.speech.common.ResponseMeta response_meta = 1; |
|||
data.speech.event.Type event = 2; |
|||
bytes data = 3; // response binary data |
|||
string text = 4; // 原文或者译文 |
|||
int32 start_time = 5; |
|||
int32 end_time = 6; |
|||
bool spk_chg = 7; |
|||
int32 muted_duration_ms = 8; |
|||
} |
|||
|
|||
service ASTService { |
|||
rpc Translate(stream TranslateRequest) returns (stream TranslateResponse); |
|||
} |
|||
@ -0,0 +1,195 @@ |
|||
syntax = "proto3"; |
|||
|
|||
package data.speech.understanding; |
|||
|
|||
option java_multiple_files = true; |
|||
option java_package = "data.speech.understanding"; |
|||
|
|||
enum Code { |
|||
ERROR_UNSPECIFIED = 0; |
|||
|
|||
SUCCESS = 21000; |
|||
|
|||
// request related |
|||
INVALID_REQUEST = 11100; |
|||
PERMISSION_DENIED = 11200; |
|||
LIMIT_QPS = 11301; |
|||
LIMIT_COUNT = 11302; |
|||
SERVER_BUSY = 11303; |
|||
INTERRUPTED = 21300; |
|||
ERROR_PARAMS = 11500; |
|||
|
|||
// audio related |
|||
LONG_AUDIO = 11101; |
|||
LARGE_PACKET = 11102; |
|||
INVALID_FORMAT = 11103; |
|||
SILENT_AUDIO = 11104; |
|||
EMPTY_AUDIO = 11105; |
|||
AUDIO_DOWNLOAD_FAIL = 21701; |
|||
|
|||
// recognition related |
|||
TIMEOUT_WAITING = 21200; |
|||
TIMEOUT_PROCESSING = 21201; |
|||
ERROR_PROCESSING = 21100; |
|||
|
|||
// others |
|||
ERROR_UNKNOWN = 29900; |
|||
} |
|||
|
|||
message Word { |
|||
string text = 1; |
|||
int32 start_time = 2; |
|||
int32 end_time = 3; |
|||
int32 blank_duration = 4; |
|||
string pronounce = 5; |
|||
double confidence = 6; |
|||
} |
|||
message UtteranceAddition {} |
|||
|
|||
message ResultAddition {} |
|||
|
|||
message Utterance { |
|||
string text = 1; |
|||
int32 start_time = 2; |
|||
int32 end_time = 3; |
|||
optional bool definite = 4; |
|||
repeated Word words = 5; |
|||
string language = 6; |
|||
double confidence = 7; |
|||
int32 speaker = 8; |
|||
int32 channel_id = 9; |
|||
|
|||
map<string, string> additions = 100; |
|||
} |
|||
|
|||
message LanguageDetail { |
|||
// language probability |
|||
double prob = 1; |
|||
} |
|||
|
|||
message Result { |
|||
string text = 1; |
|||
repeated Utterance utterances = 2; |
|||
double confidence = 3; |
|||
double global_confidence = 4; |
|||
string language = 5; |
|||
map<string, LanguageDetail> language_details = 6; |
|||
optional bool termination = 7; |
|||
double volume = 8; |
|||
double speech_rate = 9; |
|||
string result_type = 10; |
|||
string ori_text = 11; |
|||
optional bool prefetch = 12; |
|||
|
|||
map<string, string> additions = 100; |
|||
} |
|||
|
|||
message AudioInfo { |
|||
int64 duration = 1; |
|||
double speech_rate = 2; |
|||
} |
|||
|
|||
message App { string work_flow_name = 1; } |
|||
|
|||
message User { |
|||
string uid = 1; |
|||
string did = 2; |
|||
// android or ios info |
|||
string platform = 3; |
|||
string sdk_version = 4; |
|||
string app_version = 5; |
|||
} |
|||
|
|||
message Audio { |
|||
string data = 1; |
|||
string url = 2; |
|||
// url; vid... |
|||
string url_type = 3; |
|||
string format = 4; |
|||
string codec = 5; |
|||
string language = 6; |
|||
int32 rate = 7; |
|||
int32 bits = 8; |
|||
int32 channel = 9; |
|||
string tos_bucket = 10; |
|||
string tos_access_key = 11; |
|||
string audio_tos_object = 12; |
|||
string role_trn = 13; |
|||
bytes binary_data = 14; |
|||
} |
|||
|
|||
enum AsrNotificationCode { |
|||
NOTIFICATION_UNSPECIFIED = 0; |
|||
DISCONNECT = 1000; |
|||
} |
|||
|
|||
message AsrNotification { |
|||
AsrNotificationCode code = 1; |
|||
string message = 2; |
|||
} |
|||
|
|||
message ReqParams { |
|||
string model_name = 1; // 用于映射集群 |
|||
string config_name = 2; // 用于映射tcc配置 |
|||
optional bool enable_vad = 3; |
|||
optional bool enable_punc = 4; |
|||
optional bool enable_itn = 5; |
|||
optional bool enable_ddc = 6; |
|||
optional bool enable_resample = 7; |
|||
optional bool vad_signal = 8; |
|||
optional bool enable_translate = 9; |
|||
optional bool disable_end_punc = 10; |
|||
optional bool enable_speaker_info = 11; |
|||
optional bool enable_channel_split = 12; |
|||
string app_lang = 13; |
|||
string country = 14; |
|||
string audio_type = 15; |
|||
optional bool enable_lid = 16; |
|||
int32 reco2act_num_spk = 17; |
|||
int32 reco2max_num_spk = 18; |
|||
optional bool enable_interrupt = 19; |
|||
optional bool vad_segment = 20; |
|||
optional bool enable_ori_text = 21; |
|||
optional bool enable_nonstream = 22; |
|||
|
|||
optional bool show_utterances = 31; |
|||
optional bool show_words = 32; |
|||
optional bool show_duration = 33; |
|||
optional bool show_language = 34; |
|||
optional bool show_volume = 35; |
|||
optional bool show_volume_v = 362; |
|||
optional bool show_speech_rate = 37; |
|||
optional bool show_prefetch = 38; |
|||
string result_type = 39; |
|||
|
|||
int32 start_silence_time = 61; |
|||
int32 end_silence_time = 62; |
|||
int32 vad_silence_time = 63; |
|||
int32 min_silence_time = 64; |
|||
string vad_mode = 65; |
|||
int32 vad_segment_duration = 66; |
|||
int32 end_window_size = 67; |
|||
int32 force_to_speech_time = 68; |
|||
|
|||
Corpus corpus = 100; |
|||
AsrNotification notification = 101; |
|||
string sensitive_words_filter = 102; |
|||
} |
|||
|
|||
message Corpus { |
|||
// json格式，包含热词，冷词，篇章，纠错词表等 |
|||
string context = 1; |
|||
string boosting_table_id = 2; |
|||
string boosting_table_name = 3; |
|||
string boosting_id = 4; |
|||
string nnlm_id = 5; |
|||
string correct_table_id = 6; |
|||
string correct_table_name = 7; |
|||
string punc_hot_words = 8; |
|||
repeated string hot_words_list = 9; |
|||
map<string, string> glossary_list = 10; |
|||
|
|||
optional bool show_match_hot_words = 61; |
|||
optional bool show_all_match_hot_words = 62; |
|||
optional bool show_effective_hot_words = 63; |
|||
} |
|||
@ -0,0 +1,427 @@ |
|||
import Foundation |
|||
import os.log |
|||
|
|||
/** |
|||
* 阿里云百炼（Bailian）实时音视频翻译助手 (qwen3-livetranslate-flash-realtime) |
|||
* iOS 实现，使用 URLSessionWebSocketTask 和 JSON 协议。 |
|||
*/ |
|||
class AliyunBailianE2EHelper: NSObject, URLSessionWebSocketDelegate { |
|||
/** |
|||
* 回调接口 |
|||
*/ |
|||
protocol Callback: AnyObject { |
|||
func onSessionStarted(sessionId: String) |
|||
func onPartialText(sessionId: String, text: String) |
|||
func onPartialSourceText(sessionId: String, text: String) |
|||
func onFinalSourceText(sessionId: String, finalText: String) |
|||
func onFinalTranslatedText(sessionId: String, finalText: String) |
|||
func onPartialAudio(sessionId: String, data: Data) |
|||
func onSessionFinished(sessionId: String, finalText: String, finalAudio: Data) |
|||
func onSessionError(sessionId: String, code: Int, message: String) |
|||
} |
|||
|
|||
/** |
|||
* 配置结构 |
|||
*/ |
|||
struct Config { |
|||
var wsUrl: String = "wss://dashscope.aliyuncs.com/api-ws/v1/realtime" |
|||
var apiKey: String = "" |
|||
var appId: String = "qwen3-livetranslate-flash-realtime" |
|||
var sampleRate: Int = 16000 |
|||
var sourceLanguage: String = "zh" |
|||
var targetLanguage: String = "en" |
|||
var voice: String = "" |
|||
} |
|||
|
|||
private let tag = "AliyunBailianE2EHelper" |
|||
private let log = OSLog(subsystem: "com.azure.speech", category: "AliyunBailian") |
|||
|
|||
private var conf = Config() |
|||
private weak var callback: Callback? |
|||
|
|||
private var urlSession: URLSession? |
|||
private var webSocket: URLSessionWebSocketTask? |
|||
private var sessionId: String = "" |
|||
|
|||
private var fullTextBuffer = "" |
|||
private var fullAudioBuffer = Data() |
|||
|
|||
private var recvTextBuffer = "" |
|||
private var audioChunkBuffer = Data() |
|||
|
|||
private var isStarted = false |
|||
|
|||
deinit { |
|||
os_log("AliyunBailianE2EHelper deinit", log: log, type: .info) |
|||
} |
|||
|
|||
/** |
|||
* 初始化助手（异步） |
|||
* - 参数 config: 会话配置 |
|||
* - 参数 cb: 回调实现 |
|||
* - 返回: 是否成功触发连接逻辑 |
|||
*/ |
|||
func initialize(config: Config, cb: Callback) async -> Bool { |
|||
os_log("initialize: wsUrl=%{public}@", log: log, type: .info, config.wsUrl) |
|||
urlSession?.invalidateAndCancel() |
|||
urlSession = nil |
|||
|
|||
conf = config |
|||
callback = cb |
|||
|
|||
let configuration = URLSessionConfiguration.default |
|||
configuration.timeoutIntervalForRequest = 0 |
|||
configuration.timeoutIntervalForResource = 0 |
|||
|
|||
urlSession = URLSession(configuration: configuration, delegate: self, delegateQueue: OperationQueue.main) |
|||
|
|||
return startContinuousConversation() |
|||
} |
|||
|
|||
/** |
|||
* 启动会话并建立 WebSocket 连接 |
|||
*/ |
|||
func startContinuousConversation() -> Bool { |
|||
guard !isStarted else { |
|||
os_log("Already started, ignore start request", log: log, type: .info) |
|||
return true |
|||
} |
|||
guard let session = urlSession else { |
|||
os_log("URLSession is nil, cannot start", log: log, type: .error) |
|||
return false |
|||
} |
|||
|
|||
let urlWithModel: String |
|||
if conf.wsUrl.contains("?") { |
|||
urlWithModel = conf.wsUrl |
|||
} else { |
|||
urlWithModel = "\(conf.wsUrl)?model=qwen3-livetranslate-flash-realtime" |
|||
} |
|||
|
|||
guard let url = URL(string: urlWithModel) else { |
|||
os_log("Invalid wsUrl: %{public}@", log: log, type: .error, urlWithModel) |
|||
return false |
|||
} |
|||
|
|||
sessionId = UUID().uuidString |
|||
os_log("Start session: %{public}@", log: log, type: .info, sessionId) |
|||
|
|||
recvTextBuffer = "" |
|||
fullTextBuffer = "" |
|||
fullAudioBuffer.removeAll() |
|||
audioChunkBuffer.removeAll() |
|||
|
|||
var request = URLRequest(url: url) |
|||
request.timeoutInterval = 60 |
|||
request.addValue("Bearer \(conf.apiKey)", forHTTPHeaderField: "Authorization") |
|||
|
|||
webSocket = session.webSocketTask(with: request) |
|||
webSocket?.resume() |
|||
isStarted = true |
|||
return true |
|||
} |
|||
|
|||
/** |
|||
* 发送 session.update 配置会话 |
|||
*/ |
|||
private func sendSessionUpdate() { |
|||
guard let ws = webSocket else { return } |
|||
do { |
|||
var session: [String: Any] = [:] |
|||
|
|||
var transcription: [String: Any] = [:] |
|||
transcription["model"] = "qwen3-asr-flash-realtime" |
|||
transcription["language"] = conf.sourceLanguage |
|||
session["input_audio_transcription"] = transcription |
|||
|
|||
var translation: [String: Any] = [:] |
|||
translation["language"] = conf.targetLanguage |
|||
session["translation"] = translation |
|||
|
|||
if !conf.voice.isEmpty { |
|||
session["voice"] = conf.voice |
|||
} else { |
|||
if conf.targetLanguage != "yue" { |
|||
session["voice"] = "Cherry" |
|||
} else { |
|||
session["voice"] = "Kiki" |
|||
} |
|||
} |
|||
session["input_audio_format"] = "pcm16" |
|||
session["output_audio_format"] = "pcm24" |
|||
|
|||
var modalities: [String] = [] |
|||
modalities.append("text") |
|||
modalities.append("audio") |
|||
session["modalities"] = modalities |
|||
|
|||
var root: [String: Any] = [:] |
|||
root["type"] = "session.update" |
|||
root["session"] = session |
|||
|
|||
let data = try JSONSerialization.data(withJSONObject: root, options: []) |
|||
if let jsonString = String(data: data, encoding: .utf8) { |
|||
ws.send(.string(jsonString)) { [weak self] error in |
|||
if let e = error { |
|||
os_log("Error sending session.update: %{public}@", log: self?.log ?? .default, type: .error, e.localizedDescription) |
|||
} else { |
|||
os_log("Sent session.update: %{public}@", log: self?.log ?? .default, type: .info, jsonString) |
|||
} |
|||
} |
|||
} |
|||
} catch { |
|||
os_log("Error building session.update: %{public}@", log: log, type: .error, error.localizedDescription) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 将 24kHz PCM 重采样为 16kHz |
|||
*/ |
|||
private func resample24kTo16k(_ input: Data) -> Data { |
|||
if input.count < 4 { return input } |
|||
|
|||
let sampleCount = input.count / 2 |
|||
var inputSamples = [Int16](repeating: 0, count: sampleCount) |
|||
_ = input.withUnsafeBytes { raw -> Void in |
|||
guard let base = raw.bindMemory(to: Int16.self).baseAddress else { return } |
|||
for i in 0..<sampleCount { |
|||
inputSamples[i] = Int16(littleEndian: base[i]) |
|||
} |
|||
} |
|||
|
|||
let outputCount = sampleCount * 2 / 3 |
|||
if outputCount == 0 { return Data() } |
|||
var outputSamples = [Int16](repeating: 0, count: outputCount) |
|||
|
|||
for i in 0..<outputCount { |
|||
let inputIndexFloat = Float(i) * 1.5 |
|||
let index = Int(inputIndexFloat) |
|||
let frac = inputIndexFloat - Float(index) |
|||
|
|||
if index + 1 < sampleCount { |
|||
let v1 = Float(inputSamples[index]) |
|||
let v2 = Float(inputSamples[index + 1]) |
|||
let v = v1 * (1.0 - frac) + v2 * frac |
|||
outputSamples[i] = Int16(max(min(Int(v), Int(Int16.max)), Int(Int16.min))) |
|||
} else if index < sampleCount { |
|||
outputSamples[i] = inputSamples[index] |
|||
} |
|||
} |
|||
|
|||
var outData = Data(count: outputSamples.count * 2) |
|||
outData.withUnsafeMutableBytes { raw in |
|||
let ptr = raw.bindMemory(to: Int16.self).baseAddress |
|||
ptr?.assign(from: outputSamples, count: outputSamples.count) |
|||
} |
|||
return outData |
|||
} |
|||
|
|||
/** |
|||
* 处理服务端 JSON 文本消息 |
|||
*/ |
|||
private func handleJsonMessage(_ text: String) { |
|||
guard let data = text.data(using: .utf8) else { return } |
|||
guard let obj = try? JSONSerialization.jsonObject(with: data) as? [String: Any] else { return } |
|||
let type = obj["type"] as? String ?? "" |
|||
|
|||
os_log("handleJsonMessage: type=%{public}@", log: log, type: .info, type) |
|||
|
|||
switch type { |
|||
case "error": |
|||
if let errorObj = obj["error"] as? [String: Any] { |
|||
let msg = errorObj["message"] as? String ?? "Unknown error" |
|||
let codeStr = errorObj["code"] as? String ?? "" |
|||
os_log("Error from server: %{public}@ - %{public}@", log: log, type: .error, codeStr, msg) |
|||
callback?.onSessionError(sessionId: sessionId, code: 1012, message: msg) |
|||
} |
|||
case "session.created": |
|||
os_log("Session created", log: log, type: .info) |
|||
case "session.updated": |
|||
os_log("Session updated", log: log, type: .info) |
|||
case "conversation.item.input_audio_transcription.text": |
|||
if let txt = obj["text"] as? String, !txt.isEmpty { |
|||
os_log("onPartialSourceText: %{public}@", log: log, type: .info, txt) |
|||
callback?.onPartialSourceText(sessionId: sessionId, text: txt) |
|||
} |
|||
case "conversation.item.input_audio_transcription.completed": |
|||
var finalTxt = "" |
|||
if let item = obj["item"] as? [String: Any], |
|||
let content = item["content"] as? [[String: Any]] { |
|||
for part in content { |
|||
if let t = part["type"] as? String, t == "input_audio" { |
|||
finalTxt = part["transcript"] as? String ?? "" |
|||
if !finalTxt.isEmpty { break } |
|||
} |
|||
} |
|||
} |
|||
if finalTxt.isEmpty { |
|||
finalTxt = obj["transcript"] as? String ?? "" |
|||
} |
|||
if !finalTxt.isEmpty { |
|||
os_log("onFinalSourceText: %{public}@", log: log, type: .info, finalTxt) |
|||
callback?.onFinalSourceText(sessionId: sessionId, finalText: finalTxt) |
|||
} |
|||
case "response.audio_transcript.delta": |
|||
if let delta = obj["delta"] as? String, !delta.isEmpty { |
|||
recvTextBuffer.append(delta) |
|||
os_log("onPartialText: %{public}@", log: log, type: .info, delta) |
|||
callback?.onPartialText(sessionId: sessionId, text: delta) |
|||
} |
|||
case "response.audio_transcript.done": |
|||
let transcript = obj["transcript"] as? String ?? "" |
|||
let finalText = transcript.isEmpty ? recvTextBuffer : transcript |
|||
os_log("onFinalTranslatedText: %{public}@", log: log, type: .info, finalText) |
|||
callback?.onFinalTranslatedText(sessionId: sessionId, finalText: finalText) |
|||
|
|||
if !fullTextBuffer.isEmpty { |
|||
fullTextBuffer.append(" ") |
|||
} |
|||
fullTextBuffer.append(finalText) |
|||
recvTextBuffer = "" |
|||
case "response.audio.delta": |
|||
if let base64Audio = obj["delta"] as? String, !base64Audio.isEmpty { |
|||
if let audioBytes = Data(base64Encoded: base64Audio) { |
|||
let resampled = resample24kTo16k(audioBytes) |
|||
if !resampled.isEmpty { |
|||
fullAudioBuffer.append(resampled) |
|||
os_log("onPartialAudio: rawSize=%{public}d resampledSize=%{public}d", log: log, type: .info, audioBytes.count, resampled.count) |
|||
processAudioChunk(incoming: resampled, multiple: 1280) |
|||
} |
|||
} |
|||
} |
|||
case "response.done": |
|||
os_log("Response done", log: log, type: .info) |
|||
default: |
|||
break |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 音频分片累积并按 multiple 字节回调 |
|||
*/ |
|||
private func processAudioChunk(incoming: Data, multiple: Int = 1280) { |
|||
if !incoming.isEmpty { |
|||
audioChunkBuffer.append(incoming) |
|||
} |
|||
let sendLen = (audioChunkBuffer.count / multiple) * multiple |
|||
if sendLen > 0 { |
|||
let chunk = audioChunkBuffer.prefix(sendLen) |
|||
callback?.onPartialAudio(sessionId: sessionId, data: Data(chunk)) |
|||
audioChunkBuffer = Data(audioChunkBuffer.dropFirst(sendLen)) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 推送音频 PCM 数据（Base64 编码） |
|||
*/ |
|||
func pushAudioData(_ data: Data) -> Bool { |
|||
guard isStarted, let ws = webSocket else { |
|||
os_log("pushAudioData ignored: isStarted=%{public}d, ws is nil", log: log, type: .info, isStarted) |
|||
return false |
|||
} |
|||
os_log("pushAudioData: %{public}d bytes", log: log, type: .info, data.count) |
|||
do { |
|||
let base64Audio = data.base64EncodedString() |
|||
var root: [String: Any] = [:] |
|||
root["type"] = "input_audio_buffer.append" |
|||
root["audio"] = base64Audio |
|||
let jsonData = try JSONSerialization.data(withJSONObject: root, options: []) |
|||
if let jsonString = String(data: jsonData, encoding: .utf8) { |
|||
ws.send(.string(jsonString)) { [weak self] error in |
|||
if let e = error { |
|||
os_log("pushAudioData error: %{public}@", log: self?.log ?? .default, type: .error, e.localizedDescription) |
|||
} |
|||
} |
|||
} |
|||
return true |
|||
} catch { |
|||
os_log("pushAudioData serialization error: %{public}@", log: log, type: .error, error.localizedDescription) |
|||
return false |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 停止会话 |
|||
*/ |
|||
func stopContinuousConversation() -> Bool { |
|||
os_log("stopContinuousConversation", log: log, type: .info) |
|||
if let ws = webSocket { |
|||
ws.cancel(with: .normalClosure, reason: "User stopped".data(using: .utf8)) |
|||
} |
|||
isStarted = false |
|||
return true |
|||
} |
|||
|
|||
/** |
|||
* 释放资源 |
|||
*/ |
|||
func dispose() { |
|||
os_log("dispose", log: log, type: .info) |
|||
do { |
|||
try webSocket?.cancel(with: .normalClosure, reason: "dispose".data(using: .utf8)) |
|||
} catch { |
|||
} |
|||
webSocket = nil |
|||
isStarted = false |
|||
urlSession?.invalidateAndCancel() |
|||
urlSession = nil |
|||
} |
|||
|
|||
/** |
|||
* WebSocket 打开回调 |
|||
*/ |
|||
func urlSession(_ session: URLSession, webSocketTask: URLSessionWebSocketTask, didOpenWithProtocol protocol: String?) { |
|||
os_log("WebSocket didOpen", log: log, type: .info) |
|||
sendSessionUpdate() |
|||
callback?.onSessionStarted(sessionId: sessionId) |
|||
receiveLoop() |
|||
} |
|||
|
|||
/** |
|||
* WebSocket 接收循环 |
|||
*/ |
|||
private func receiveLoop() { |
|||
webSocket?.receive { [weak self] result in |
|||
guard let self = self else { return } |
|||
switch result { |
|||
case .failure(let error): |
|||
os_log("WebSocket receive error: %{public}@", log: self.log, type: .error, error.localizedDescription) |
|||
self.callback?.onSessionError(sessionId: self.sessionId, code: 1011, message: error.localizedDescription) |
|||
case .success(let message): |
|||
switch message { |
|||
case .string(let text): |
|||
self.handleJsonMessage(text) |
|||
case .data(let data): |
|||
if let text = String(data: data, encoding: .utf8) { |
|||
self.handleJsonMessage(text) |
|||
} |
|||
@unknown default: |
|||
break |
|||
} |
|||
self.receiveLoop() |
|||
} |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* WebSocket 关闭回调 |
|||
*/ |
|||
func urlSession(_ session: URLSession, webSocketTask: URLSessionWebSocketTask, didCloseWith closeCode: URLSessionWebSocketTask.CloseCode, reason: Data?) { |
|||
let reasonStr = String(data: reason ?? Data(), encoding: .utf8) ?? "" |
|||
os_log("WebSocket didClose code=%{public}d reason=%{public}@", log: log, type: .info, closeCode.rawValue, reasonStr) |
|||
isStarted = false |
|||
callback?.onSessionFinished(sessionId: sessionId, finalText: fullTextBuffer, finalAudio: fullAudioBuffer) |
|||
} |
|||
|
|||
/** |
|||
* 任务完成回调（错误处理） |
|||
*/ |
|||
func urlSession(_ session: URLSession, task: URLSessionTask, didCompleteWithError error: Error?) { |
|||
if let e = error { |
|||
os_log("WebSocket task error: %{public}@", log: log, type: .error, e.localizedDescription) |
|||
callback?.onSessionError(sessionId: sessionId, code: 1012, message: e.localizedDescription) |
|||
} |
|||
isStarted = false |
|||
} |
|||
} |
|||
|
|||
@ -0,0 +1,650 @@ |
|||
import Foundation |
|||
import os.log |
|||
|
|||
/** |
|||
* Doubao AST 流式翻译助手（iOS） |
|||
* 提供与 Android `DoubaoE2ETranslateHelper` 类似的接口： |
|||
* - 初始化并建立 WebSocket 连接 |
|||
* - 推送音频分片到服务端 |
|||
* - 接收增量文本与音频分片 |
|||
* - 在会话结束时回传完整文本与合成音频 |
|||
*/ |
|||
class DoubaoE2ETranslateHelper: NSObject, URLSessionWebSocketDelegate { |
|||
/** |
|||
* 回调接口 |
|||
*/ |
|||
protocol Callback: AnyObject { |
|||
func onSessionStarted(sessionId: String) |
|||
func onPartialText(sessionId: String, text: String) |
|||
func onPartialSourceText(sessionId: String, text: String) |
|||
func onFinalSourceText(sessionId: String, finalText: String) |
|||
func onFinalTranslatedText(sessionId: String, finalText: String) |
|||
func onPartialAudio(sessionId: String, data: Data) |
|||
func onSessionFinished(sessionId: String, finalText: String, finalAudio: Data) |
|||
func onSessionError(sessionId: String, code: Int, message: String) |
|||
} |
|||
|
|||
/** |
|||
* 客户端配置 |
|||
*/ |
|||
struct Config { |
|||
var wsUrl: String = "wss://openspeech.bytedance.com/api/v4/ast/v2/translate" |
|||
var appKey: String = "" |
|||
var accessKey: String = "" |
|||
var resourceId: String = "volc.service_type.10053" |
|||
var sourceLanguage: String = "zh" |
|||
var targetLanguage: String = "en" |
|||
var clientWaitMs: Int64 = 60_000 |
|||
} |
|||
|
|||
private let tag = "DoubaoE2ETranslateHelper" |
|||
private let log = OSLog(subsystem: "com.azure.speech", category: "DoubaoE2E") |
|||
|
|||
private var conf = Config() |
|||
private weak var callback: Callback? |
|||
|
|||
private var urlSession: URLSession? |
|||
private var webSocket: URLSessionWebSocketTask? |
|||
private var sessionId: String = "" |
|||
|
|||
private var recvAudio = Data() |
|||
private var recvText: [String] = [] |
|||
private var recvSourceText: [String] = [] |
|||
|
|||
private var audioChunkBuffer = Data() |
|||
private var isStarted = false |
|||
|
|||
private let lock = NSLock() |
|||
private var isWebSocketConnected = false |
|||
private var pendingAudioData: [Data] = [] |
|||
private var keepAliveTimer: Timer? |
|||
private var lastAudioSentAt = Date.distantPast |
|||
private let keepAliveIntervalSec: TimeInterval = 1.0 |
|||
private let keepAliveIdleThresholdSec: TimeInterval = 2.5 |
|||
private let keepAliveSilentChunk = Data(repeating: 0, count: 640) |
|||
private var hasSentInitialWarmupPacket = false |
|||
|
|||
/** |
|||
* 初始化助手（异步） |
|||
* 返回是否成功触发了连接逻辑 |
|||
*/ |
|||
func initialize(config: Config, cb: Callback) async -> Bool { |
|||
// 清理旧的 Session |
|||
if let session = urlSession { |
|||
session.invalidateAndCancel() |
|||
} |
|||
|
|||
os_log("Initialize with wsUrl: %{public}@", log: log, type: .info, config.wsUrl) |
|||
conf = config |
|||
// 去除可能的空白字符 |
|||
conf.appKey = conf.appKey.trimmingCharacters(in: .whitespacesAndNewlines) |
|||
conf.accessKey = conf.accessKey.trimmingCharacters(in: .whitespacesAndNewlines) |
|||
conf.resourceId = conf.resourceId.trimmingCharacters(in: .whitespacesAndNewlines) |
|||
|
|||
callback = cb |
|||
let configuration = URLSessionConfiguration.default |
|||
configuration.timeoutIntervalForRequest = 0 // 恢复为0（无超时），因为 WebSocket 是长连接 |
|||
configuration.timeoutIntervalForResource = 0 |
|||
// 使用主队列，方便调试且避免并发问题 |
|||
urlSession = URLSession(configuration: configuration, delegate: self, delegateQueue: OperationQueue.main) |
|||
|
|||
// 等待连接建立（简单的延迟模拟，或者通过回调通知） |
|||
// 这里仅启动连接，实际连接结果是异步的 |
|||
return startContinuousTranslation() |
|||
} |
|||
|
|||
/** |
|||
* 重启会话（用于超时自动重连） |
|||
*/ |
|||
private func restartSession() { |
|||
os_log("Restarting session due to timeout...", log: log, type: .info) |
|||
stopKeepAliveTimer() |
|||
hasSentInitialWarmupPacket = false |
|||
|
|||
webSocket?.cancel(with: .normalClosure, reason: "restarting".data(using: .utf8)) |
|||
webSocket = nil |
|||
isStarted = false |
|||
lock.lock() |
|||
isWebSocketConnected = false |
|||
lock.unlock() |
|||
|
|||
DispatchQueue.main.asyncAfter(deadline: .now() + 0.2) { [weak self] in |
|||
guard let self = self else { return } |
|||
if !self.isStarted { |
|||
_ = self.startContinuousTranslation() |
|||
} |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 启动会话并建立 WebSocket 连接 |
|||
*/ |
|||
func startContinuousTranslation() -> Bool { |
|||
guard !isStarted else { |
|||
os_log("Already started, ignoring start request", log: log, type: .info) |
|||
return true |
|||
} |
|||
guard let url = URL(string: conf.wsUrl) else { |
|||
os_log("Invalid WebSocket URL: %{public}@", log: log, type: .error, conf.wsUrl) |
|||
return false |
|||
} |
|||
|
|||
sessionId = UUID().uuidString |
|||
os_log("Starting new session: %{public}@", log: log, type: .info, sessionId) |
|||
|
|||
recvAudio.removeAll() |
|||
recvText.removeAll() |
|||
recvSourceText.removeAll() |
|||
hasSentInitialWarmupPacket = false |
|||
|
|||
lock.lock() |
|||
isWebSocketConnected = false |
|||
pendingAudioData.removeAll() |
|||
lock.unlock() |
|||
|
|||
var request = URLRequest(url: url) |
|||
request.timeoutInterval = 60 // 握手超时可以设置，但连接成功后不应受此限制 |
|||
|
|||
// 打印 Header 信息（部分脱敏） |
|||
let appKeyMasked = conf.appKey.count > 4 ? String(conf.appKey.prefix(4)) + "***" : conf.appKey |
|||
let accessKeyMasked = conf.accessKey.count > 4 ? String(conf.accessKey.prefix(4)) + "***" : "N/A" |
|||
os_log("Request Headers: AppKey=%{public}@, ResourceId=%{public}@", log: log, type: .info, appKeyMasked, conf.resourceId) |
|||
|
|||
request.addValue(conf.appKey, forHTTPHeaderField: "X-Api-App-Key") |
|||
request.addValue(conf.accessKey, forHTTPHeaderField: "X-Api-Access-Key") |
|||
request.addValue(conf.resourceId, forHTTPHeaderField: "X-Api-Resource-Id") |
|||
request.addValue(UUID().uuidString, forHTTPHeaderField: "X-Api-Connect-Id") |
|||
|
|||
// 尝试添加 Protocol 头 |
|||
// request.addValue("ast-v2", forHTTPHeaderField: "Sec-WebSocket-Protocol") |
|||
|
|||
if let session = urlSession { |
|||
webSocket = session.webSocketTask(with: request) |
|||
} else { |
|||
os_log("Failed to create WebSocket task: urlSession is nil", log: log, type: .error) |
|||
return false |
|||
} |
|||
webSocket?.resume() |
|||
isStarted = true |
|||
return true |
|||
} |
|||
|
|||
/** |
|||
* 推送音频数据（PCM/WAV） |
|||
*/ |
|||
func pushAudioData(_ data: Data) -> Bool { |
|||
// 先检查 isStarted，避免未启动就调用 |
|||
guard isStarted else { |
|||
os_log("Push audio failed: not started", log: log, type: .error) |
|||
return false |
|||
} |
|||
|
|||
// 尝试获取 webSocket，如果为空，说明还在连接中，应该进入缓冲逻辑 |
|||
// 如果已经连接成功但 webSocket 还是 nil，那就是异常状态 |
|||
|
|||
lock.lock() |
|||
// 1. 如果已连接且有 ws 实例 -> 直接发 |
|||
if isWebSocketConnected, let ws = webSocket { |
|||
lock.unlock() |
|||
return sendAudioData(ws, data: data) |
|||
} |
|||
// 2. 否则（连接中、ws 为空、或者刚断开）-> 缓冲 |
|||
else { |
|||
// os_log("Buffering audio data, size: %d", log: log, type: .debug, data.count) |
|||
pendingAudioData.append(data) |
|||
lock.unlock() |
|||
return true |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 发送音频分片（Protobuf） |
|||
*/ |
|||
private func sendAudioData(_ ws: URLSessionWebSocketTask, data: Data) -> Bool { |
|||
if data.isEmpty { return true } |
|||
let req = makeChunkRequest(sessionId: sessionId, chunk: data) |
|||
guard let msg = try? req.serializedData() else { return false } |
|||
ws.send(.data(msg)) { [weak self] error in |
|||
if let e = error { |
|||
os_log("发送音频失败: %{public}@", log: self?.log ?? .default, type: .error, e.localizedDescription) |
|||
} else { |
|||
self?.markAudioSentNow() |
|||
} |
|||
} |
|||
return true |
|||
} |
|||
|
|||
/** |
|||
* 停止会话并发送结束请求（Protobuf） |
|||
*/ |
|||
func stopContinuousTranslation() -> Bool { |
|||
os_log("Stopping continuous translation", log: log, type: .info) |
|||
guard let ws = webSocket else { return false } |
|||
let req = makeFinishRequest(sessionId: sessionId) |
|||
guard let msg = try? req.serializedData() else { return false } |
|||
ws.send(.data(msg)) { _ in } |
|||
return true |
|||
} |
|||
|
|||
/** |
|||
* 释放资源 |
|||
*/ |
|||
func dispose() { |
|||
os_log("Disposing resources", log: log, type: .info) |
|||
stopKeepAliveTimer() |
|||
do { try webSocket?.cancel(with: .normalClosure, reason: "dispose".data(using: .utf8)) } catch { } |
|||
webSocket = nil |
|||
isStarted = false |
|||
lock.lock() |
|||
isWebSocketConnected = false |
|||
pendingAudioData.removeAll() |
|||
lock.unlock() |
|||
urlSession?.invalidateAndCancel() |
|||
urlSession = nil |
|||
} |
|||
|
|||
/** |
|||
* WebSocket打开回调（发送 Protobuf StartSession） |
|||
*/ |
|||
func urlSession(_ session: URLSession, webSocketTask: URLSessionWebSocketTask, didOpenWithProtocol protocol: String?) { |
|||
os_log("WebSocket didOpen", log: log, type: .info) |
|||
markAudioSentNow() |
|||
startKeepAliveTimer() |
|||
let startReq = makeStartRequest(sessionId: sessionId) |
|||
if let data = try? startReq.serializedData() { |
|||
os_log("Sending StartSession (protobuf), size=%{public}d", log: log, type: .info, data.count) |
|||
let message = URLSessionWebSocketTask.Message.data(data) |
|||
webSocketTask.send(message) { [weak self] error in |
|||
if let e = error { |
|||
os_log("StartSession 发送失败: %{public}@", log: self?.log ?? .default, type: .error, e.localizedDescription) |
|||
} else { |
|||
os_log("StartSession sent successfully", log: self?.log ?? .default, type: .info) |
|||
self?.sendInitialWarmupPacketIfNeeded(ws: webSocketTask) |
|||
} |
|||
} |
|||
} |
|||
|
|||
lock.lock() |
|||
isWebSocketConnected = true |
|||
let pending = pendingAudioData |
|||
pendingAudioData.removeAll() |
|||
lock.unlock() |
|||
|
|||
if !pending.isEmpty { |
|||
os_log("Flushing %d buffered audio chunks", log: log, type: .info, pending.count) |
|||
} |
|||
|
|||
for data in pending { |
|||
_ = sendAudioData(webSocketTask, data: data) |
|||
} |
|||
|
|||
callback?.onSessionStarted(sessionId: sessionId) |
|||
receiveLoop() |
|||
} |
|||
|
|||
/** |
|||
* WebSocket关闭回调 |
|||
*/ |
|||
func urlSession(_ session: URLSession, webSocketTask: URLSessionWebSocketTask, didCloseWith closeCode: URLSessionWebSocketTask.CloseCode, reason: Data?) { |
|||
let reasonStr = String(data: reason ?? Data(), encoding: .utf8) ?? "" |
|||
os_log("WebSocket didClose, code: %d, reason: %{public}@", log: log, type: .info, closeCode.rawValue, reasonStr) |
|||
stopKeepAliveTimer() |
|||
lock.lock() |
|||
isWebSocketConnected = false |
|||
lock.unlock() |
|||
isStarted = false |
|||
} |
|||
|
|||
/** |
|||
* 任务完成回调（处理错误） |
|||
*/ |
|||
func urlSession(_ session: URLSession, task: URLSessionTask, didCompleteWithError error: Error?) { |
|||
if let e = error { |
|||
os_log("WebSocket task failed: %{public}@", log: log, type: .error, e.localizedDescription) |
|||
callback?.onSessionError(sessionId: sessionId, code: 1012, message: e.localizedDescription) |
|||
stopKeepAliveTimer() |
|||
|
|||
lock.lock() |
|||
isWebSocketConnected = false |
|||
lock.unlock() |
|||
isStarted = false |
|||
} |
|||
} |
|||
|
|||
private func startKeepAliveTimer() { |
|||
stopKeepAliveTimer() |
|||
keepAliveTimer = Timer.scheduledTimer(withTimeInterval: keepAliveIntervalSec, repeats: true) { [weak self] _ in |
|||
self?.sendKeepAliveIfNeeded() |
|||
} |
|||
if let timer = keepAliveTimer { |
|||
RunLoop.main.add(timer, forMode: .common) |
|||
} |
|||
} |
|||
|
|||
private func stopKeepAliveTimer() { |
|||
keepAliveTimer?.invalidate() |
|||
keepAliveTimer = nil |
|||
} |
|||
|
|||
private func markAudioSentNow() { |
|||
lastAudioSentAt = Date() |
|||
} |
|||
|
|||
private func sendKeepAliveIfNeeded() { |
|||
lock.lock() |
|||
let canSend = isStarted && isWebSocketConnected |
|||
let ws = webSocket |
|||
lock.unlock() |
|||
guard canSend, let ws = ws else { return } |
|||
|
|||
let idle = Date().timeIntervalSince(lastAudioSentAt) |
|||
if idle < keepAliveIdleThresholdSec { return } |
|||
|
|||
os_log("Sending keepalive silent chunk, idle=%.2fs", log: log, type: .debug, idle) |
|||
_ = sendAudioData(ws, data: keepAliveSilentChunk) |
|||
} |
|||
|
|||
private func sendInitialWarmupPacketIfNeeded(ws: URLSessionWebSocketTask) { |
|||
guard !hasSentInitialWarmupPacket else { return } |
|||
hasSentInitialWarmupPacket = true |
|||
os_log("Sending initial warmup packet to avoid first-packet timeout", log: log, type: .info) |
|||
_ = sendAudioData(ws, data: keepAliveSilentChunk) |
|||
} |
|||
|
|||
/** |
|||
* 接收消息循环 |
|||
*/ |
|||
private func receiveLoop() { |
|||
webSocket?.receive { [weak self] result in |
|||
guard let self = self else { return } |
|||
switch result { |
|||
case .failure(let error): |
|||
self.callback?.onSessionError(sessionId: self.sessionId, code: 1011, message: error.localizedDescription) |
|||
case .success(let message): |
|||
self.handleMessage(message) |
|||
self.receiveLoop() |
|||
} |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 处理服务端消息（优先按 Protobuf 解码） |
|||
*/ |
|||
private func handleMessage(_ message: URLSessionWebSocketTask.Message) { |
|||
switch message { |
|||
case .data(let data): |
|||
handleProtoMessage(data) |
|||
case .string(let text): |
|||
handleTextMessage(text) |
|||
@unknown default: |
|||
break |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 处理二进制消息（Protobuf TranslateResponse） |
|||
*/ |
|||
private func handleProtoMessage(_ data: Data) { |
|||
do { |
|||
let resp = try Data_Speech_Ast_TranslateResponse(serializedData: data) |
|||
handleProtoResponse(resp) |
|||
} catch { |
|||
os_log("Proto parse failed, len=%{public}d error=%{public}@", log: log, type: .error, data.count, error.localizedDescription) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 按事件类型处理 Protobuf TranslateResponse |
|||
*/ |
|||
private func handleProtoResponse(_ resp: Data_Speech_Ast_TranslateResponse) { |
|||
let event = resp.event |
|||
let text = resp.text |
|||
let audioData = resp.data |
|||
|
|||
if event != .usageResponse { |
|||
os_log("Handling proto event: %{public}d", log: log, type: .info, event.rawValue) |
|||
os_log("Proto payload: textLen=%{public}d audioLen=%{public}d", |
|||
log: log, type: .info, text.count, audioData.count) |
|||
} |
|||
|
|||
if event == .usageResponse { |
|||
return |
|||
} |
|||
|
|||
if event == .sessionFailed || event == .sessionCanceled { |
|||
let message = resp.hasResponseMeta ? resp.responseMeta.message : "" |
|||
os_log("Session error event: %{public}d, message: %{public}@", log: log, type: .error, event.rawValue, message) |
|||
|
|||
if message.contains("Timeout waiting next packet") { |
|||
os_log("Detected timeout error, triggering auto-restart", log: log, type: .info) |
|||
restartSession() |
|||
return |
|||
} |
|||
|
|||
callback?.onSessionError(sessionId: sessionId, code: 1011, message: message) |
|||
webSocket?.cancel(with: .goingAway, reason: message.data(using: .utf8)) |
|||
return |
|||
} |
|||
|
|||
if event == .sessionFinished { |
|||
let finalText = recvText.joined(separator: " ") |
|||
let finalAudio = recvAudio |
|||
os_log("SessionFinished: finalTextLen=%{public}d finalAudioLen=%{public}d", |
|||
log: log, type: .info, finalText.count, finalAudio.count) |
|||
callback?.onSessionFinished(sessionId: sessionId, finalText: finalText, finalAudio: finalAudio) |
|||
webSocket?.cancel(with: .normalClosure, reason: "Completed".data(using: .utf8)) |
|||
return |
|||
} |
|||
|
|||
if !audioData.isEmpty { |
|||
recvAudio.append(audioData) |
|||
os_log("Append audio chunk: chunkSize=%{public}d totalSize=%{public}d", |
|||
log: log, type: .info, audioData.count, recvAudio.count) |
|||
processAudioChunk(incoming: audioData, multiple: 1280) |
|||
} |
|||
|
|||
if !text.isEmpty { |
|||
switch event { |
|||
case .sourceSubtitleStart: |
|||
os_log("SourceSubtitleStart", log: log, type: .info) |
|||
recvSourceText.removeAll() |
|||
case .sourceSubtitleResponse: |
|||
os_log("SourceSubtitleResponse text=%{public}@", log: log, type: .info, text) |
|||
recvSourceText.append(text) |
|||
callback?.onPartialSourceText(sessionId: sessionId, text: text) |
|||
case .sourceSubtitleEnd: |
|||
os_log("SourceSubtitleEnd, finalSrcLen=%{public}d", |
|||
log: log, type: .info, recvSourceText.joined(separator: " ").count) |
|||
callback?.onFinalSourceText(sessionId: sessionId, finalText: recvSourceText.joined(separator: " ")) |
|||
case .translationSubtitleStart: |
|||
os_log("TranslationSubtitleStart", log: log, type: .info) |
|||
recvText.removeAll() |
|||
case .translationSubtitleResponse: |
|||
os_log("TranslationSubtitleResponse text=%{public}@", log: log, type: .info, text) |
|||
recvText.append(text) |
|||
callback?.onPartialText(sessionId: sessionId, text: text) |
|||
case .translationSubtitleEnd: |
|||
os_log("TranslationSubtitleEnd, finalTgtLen=%{public}d", |
|||
log: log, type: .info, recvText.joined(separator: " ").count) |
|||
callback?.onFinalTranslatedText(sessionId: sessionId, finalText: recvText.joined(separator: " ")) |
|||
startNewSubSession() |
|||
default: |
|||
os_log("Other text event=%{public}d text=%{public}@", log: log, type: .info, event.rawValue, text) |
|||
// recvText.append(text) |
|||
// callback?.onPartialText(sessionId: sessionId, text: text) |
|||
} |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 处理文本消息（保留 JSON 兜底逻辑） |
|||
*/ |
|||
private func handleTextMessage(_ text: String) { |
|||
guard let data = text.data(using: .utf8), |
|||
let obj = try? JSONSerialization.jsonObject(with: data) as? [String: Any] else { return } |
|||
let event = obj["event"] as? String ?? "" |
|||
let msgText = obj["text"] as? String |
|||
|
|||
if event != "UsageResponse" { |
|||
os_log("Handling text event: %{public}@", log: log, type: .info, event) |
|||
} |
|||
|
|||
if event == "UsageResponse" { return } |
|||
if event == "SessionFailed" || event == "SessionCanceled" { |
|||
let meta = obj["responseMeta"] as? [String: Any] |
|||
let message = meta?["message"] as? String ?? "" |
|||
os_log("Session error event(text): %{public}@, message: %{public}@", log: log, type: .error, event, message) |
|||
|
|||
if message.contains("Timeout waiting next packet") { |
|||
os_log("Detected timeout error (text), triggering auto-restart", log: log, type: .info) |
|||
restartSession() |
|||
return |
|||
} |
|||
|
|||
callback?.onSessionError(sessionId: sessionId, code: 1011, message: message) |
|||
webSocket?.cancel(with: .goingAway, reason: message.data(using: .utf8)) |
|||
return |
|||
} |
|||
if event == "SessionFinished" { |
|||
let finalText = recvText.joined(separator: " ") |
|||
let finalAudio = recvAudio |
|||
callback?.onSessionFinished(sessionId: sessionId, finalText: finalText, finalAudio: finalAudio) |
|||
webSocket?.cancel(with: .normalClosure, reason: "Completed".data(using: .utf8)) |
|||
return |
|||
} |
|||
if let audioB64 = (obj["data"] as? String), let audioData = Data(base64Encoded: audioB64) { |
|||
recvAudio.append(audioData) |
|||
processAudioChunk(incoming: audioData, multiple: 1280) |
|||
} |
|||
if let t = msgText, !t.isEmpty { |
|||
switch event { |
|||
case "SourceSubtitleStart": |
|||
recvSourceText.removeAll() |
|||
case "SourceSubtitleResponse": |
|||
recvSourceText.append(t) |
|||
callback?.onPartialSourceText(sessionId: sessionId, text: t) |
|||
case "SourceSubtitleEnd": |
|||
callback?.onFinalSourceText(sessionId: sessionId, finalText: recvSourceText.joined(separator: " ")) |
|||
case "TranslationSubtitleStart": |
|||
recvText.removeAll() |
|||
case "TranslationSubtitleResponse": |
|||
recvText.append(t) |
|||
callback?.onPartialText(sessionId: sessionId, text: t) |
|||
case "TranslationSubtitleEnd": |
|||
callback?.onFinalTranslatedText(sessionId: sessionId, finalText: recvText.joined(separator: " ")) |
|||
startNewSubSession() |
|||
default: |
|||
break |
|||
// recvText.append(t) |
|||
// callback?.onPartialText(sessionId: sessionId, text: t) |
|||
} |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 新开子会话（句末切分） |
|||
*/ |
|||
private func startNewSubSession() { |
|||
sessionId = UUID().uuidString |
|||
recvText.removeAll() |
|||
recvSourceText.removeAll() |
|||
} |
|||
|
|||
/** |
|||
* 分片处理音频并按 1280 字节回调 |
|||
*/ |
|||
private func processAudioChunk(incoming: Data, multiple: Int) { |
|||
if !incoming.isEmpty { audioChunkBuffer.append(incoming) } |
|||
let sendLen = (audioChunkBuffer.count / multiple) * multiple |
|||
if sendLen > 0 { |
|||
let chunk = audioChunkBuffer.prefix(sendLen) |
|||
callback?.onPartialAudio(sessionId: sessionId, data: Data(chunk)) |
|||
audioChunkBuffer = Data(audioChunkBuffer.dropFirst(sendLen)) |
|||
} |
|||
} |
|||
|
|||
/** |
|||
* 构造 StartSession 请求（Protobuf） |
|||
*/ |
|||
private func makeStartRequest(sessionId: String) -> Data_Speech_Ast_TranslateRequest { |
|||
var user = Data_Speech_Understanding_User() |
|||
user.uid = "ast_ios_client" |
|||
user.did = "ast_ios_client" |
|||
user.platform = "ios" |
|||
|
|||
var sourceAudio = Data_Speech_Understanding_Audio() |
|||
sourceAudio.format = "wav" |
|||
sourceAudio.rate = 16000 |
|||
sourceAudio.bits = 16 |
|||
sourceAudio.channel = 1 |
|||
|
|||
var targetAudio = Data_Speech_Understanding_Audio() |
|||
targetAudio.format = "pcm" |
|||
targetAudio.rate = 16000 |
|||
|
|||
var params = Data_Speech_Ast_ReqParams() |
|||
params.mode = "s2s" |
|||
params.sourceLanguage = conf.sourceLanguage |
|||
params.targetLanguage = conf.targetLanguage |
|||
|
|||
var meta = Data_Speech_Common_RequestMeta() |
|||
meta.sessionID = sessionId |
|||
|
|||
var req = Data_Speech_Ast_TranslateRequest() |
|||
req.requestMeta = meta |
|||
req.event = .startSession |
|||
req.user = user |
|||
req.sourceAudio = sourceAudio |
|||
req.targetAudio = targetAudio |
|||
req.request = params |
|||
return req |
|||
} |
|||
|
|||
/** |
|||
* 构造音频分片请求（Protobuf） |
|||
*/ |
|||
private func makeChunkRequest(sessionId: String, chunk: Data) -> Data_Speech_Ast_TranslateRequest { |
|||
var sourceAudio = Data_Speech_Understanding_Audio() |
|||
sourceAudio.format = "wav" |
|||
sourceAudio.rate = 16000 |
|||
sourceAudio.bits = 16 |
|||
sourceAudio.channel = 1 |
|||
sourceAudio.binaryData = chunk |
|||
|
|||
var meta = Data_Speech_Common_RequestMeta() |
|||
meta.sessionID = sessionId |
|||
|
|||
var req = Data_Speech_Ast_TranslateRequest() |
|||
req.requestMeta = meta |
|||
req.event = .taskRequest |
|||
var user = Data_Speech_Understanding_User() |
|||
user.uid = "ast_ios_client" |
|||
user.did = "ast_ios_client" |
|||
req.user = user |
|||
req.sourceAudio = sourceAudio |
|||
return req |
|||
} |
|||
|
|||
/** |
|||
* 构造 FinishSession 请求（Protobuf） |
|||
*/ |
|||
private func makeFinishRequest(sessionId: String) -> Data_Speech_Ast_TranslateRequest { |
|||
var meta = Data_Speech_Common_RequestMeta() |
|||
meta.sessionID = sessionId |
|||
|
|||
var user = Data_Speech_Understanding_User() |
|||
user.uid = "ast_ios_client" |
|||
user.did = "ast_ios_client" |
|||
|
|||
var sourceAudio = Data_Speech_Understanding_Audio() |
|||
|
|||
var req = Data_Speech_Ast_TranslateRequest() |
|||
req.requestMeta = meta |
|||
req.event = .finishSession |
|||
req.user = user |
|||
req.sourceAudio = sourceAudio |
|||
return req |
|||
} |
|||
|
|||
|
|||
} |
|||
@ -0,0 +1,301 @@ |
|||
// DO NOT EDIT. |
|||
// swift-format-ignore-file |
|||
// swiftlint:disable all |
|||
// |
|||
// Generated by the Swift generator plugin for the protocol buffer compiler. |
|||
// Source: common/events.proto |
|||
// |
|||
// For information on using the generated types, please see the documentation: |
|||
// https://github.com/apple/swift-protobuf/ |
|||
|
|||
import SwiftProtobuf |
|||
|
|||
// If the compiler emits an error on this type, it is because this file |
|||
// was generated by a version of the `protoc` Swift plug-in that is |
|||
// incompatible with the version of SwiftProtobuf to which you are linking. |
|||
// Please ensure that you are building against the same version of the API |
|||
// that was used to generate this file. |
|||
fileprivate struct _GeneratedWithProtocGenSwiftVersion: SwiftProtobuf.ProtobufAPIVersionCheck { |
|||
struct _2: SwiftProtobuf.ProtobufAPIVersion_2 {} |
|||
typealias Version = _2 |
|||
} |
|||
|
|||
enum Data_Speech_Event_Type: SwiftProtobuf.Enum, Swift.CaseIterable { |
|||
typealias RawValue = Int |
|||
|
|||
/// 默认事件，适用于不使用事件的方案或不需要传递事件的情况， |
|||
/// 或者对于使用事件的方案，可以通过非0值来校验事件的合法性 |
|||
case none // = 0 |
|||
|
|||
/// 1 ~ 49 为上行Connection事件 |
|||
case startConnection // = 1 |
|||
|
|||
/// Alias of "StartConnection" |
|||
static let startTask = startConnection |
|||
case finishConnection // = 2 |
|||
|
|||
/// Alias of "FinishConnection" |
|||
static let finishTask = finishConnection |
|||
|
|||
/// 50 ~ 99 为下行Connection事件 |
|||
/// 成功建连 |
|||
case connectionStarted // = 50 |
|||
|
|||
/// Alias of "ConnectionStarted" |
|||
static let taskStarted = connectionStarted |
|||
|
|||
/// 建连失败（可能是无法通过权限认证） |
|||
case connectionFailed // = 51 |
|||
|
|||
/// Alias of "ConnectionFailed" |
|||
static let taskFailed = connectionFailed |
|||
|
|||
/// 连接结束 |
|||
case connectionFinished // = 52 |
|||
|
|||
/// Alias of "ConnectionFinished" |
|||
static let taskFinished = connectionFinished |
|||
|
|||
/// 100 ~ 149 为上行Session事件 |
|||
case startSession // = 100 |
|||
case cancelSession // = 101 |
|||
case finishSession // = 102 |
|||
|
|||
/// 150 ～ 199 为下行Session事件 |
|||
case sessionStarted // = 150 |
|||
case sessionCanceled // = 151 |
|||
case sessionFinished // = 152 |
|||
case sessionFailed // = 153 |
|||
|
|||
/// 用量事件 |
|||
case usageResponse // = 154 |
|||
|
|||
/// Alias of "UsageResponse" |
|||
static let chargeData = usageResponse |
|||
|
|||
/// 200 ~ 249 为上行通用事件 |
|||
case taskRequest // = 200 |
|||
case updateConfig // = 201 |
|||
case imageRequest // = 202 |
|||
|
|||
/// 250 ~ 299 为下行通用事件 |
|||
case audioMuted // = 250 |
|||
|
|||
/// 300 ~ 349 为上行TTS事件 |
|||
case sayHello // = 300 |
|||
|
|||
/// 350 ~ 399 为下行TTS事件 |
|||
case ttssentenceStart // = 350 |
|||
case ttssentenceEnd // = 351 |
|||
case ttsresponse // = 352 |
|||
case ttsended // = 359 |
|||
case podcastRoundStart // = 360 |
|||
case podcastRoundResponse // = 361 |
|||
case podcastRoundEnd // = 362 |
|||
|
|||
/// 450 ~ 499 为下行ASR事件 |
|||
case asrinfo // = 450 |
|||
case asrresponse // = 451 |
|||
case asrended // = 459 |
|||
|
|||
/// 500 ~ 549 为上行对话事件 |
|||
/// (Ground-Truth-Alignment) text for speech synthesis |
|||
case chatTtstext // = 500 |
|||
case chatTextQuery // = 501 |
|||
|
|||
/// 550 ~ 599 为下行对话事件 |
|||
case chatResponse // = 550 |
|||
case chimeInStart // = 551 |
|||
case chimeInEnd // = 552 |
|||
case chatEnded // = 559 |
|||
case thinkStart // = 560 |
|||
case thinkResponse // = 561 |
|||
case thinkEnd // = 562 |
|||
case toolOutput // = 563 |
|||
case fcresponseStart // = 564 |
|||
case fcresponse // = 565 |
|||
case fcresponseEnd // = 566 |
|||
|
|||
/// 650 ~ 699 为下行对话事件 |
|||
/// Events for source (original) language subtitle. |
|||
case sourceSubtitleStart // = 650 |
|||
case sourceSubtitleResponse // = 651 |
|||
case sourceSubtitleEnd // = 652 |
|||
|
|||
/// Events for target (translation) language subtitle. |
|||
case translationSubtitleStart // = 653 |
|||
case translationSubtitleResponse // = 654 |
|||
case translationSubtitleEnd // = 655 |
|||
case UNRECOGNIZED(Int) |
|||
|
|||
init() { |
|||
self = .none |
|||
} |
|||
|
|||
init?(rawValue: Int) { |
|||
switch rawValue { |
|||
case 0: self = .none |
|||
case 1: self = .startConnection |
|||
case 2: self = .finishConnection |
|||
case 50: self = .connectionStarted |
|||
case 51: self = .connectionFailed |
|||
case 52: self = .connectionFinished |
|||
case 100: self = .startSession |
|||
case 101: self = .cancelSession |
|||
case 102: self = .finishSession |
|||
case 150: self = .sessionStarted |
|||
case 151: self = .sessionCanceled |
|||
case 152: self = .sessionFinished |
|||
case 153: self = .sessionFailed |
|||
case 154: self = .usageResponse |
|||
case 200: self = .taskRequest |
|||
case 201: self = .updateConfig |
|||
case 202: self = .imageRequest |
|||
case 250: self = .audioMuted |
|||
case 300: self = .sayHello |
|||
case 350: self = .ttssentenceStart |
|||
case 351: self = .ttssentenceEnd |
|||
case 352: self = .ttsresponse |
|||
case 359: self = .ttsended |
|||
case 360: self = .podcastRoundStart |
|||
case 361: self = .podcastRoundResponse |
|||
case 362: self = .podcastRoundEnd |
|||
case 450: self = .asrinfo |
|||
case 451: self = .asrresponse |
|||
case 459: self = .asrended |
|||
case 500: self = .chatTtstext |
|||
case 501: self = .chatTextQuery |
|||
case 550: self = .chatResponse |
|||
case 551: self = .chimeInStart |
|||
case 552: self = .chimeInEnd |
|||
case 559: self = .chatEnded |
|||
case 560: self = .thinkStart |
|||
case 561: self = .thinkResponse |
|||
case 562: self = .thinkEnd |
|||
case 563: self = .toolOutput |
|||
case 564: self = .fcresponseStart |
|||
case 565: self = .fcresponse |
|||
case 566: self = .fcresponseEnd |
|||
case 650: self = .sourceSubtitleStart |
|||
case 651: self = .sourceSubtitleResponse |
|||
case 652: self = .sourceSubtitleEnd |
|||
case 653: self = .translationSubtitleStart |
|||
case 654: self = .translationSubtitleResponse |
|||
case 655: self = .translationSubtitleEnd |
|||
default: self = .UNRECOGNIZED(rawValue) |
|||
} |
|||
} |
|||
|
|||
var rawValue: Int { |
|||
switch self { |
|||
case .none: return 0 |
|||
case .startConnection: return 1 |
|||
case .finishConnection: return 2 |
|||
case .connectionStarted: return 50 |
|||
case .connectionFailed: return 51 |
|||
case .connectionFinished: return 52 |
|||
case .startSession: return 100 |
|||
case .cancelSession: return 101 |
|||
case .finishSession: return 102 |
|||
case .sessionStarted: return 150 |
|||
case .sessionCanceled: return 151 |
|||
case .sessionFinished: return 152 |
|||
case .sessionFailed: return 153 |
|||
case .usageResponse: return 154 |
|||
case .taskRequest: return 200 |
|||
case .updateConfig: return 201 |
|||
case .imageRequest: return 202 |
|||
case .audioMuted: return 250 |
|||
case .sayHello: return 300 |
|||
case .ttssentenceStart: return 350 |
|||
case .ttssentenceEnd: return 351 |
|||
case .ttsresponse: return 352 |
|||
case .ttsended: return 359 |
|||
case .podcastRoundStart: return 360 |
|||
case .podcastRoundResponse: return 361 |
|||
case .podcastRoundEnd: return 362 |
|||
case .asrinfo: return 450 |
|||
case .asrresponse: return 451 |
|||
case .asrended: return 459 |
|||
case .chatTtstext: return 500 |
|||
case .chatTextQuery: return 501 |
|||
case .chatResponse: return 550 |
|||
case .chimeInStart: return 551 |
|||
case .chimeInEnd: return 552 |
|||
case .chatEnded: return 559 |
|||
case .thinkStart: return 560 |
|||
case .thinkResponse: return 561 |
|||
case .thinkEnd: return 562 |
|||
case .toolOutput: return 563 |
|||
case .fcresponseStart: return 564 |
|||
case .fcresponse: return 565 |
|||
case .fcresponseEnd: return 566 |
|||
case .sourceSubtitleStart: return 650 |
|||
case .sourceSubtitleResponse: return 651 |
|||
case .sourceSubtitleEnd: return 652 |
|||
case .translationSubtitleStart: return 653 |
|||
case .translationSubtitleResponse: return 654 |
|||
case .translationSubtitleEnd: return 655 |
|||
case .UNRECOGNIZED(let i): return i |
|||
} |
|||
} |
|||
|
|||
// The compiler won't synthesize support with the UNRECOGNIZED case. |
|||
static let allCases: [Data_Speech_Event_Type] = [ |
|||
.none, |
|||
.startConnection, |
|||
.finishConnection, |
|||
.connectionStarted, |
|||
.connectionFailed, |
|||
.connectionFinished, |
|||
.startSession, |
|||
.cancelSession, |
|||
.finishSession, |
|||
.sessionStarted, |
|||
.sessionCanceled, |
|||
.sessionFinished, |
|||
.sessionFailed, |
|||
.usageResponse, |
|||
.taskRequest, |
|||
.updateConfig, |
|||
.imageRequest, |
|||
.audioMuted, |
|||
.sayHello, |
|||
.ttssentenceStart, |
|||
.ttssentenceEnd, |
|||
.ttsresponse, |
|||
.ttsended, |
|||
.podcastRoundStart, |
|||
.podcastRoundResponse, |
|||
.podcastRoundEnd, |
|||
.asrinfo, |
|||
.asrresponse, |
|||
.asrended, |
|||
.chatTtstext, |
|||
.chatTextQuery, |
|||
.chatResponse, |
|||
.chimeInStart, |
|||
.chimeInEnd, |
|||
.chatEnded, |
|||
.thinkStart, |
|||
.thinkResponse, |
|||
.thinkEnd, |
|||
.toolOutput, |
|||
.fcresponseStart, |
|||
.fcresponse, |
|||
.fcresponseEnd, |
|||
.sourceSubtitleStart, |
|||
.sourceSubtitleResponse, |
|||
.sourceSubtitleEnd, |
|||
.translationSubtitleStart, |
|||
.translationSubtitleResponse, |
|||
.translationSubtitleEnd, |
|||
] |
|||
|
|||
} |
|||
|
|||
// MARK: - Code below here is support for the SwiftProtobuf runtime. |
|||
|
|||
extension Data_Speech_Event_Type: SwiftProtobuf._ProtoNameProviding { |
|||
static let _protobuf_nameMap = SwiftProtobuf._NameMap(bytecode: "\0\u{2}\0None\0\u{9}StartConnection\0\u{1}StartTask\0\u{9}FinishConnection\0\u{1}FinishTask\0\u{a}0ConnectionStarted\0\u{1}TaskStarted\0\u{9}ConnectionFailed\0\u{1}TaskFailed\0\u{9}ConnectionFinished\0\u{1}TaskFinished\0\u{2}0StartSession\0\u{1}CancelSession\0\u{1}FinishSession\0\u{2}0SessionStarted\0\u{1}SessionCanceled\0\u{1}SessionFinished\0\u{1}SessionFailed\0\u{9}UsageResponse\0\u{1}ChargeData\0\u{2}.TaskRequest\0\u{1}UpdateConfig\0\u{1}ImageRequest\0\u{2}0AudioMuted\0\u{2}2SayHello\0\u{2}2TTSSentenceStart\0\u{1}TTSSentenceEnd\0\u{1}TTSResponse\0\u{2}\u{7}TTSEnded\0\u{1}PodcastRoundStart\0\u{1}PodcastRoundResponse\0\u{1}PodcastRoundEnd\0\u{2}X\u{1}ASRInfo\0\u{1}ASRResponse\0\u{2}\u{8}ASREnded\0\u{2})ChatTTSText\0\u{1}ChatTextQuery\0\u{2}1ChatResponse\0\u{1}ChimeInStart\0\u{1}ChimeInEnd\0\u{2}\u{7}ChatEnded\0\u{1}ThinkStart\0\u{1}ThinkResponse\0\u{1}ThinkEnd\0\u{1}ToolOutput\0\u{1}FCResponseStart\0\u{1}FCResponse\0\u{1}FCResponseEnd\0\u{2}T\u{1}SourceSubtitleStart\0\u{1}SourceSubtitleResponse\0\u{1}SourceSubtitleEnd\0\u{1}TranslationSubtitleStart\0\u{1}TranslationSubtitleResponse\0\u{1}TranslationSubtitleEnd\0") |
|||
} |
|||
@ -0,0 +1,350 @@ |
|||
// DO NOT EDIT. |
|||
// swift-format-ignore-file |
|||
// swiftlint:disable all |
|||
// |
|||
// Generated by the Swift generator plugin for the protocol buffer compiler. |
|||
// Source: common/rpcmeta.proto |
|||
// |
|||
// For information on using the generated types, please see the documentation: |
|||
// https://github.com/apple/swift-protobuf/ |
|||
|
|||
import SwiftProtobuf |
|||
|
|||
// If the compiler emits an error on this type, it is because this file |
|||
// was generated by a version of the `protoc` Swift plug-in that is |
|||
// incompatible with the version of SwiftProtobuf to which you are linking. |
|||
// Please ensure that you are building against the same version of the API |
|||
// that was used to generate this file. |
|||
fileprivate struct _GeneratedWithProtocGenSwiftVersion: SwiftProtobuf.ProtobufAPIVersionCheck { |
|||
struct _2: SwiftProtobuf.ProtobufAPIVersion_2 {} |
|||
typealias Version = _2 |
|||
} |
|||
|
|||
struct Data_Speech_Common_RequestMeta: Sendable { |
|||
// SwiftProtobuf.Message conformance is added in an extension below. See the |
|||
// `Message` and `Message+*Additions` files in the SwiftProtobuf library for |
|||
// methods supported on all messages. |
|||
|
|||
/// Required. |
|||
/// Backend endpoint name. Gateway will also use it to infer HTTP/WebSocket URL |
|||
/// path. |
|||
var endpoint: String = String() |
|||
|
|||
/// Required. |
|||
/// For historical reasons, this can be either a volcanic appid or a sail |
|||
/// platform appKey. If both are transmitted at the same time, you need to put |
|||
/// the volcengine appid in the AppID |
|||
var appKey: String = String() |
|||
|
|||
/// Optional. |
|||
/// appID in volcengine |
|||
var appID: String = String() |
|||
|
|||
/// - Required by gateway, you must provide a ResourceID on calling gateway. |
|||
/// - Optional for backend services, gateway doesn't need to pass this to |
|||
/// backend. |
|||
var resourceID: String = String() |
|||
|
|||
/// Required for Websocket connection (SDK -> Gateway), optional otherwise. |
|||
/// In websocket communication, a connection can be reused for multiple |
|||
/// simultaneous sessions. Servers use this ID mainly for debugging purposes. |
|||
var connectionID: String = String() |
|||
|
|||
/// Required. Session is the minimum unit that performs a specific task. The |
|||
/// request data in a session can be devided into a number data packets |
|||
/// transmitted (but not necessarily processed by the server) in a sequential |
|||
/// manner. Each data packet can be denoted by a "Sequence" number (although |
|||
/// populating the "Sequence" field is not mandatory). In the simplest case, |
|||
/// all request data in a session is sent as a lump (ie, only one packet). |
|||
var sessionID: String = String() |
|||
|
|||
/// Optional, if not passed, default value 0 will be assumed. For streaming |
|||
/// RPC, this field can be omitted (because the order of sending is preserved |
|||
/// at the receiving side) unless the client needs the acknowledgement of |
|||
/// packet receipt from the server side. |
|||
var sequence: Int32 = 0 |
|||
|
|||
var unknownFields = SwiftProtobuf.UnknownStorage() |
|||
|
|||
init() {} |
|||
} |
|||
|
|||
struct Data_Speech_Common_BillingItem: Sendable { |
|||
// SwiftProtobuf.Message conformance is added in an extension below. See the |
|||
// `Message` and `Message+*Additions` files in the SwiftProtobuf library for |
|||
// methods supported on all messages. |
|||
|
|||
/// Required. |
|||
/// Billing unit, eg: |
|||
/// - minute (billed by the number of minutes used) |
|||
/// - word (billed by the number of words submitted or generated) |
|||
/// - call (billed by the number of RPC/HTTP/Websocket calls) |
|||
var unit: String = String() |
|||
|
|||
/// Optional. |
|||
/// The amount that the consumer consumed counted by `BillingItem.Unit`. We use |
|||
/// a float number because sometimes we want precision better than whole |
|||
/// numbers. |
|||
var quantity: Float = 0 |
|||
|
|||
var unknownFields = SwiftProtobuf.UnknownStorage() |
|||
|
|||
init() {} |
|||
} |
|||
|
|||
struct Data_Speech_Common_Billing: Sendable { |
|||
// SwiftProtobuf.Message conformance is added in an extension below. See the |
|||
// `Message` and `Message+*Additions` files in the SwiftProtobuf library for |
|||
// methods supported on all messages. |
|||
|
|||
/// Optional. |
|||
/// For new billing items, use this field. This is a list because there may be |
|||
/// more than one billing items in one request/session. |
|||
var items: [Data_Speech_Common_BillingItem] = [] |
|||
|
|||
/// Optional. |
|||
/// For commodities that are priced w.r.t (typically audio/video) duration in |
|||
/// milliseconds. |
|||
var durationMsec: Int64 = 0 |
|||
|
|||
/// Optional. |
|||
/// For commodities priced w.r.t text length (number of words/tokens). |
|||
var wordCount: Int64 = 0 |
|||
|
|||
var unknownFields = SwiftProtobuf.UnknownStorage() |
|||
|
|||
init() {} |
|||
} |
|||
|
|||
struct Data_Speech_Common_ResponseMeta: Sendable { |
|||
// SwiftProtobuf.Message conformance is added in an extension below. See the |
|||
// `Message` and `Message+*Additions` files in the SwiftProtobuf library for |
|||
// methods supported on all messages. |
|||
|
|||
/// Required. |
|||
/// The same SessionID in RequestMeta. |
|||
var sessionID: String = String() |
|||
|
|||
/// Optional. |
|||
/// The same sequence in RequestMeta (except that server *may* turn a positive |
|||
/// sequence to its negative counterpart). |
|||
var sequence: Int32 = 0 |
|||
|
|||
/// Optional. |
|||
/// Response status code. |
|||
var statusCode: Int32 = 0 |
|||
|
|||
/// Optional. |
|||
/// Detailed status information. |
|||
var message: String = String() |
|||
|
|||
/// Billing information in case only backend servers can collect this |
|||
/// information. |
|||
var billing: Data_Speech_Common_Billing { |
|||
get {return _billing ?? Data_Speech_Common_Billing()} |
|||
set {_billing = newValue} |
|||
} |
|||
/// Returns true if `billing` has been explicitly set. |
|||
var hasBilling: Bool {return self._billing != nil} |
|||
/// Clears the value of `billing`. Subsequent reads from it will return its default value. |
|||
mutating func clearBilling() {self._billing = nil} |
|||
|
|||
var unknownFields = SwiftProtobuf.UnknownStorage() |
|||
|
|||
init() {} |
|||
|
|||
fileprivate var _billing: Data_Speech_Common_Billing? = nil |
|||
} |
|||
|
|||
// MARK: - Code below here is support for the SwiftProtobuf runtime. |
|||
|
|||
fileprivate let _protobuf_package = "data.speech.common" |
|||
|
|||
extension Data_Speech_Common_RequestMeta: SwiftProtobuf.Message, SwiftProtobuf._MessageImplementationBase, SwiftProtobuf._ProtoNameProviding { |
|||
static let protoMessageName: String = _protobuf_package + ".RequestMeta" |
|||
static let _protobuf_nameMap = SwiftProtobuf._NameMap(bytecode: "\0\u{5}Endpoint\0endpoint\0\u{5}AppKey\0app_key\0\u{5}AppID\0app_id\0\u{5}ResourceID\0resource_id\0\u{5}ConnectionID\0connection_id\0\u{5}SessionID\0session_id\0\u{5}Sequence\0sequence\0") |
|||
|
|||
mutating func decodeMessage<D: SwiftProtobuf.Decoder>(decoder: inout D) throws { |
|||
while let fieldNumber = try decoder.nextFieldNumber() { |
|||
// The use of inline closures is to circumvent an issue where the compiler |
|||
// allocates stack space for every case branch when no optimizations are |
|||
// enabled. https://github.com/apple/swift-protobuf/issues/1034 |
|||
switch fieldNumber { |
|||
case 1: try { try decoder.decodeSingularStringField(value: &self.endpoint) }() |
|||
case 2: try { try decoder.decodeSingularStringField(value: &self.appKey) }() |
|||
case 3: try { try decoder.decodeSingularStringField(value: &self.appID) }() |
|||
case 4: try { try decoder.decodeSingularStringField(value: &self.resourceID) }() |
|||
case 5: try { try decoder.decodeSingularStringField(value: &self.connectionID) }() |
|||
case 6: try { try decoder.decodeSingularStringField(value: &self.sessionID) }() |
|||
case 7: try { try decoder.decodeSingularInt32Field(value: &self.sequence) }() |
|||
default: break |
|||
} |
|||
} |
|||
} |
|||
|
|||
func traverse<V: SwiftProtobuf.Visitor>(visitor: inout V) throws { |
|||
if !self.endpoint.isEmpty { |
|||
try visitor.visitSingularStringField(value: self.endpoint, fieldNumber: 1) |
|||
} |
|||
if !self.appKey.isEmpty { |
|||
try visitor.visitSingularStringField(value: self.appKey, fieldNumber: 2) |
|||
} |
|||
if !self.appID.isEmpty { |
|||
try visitor.visitSingularStringField(value: self.appID, fieldNumber: 3) |
|||
} |
|||
if !self.resourceID.isEmpty { |
|||
try visitor.visitSingularStringField(value: self.resourceID, fieldNumber: 4) |
|||
} |
|||
if !self.connectionID.isEmpty { |
|||
try visitor.visitSingularStringField(value: self.connectionID, fieldNumber: 5) |
|||
} |
|||
if !self.sessionID.isEmpty { |
|||
try visitor.visitSingularStringField(value: self.sessionID, fieldNumber: 6) |
|||
} |
|||
if self.sequence != 0 { |
|||
try visitor.visitSingularInt32Field(value: self.sequence, fieldNumber: 7) |
|||
} |
|||
try unknownFields.traverse(visitor: &visitor) |
|||
} |
|||
|
|||
static func ==(lhs: Data_Speech_Common_RequestMeta, rhs: Data_Speech_Common_RequestMeta) -> Bool { |
|||
if lhs.endpoint != rhs.endpoint {return false} |
|||
if lhs.appKey != rhs.appKey {return false} |
|||
if lhs.appID != rhs.appID {return false} |
|||
if lhs.resourceID != rhs.resourceID {return false} |
|||
if lhs.connectionID != rhs.connectionID {return false} |
|||
if lhs.sessionID != rhs.sessionID {return false} |
|||
if lhs.sequence != rhs.sequence {return false} |
|||
if lhs.unknownFields != rhs.unknownFields {return false} |
|||
return true |
|||
} |
|||
} |
|||
|
|||
extension Data_Speech_Common_BillingItem: SwiftProtobuf.Message, SwiftProtobuf._MessageImplementationBase, SwiftProtobuf._ProtoNameProviding { |
|||
static let protoMessageName: String = _protobuf_package + ".BillingItem" |
|||
static let _protobuf_nameMap = SwiftProtobuf._NameMap(bytecode: "\0\u{5}Unit\0unit\0\u{5}Quantity\0quantity\0") |
|||
|
|||
mutating func decodeMessage<D: SwiftProtobuf.Decoder>(decoder: inout D) throws { |
|||
while let fieldNumber = try decoder.nextFieldNumber() { |
|||
// The use of inline closures is to circumvent an issue where the compiler |
|||
// allocates stack space for every case branch when no optimizations are |
|||
// enabled. https://github.com/apple/swift-protobuf/issues/1034 |
|||
switch fieldNumber { |
|||
case 1: try { try decoder.decodeSingularStringField(value: &self.unit) }() |
|||
case 2: try { try decoder.decodeSingularFloatField(value: &self.quantity) }() |
|||
default: break |
|||
} |
|||
} |
|||
} |
|||
|
|||
func traverse<V: SwiftProtobuf.Visitor>(visitor: inout V) throws { |
|||
if !self.unit.isEmpty { |
|||
try visitor.visitSingularStringField(value: self.unit, fieldNumber: 1) |
|||
} |
|||
if self.quantity.bitPattern != 0 { |
|||
try visitor.visitSingularFloatField(value: self.quantity, fieldNumber: 2) |
|||
} |
|||
try unknownFields.traverse(visitor: &visitor) |
|||
} |
|||
|
|||
static func ==(lhs: Data_Speech_Common_BillingItem, rhs: Data_Speech_Common_BillingItem) -> Bool { |
|||
if lhs.unit != rhs.unit {return false} |
|||
if lhs.quantity != rhs.quantity {return false} |
|||
if lhs.unknownFields != rhs.unknownFields {return false} |
|||
return true |
|||
} |
|||
} |
|||
|
|||
extension Data_Speech_Common_Billing: SwiftProtobuf.Message, SwiftProtobuf._MessageImplementationBase, SwiftProtobuf._ProtoNameProviding { |
|||
static let protoMessageName: String = _protobuf_package + ".Billing" |
|||
static let _protobuf_nameMap = SwiftProtobuf._NameMap(bytecode: "\0\u{5}Items\0items\0\u{5}DurationMsec\0duration_msec\0\u{5}WordCount\0word_count\0") |
|||
|
|||
mutating func decodeMessage<D: SwiftProtobuf.Decoder>(decoder: inout D) throws { |
|||
while let fieldNumber = try decoder.nextFieldNumber() { |
|||
// The use of inline closures is to circumvent an issue where the compiler |
|||
// allocates stack space for every case branch when no optimizations are |
|||
// enabled. https://github.com/apple/swift-protobuf/issues/1034 |
|||
switch fieldNumber { |
|||
case 1: try { try decoder.decodeRepeatedMessageField(value: &self.items) }() |
|||
case 2: try { try decoder.decodeSingularInt64Field(value: &self.durationMsec) }() |
|||
case 3: try { try decoder.decodeSingularInt64Field(value: &self.wordCount) }() |
|||
default: break |
|||
} |
|||
} |
|||
} |
|||
|
|||
func traverse<V: SwiftProtobuf.Visitor>(visitor: inout V) throws { |
|||
if !self.items.isEmpty { |
|||
try visitor.visitRepeatedMessageField(value: self.items, fieldNumber: 1) |
|||
} |
|||
if self.durationMsec != 0 { |
|||
try visitor.visitSingularInt64Field(value: self.durationMsec, fieldNumber: 2) |
|||
} |
|||
if self.wordCount != 0 { |
|||
try visitor.visitSingularInt64Field(value: self.wordCount, fieldNumber: 3) |
|||
} |
|||
try unknownFields.traverse(visitor: &visitor) |
|||
} |
|||
|
|||
static func ==(lhs: Data_Speech_Common_Billing, rhs: Data_Speech_Common_Billing) -> Bool { |
|||
if lhs.items != rhs.items {return false} |
|||
if lhs.durationMsec != rhs.durationMsec {return false} |
|||
if lhs.wordCount != rhs.wordCount {return false} |
|||
if lhs.unknownFields != rhs.unknownFields {return false} |
|||
return true |
|||
} |
|||
} |
|||
|
|||
extension Data_Speech_Common_ResponseMeta: SwiftProtobuf.Message, SwiftProtobuf._MessageImplementationBase, SwiftProtobuf._ProtoNameProviding { |
|||
static let protoMessageName: String = _protobuf_package + ".ResponseMeta" |
|||
static let _protobuf_nameMap = SwiftProtobuf._NameMap(bytecode: "\0\u{5}SessionID\0session_id\0\u{5}Sequence\0sequence\0\u{5}StatusCode\0status_code\0\u{5}Message\0message\0\u{5}Billing\0billing\0") |
|||
|
|||
mutating func decodeMessage<D: SwiftProtobuf.Decoder>(decoder: inout D) throws { |
|||
while let fieldNumber = try decoder.nextFieldNumber() { |
|||
// The use of inline closures is to circumvent an issue where the compiler |
|||
// allocates stack space for every case branch when no optimizations are |
|||
// enabled. https://github.com/apple/swift-protobuf/issues/1034 |
|||
switch fieldNumber { |
|||
case 1: try { try decoder.decodeSingularStringField(value: &self.sessionID) }() |
|||
case 2: try { try decoder.decodeSingularInt32Field(value: &self.sequence) }() |
|||
case 3: try { try decoder.decodeSingularInt32Field(value: &self.statusCode) }() |
|||
case 4: try { try decoder.decodeSingularStringField(value: &self.message) }() |
|||
case 5: try { try decoder.decodeSingularMessageField(value: &self._billing) }() |
|||
default: break |
|||
} |
|||
} |
|||
} |
|||
|
|||
func traverse<V: SwiftProtobuf.Visitor>(visitor: inout V) throws { |
|||
// The use of inline closures is to circumvent an issue where the compiler |
|||
// allocates stack space for every if/case branch local when no optimizations |
|||
// are enabled. https://github.com/apple/swift-protobuf/issues/1034 and |
|||
// https://github.com/apple/swift-protobuf/issues/1182 |
|||
if !self.sessionID.isEmpty { |
|||
try visitor.visitSingularStringField(value: self.sessionID, fieldNumber: 1) |
|||
} |
|||
if self.sequence != 0 { |
|||
try visitor.visitSingularInt32Field(value: self.sequence, fieldNumber: 2) |
|||
} |
|||
if self.statusCode != 0 { |
|||
try visitor.visitSingularInt32Field(value: self.statusCode, fieldNumber: 3) |
|||
} |
|||
if !self.message.isEmpty { |
|||
try visitor.visitSingularStringField(value: self.message, fieldNumber: 4) |
|||
} |
|||
try { if let v = self._billing { |
|||
try visitor.visitSingularMessageField(value: v, fieldNumber: 5) |
|||
} }() |
|||
try unknownFields.traverse(visitor: &visitor) |
|||
} |
|||
|
|||
static func ==(lhs: Data_Speech_Common_ResponseMeta, rhs: Data_Speech_Common_ResponseMeta) -> Bool { |
|||
if lhs.sessionID != rhs.sessionID {return false} |
|||
if lhs.sequence != rhs.sequence {return false} |
|||
if lhs.statusCode != rhs.statusCode {return false} |
|||
if lhs.message != rhs.message {return false} |
|||
if lhs._billing != rhs._billing {return false} |
|||
if lhs.unknownFields != rhs.unknownFields {return false} |
|||
return true |
|||
} |
|||
} |
|||
@ -0,0 +1,461 @@ |
|||
// DO NOT EDIT. |
|||
// swift-format-ignore-file |
|||
// swiftlint:disable all |
|||
// |
|||
// Generated by the Swift generator plugin for the protocol buffer compiler. |
|||
// Source: products/understanding/ast/ast_service.proto |
|||
// |
|||
// For information on using the generated types, please see the documentation: |
|||
// https://github.com/apple/swift-protobuf/ |
|||
|
|||
import Foundation |
|||
import SwiftProtobuf |
|||
|
|||
// If the compiler emits an error on this type, it is because this file |
|||
// was generated by a version of the `protoc` Swift plug-in that is |
|||
// incompatible with the version of SwiftProtobuf to which you are linking. |
|||
// Please ensure that you are building against the same version of the API |
|||
// that was used to generate this file. |
|||
fileprivate struct _GeneratedWithProtocGenSwiftVersion: SwiftProtobuf.ProtobufAPIVersionCheck { |
|||
struct _2: SwiftProtobuf.ProtobufAPIVersion_2 {} |
|||
typealias Version = _2 |
|||
} |
|||
|
|||
struct Data_Speech_Ast_ReqParams: @unchecked Sendable { |
|||
// SwiftProtobuf.Message conformance is added in an extension below. See the |
|||
// `Message` and `Message+*Additions` files in the SwiftProtobuf library for |
|||
// methods supported on all messages. |
|||
|
|||
/// 可能是s2t , s2s 选一个, 控制是否需要语音 |
|||
var mode: String { |
|||
get {return _storage._mode} |
|||
set {_uniqueStorage()._mode = newValue} |
|||
} |
|||
|
|||
/// 源语言 |
|||
var sourceLanguage: String { |
|||
get {return _storage._sourceLanguage} |
|||
set {_uniqueStorage()._sourceLanguage = newValue} |
|||
} |
|||
|
|||
/// 目标语言 |
|||
var targetLanguage: String { |
|||
get {return _storage._targetLanguage} |
|||
set {_uniqueStorage()._targetLanguage = newValue} |
|||
} |
|||
|
|||
var speakerID: String { |
|||
get {return _storage._speakerID} |
|||
set {_uniqueStorage()._speakerID = newValue} |
|||
} |
|||
|
|||
var corpus: Data_Speech_Understanding_Corpus { |
|||
get {return _storage._corpus ?? Data_Speech_Understanding_Corpus()} |
|||
set {_uniqueStorage()._corpus = newValue} |
|||
} |
|||
/// Returns true if `corpus` has been explicitly set. |
|||
var hasCorpus: Bool {return _storage._corpus != nil} |
|||
/// Clears the value of `corpus`. Subsequent reads from it will return its default value. |
|||
mutating func clearCorpus() {_uniqueStorage()._corpus = nil} |
|||
|
|||
var unknownFields = SwiftProtobuf.UnknownStorage() |
|||
|
|||
init() {} |
|||
|
|||
fileprivate var _storage = _StorageClass.defaultInstance |
|||
} |
|||
|
|||
struct Data_Speech_Ast_TranslateRequest: @unchecked Sendable { |
|||
// SwiftProtobuf.Message conformance is added in an extension below. See the |
|||
// `Message` and `Message+*Additions` files in the SwiftProtobuf library for |
|||
// methods supported on all messages. |
|||
|
|||
var requestMeta: Data_Speech_Common_RequestMeta { |
|||
get {return _storage._requestMeta ?? Data_Speech_Common_RequestMeta()} |
|||
set {_uniqueStorage()._requestMeta = newValue} |
|||
} |
|||
/// Returns true if `requestMeta` has been explicitly set. |
|||
var hasRequestMeta: Bool {return _storage._requestMeta != nil} |
|||
/// Clears the value of `requestMeta`. Subsequent reads from it will return its default value. |
|||
mutating func clearRequestMeta() {_uniqueStorage()._requestMeta = nil} |
|||
|
|||
var event: Data_Speech_Event_Type { |
|||
get {return _storage._event} |
|||
set {_uniqueStorage()._event = newValue} |
|||
} |
|||
|
|||
var user: Data_Speech_Understanding_User { |
|||
get {return _storage._user ?? Data_Speech_Understanding_User()} |
|||
set {_uniqueStorage()._user = newValue} |
|||
} |
|||
/// Returns true if `user` has been explicitly set. |
|||
var hasUser: Bool {return _storage._user != nil} |
|||
/// Clears the value of `user`. Subsequent reads from it will return its default value. |
|||
mutating func clearUser() {_uniqueStorage()._user = nil} |
|||
|
|||
/// bytes data = 2; // request binary data |
|||
var sourceAudio: Data_Speech_Understanding_Audio { |
|||
get {return _storage._sourceAudio ?? Data_Speech_Understanding_Audio()} |
|||
set {_uniqueStorage()._sourceAudio = newValue} |
|||
} |
|||
/// Returns true if `sourceAudio` has been explicitly set. |
|||
var hasSourceAudio: Bool {return _storage._sourceAudio != nil} |
|||
/// Clears the value of `sourceAudio`. Subsequent reads from it will return its default value. |
|||
mutating func clearSourceAudio() {_uniqueStorage()._sourceAudio = nil} |
|||
|
|||
/// 目标音频信息，只需要传format（pcm/ogg）、rate、bits、channel这些 |
|||
var targetAudio: Data_Speech_Understanding_Audio { |
|||
get {return _storage._targetAudio ?? Data_Speech_Understanding_Audio()} |
|||
set {_uniqueStorage()._targetAudio = newValue} |
|||
} |
|||
/// Returns true if `targetAudio` has been explicitly set. |
|||
var hasTargetAudio: Bool {return _storage._targetAudio != nil} |
|||
/// Clears the value of `targetAudio`. Subsequent reads from it will return its default value. |
|||
mutating func clearTargetAudio() {_uniqueStorage()._targetAudio = nil} |
|||
|
|||
/// 请求参数 |
|||
var request: Data_Speech_Ast_ReqParams { |
|||
get {return _storage._request ?? Data_Speech_Ast_ReqParams()} |
|||
set {_uniqueStorage()._request = newValue} |
|||
} |
|||
/// Returns true if `request` has been explicitly set. |
|||
var hasRequest: Bool {return _storage._request != nil} |
|||
/// Clears the value of `request`. Subsequent reads from it will return its default value. |
|||
mutating func clearRequest() {_uniqueStorage()._request = nil} |
|||
|
|||
/// 是否开启降噪 |
|||
var denoise: Bool { |
|||
get {return _storage._denoise ?? false} |
|||
set {_uniqueStorage()._denoise = newValue} |
|||
} |
|||
/// Returns true if `denoise` has been explicitly set. |
|||
var hasDenoise: Bool {return _storage._denoise != nil} |
|||
/// Clears the value of `denoise`. Subsequent reads from it will return its default value. |
|||
mutating func clearDenoise() {_uniqueStorage()._denoise = nil} |
|||
|
|||
var unknownFields = SwiftProtobuf.UnknownStorage() |
|||
|
|||
init() {} |
|||
|
|||
fileprivate var _storage = _StorageClass.defaultInstance |
|||
} |
|||
|
|||
struct Data_Speech_Ast_TranslateResponse: Sendable { |
|||
// SwiftProtobuf.Message conformance is added in an extension below. See the |
|||
// `Message` and `Message+*Additions` files in the SwiftProtobuf library for |
|||
// methods supported on all messages. |
|||
|
|||
var responseMeta: Data_Speech_Common_ResponseMeta { |
|||
get {return _responseMeta ?? Data_Speech_Common_ResponseMeta()} |
|||
set {_responseMeta = newValue} |
|||
} |
|||
/// Returns true if `responseMeta` has been explicitly set. |
|||
var hasResponseMeta: Bool {return self._responseMeta != nil} |
|||
/// Clears the value of `responseMeta`. Subsequent reads from it will return its default value. |
|||
mutating func clearResponseMeta() {self._responseMeta = nil} |
|||
|
|||
var event: Data_Speech_Event_Type = .none |
|||
|
|||
/// response binary data |
|||
var data: Data = Data() |
|||
|
|||
/// 原文或者译文 |
|||
var text: String = String() |
|||
|
|||
var startTime: Int32 = 0 |
|||
|
|||
var endTime: Int32 = 0 |
|||
|
|||
var spkChg: Bool = false |
|||
|
|||
var mutedDurationMs: Int32 = 0 |
|||
|
|||
var unknownFields = SwiftProtobuf.UnknownStorage() |
|||
|
|||
init() {} |
|||
|
|||
fileprivate var _responseMeta: Data_Speech_Common_ResponseMeta? = nil |
|||
} |
|||
|
|||
// MARK: - Code below here is support for the SwiftProtobuf runtime. |
|||
|
|||
fileprivate let _protobuf_package = "data.speech.ast" |
|||
|
|||
extension Data_Speech_Ast_ReqParams: SwiftProtobuf.Message, SwiftProtobuf._MessageImplementationBase, SwiftProtobuf._ProtoNameProviding { |
|||
static let protoMessageName: String = _protobuf_package + ".ReqParams" |
|||
static let _protobuf_nameMap = SwiftProtobuf._NameMap(bytecode: "\0\u{1}mode\0\u{3}source_language\0\u{3}target_language\0\u{3}speaker_id\0\u{2}`\u{1}corpus\0") |
|||
|
|||
fileprivate class _StorageClass { |
|||
var _mode: String = String() |
|||
var _sourceLanguage: String = String() |
|||
var _targetLanguage: String = String() |
|||
var _speakerID: String = String() |
|||
var _corpus: Data_Speech_Understanding_Corpus? = nil |
|||
|
|||
// This property is used as the initial default value for new instances of the type. |
|||
// The type itself is protecting the reference to its storage via CoW semantics. |
|||
// This will force a copy to be made of this reference when the first mutation occurs; |
|||
// hence, it is safe to mark this as `nonisolated(unsafe)`. |
|||
static nonisolated(unsafe) let defaultInstance = _StorageClass() |
|||
|
|||
private init() {} |
|||
|
|||
init(copying source: _StorageClass) { |
|||
_mode = source._mode |
|||
_sourceLanguage = source._sourceLanguage |
|||
_targetLanguage = source._targetLanguage |
|||
_speakerID = source._speakerID |
|||
_corpus = source._corpus |
|||
} |
|||
} |
|||
|
|||
fileprivate mutating func _uniqueStorage() -> _StorageClass { |
|||
if !isKnownUniquelyReferenced(&_storage) { |
|||
_storage = _StorageClass(copying: _storage) |
|||
} |
|||
return _storage |
|||
} |
|||
|
|||
mutating func decodeMessage<D: SwiftProtobuf.Decoder>(decoder: inout D) throws { |
|||
_ = _uniqueStorage() |
|||
try withExtendedLifetime(_storage) { (_storage: _StorageClass) in |
|||
while let fieldNumber = try decoder.nextFieldNumber() { |
|||
// The use of inline closures is to circumvent an issue where the compiler |
|||
// allocates stack space for every case branch when no optimizations are |
|||
// enabled. https://github.com/apple/swift-protobuf/issues/1034 |
|||
switch fieldNumber { |
|||
case 1: try { try decoder.decodeSingularStringField(value: &_storage._mode) }() |
|||
case 2: try { try decoder.decodeSingularStringField(value: &_storage._sourceLanguage) }() |
|||
case 3: try { try decoder.decodeSingularStringField(value: &_storage._targetLanguage) }() |
|||
case 4: try { try decoder.decodeSingularStringField(value: &_storage._speakerID) }() |
|||
case 100: try { try decoder.decodeSingularMessageField(value: &_storage._corpus) }() |
|||
default: break |
|||
} |
|||
} |
|||
} |
|||
} |
|||
|
|||
func traverse<V: SwiftProtobuf.Visitor>(visitor: inout V) throws { |
|||
try withExtendedLifetime(_storage) { (_storage: _StorageClass) in |
|||
// The use of inline closures is to circumvent an issue where the compiler |
|||
// allocates stack space for every if/case branch local when no optimizations |
|||
// are enabled. https://github.com/apple/swift-protobuf/issues/1034 and |
|||
// https://github.com/apple/swift-protobuf/issues/1182 |
|||
if !_storage._mode.isEmpty { |
|||
try visitor.visitSingularStringField(value: _storage._mode, fieldNumber: 1) |
|||
} |
|||
if !_storage._sourceLanguage.isEmpty { |
|||
try visitor.visitSingularStringField(value: _storage._sourceLanguage, fieldNumber: 2) |
|||
} |
|||
if !_storage._targetLanguage.isEmpty { |
|||
try visitor.visitSingularStringField(value: _storage._targetLanguage, fieldNumber: 3) |
|||
} |
|||
if !_storage._speakerID.isEmpty { |
|||
try visitor.visitSingularStringField(value: _storage._speakerID, fieldNumber: 4) |
|||
} |
|||
try { if let v = _storage._corpus { |
|||
try visitor.visitSingularMessageField(value: v, fieldNumber: 100) |
|||
} }() |
|||
} |
|||
try unknownFields.traverse(visitor: &visitor) |
|||
} |
|||
|
|||
static func ==(lhs: Data_Speech_Ast_ReqParams, rhs: Data_Speech_Ast_ReqParams) -> Bool { |
|||
if lhs._storage !== rhs._storage { |
|||
let storagesAreEqual: Bool = withExtendedLifetime((lhs._storage, rhs._storage)) { (_args: (_StorageClass, _StorageClass)) in |
|||
let _storage = _args.0 |
|||
let rhs_storage = _args.1 |
|||
if _storage._mode != rhs_storage._mode {return false} |
|||
if _storage._sourceLanguage != rhs_storage._sourceLanguage {return false} |
|||
if _storage._targetLanguage != rhs_storage._targetLanguage {return false} |
|||
if _storage._speakerID != rhs_storage._speakerID {return false} |
|||
if _storage._corpus != rhs_storage._corpus {return false} |
|||
return true |
|||
} |
|||
if !storagesAreEqual {return false} |
|||
} |
|||
if lhs.unknownFields != rhs.unknownFields {return false} |
|||
return true |
|||
} |
|||
} |
|||
|
|||
extension Data_Speech_Ast_TranslateRequest: SwiftProtobuf.Message, SwiftProtobuf._MessageImplementationBase, SwiftProtobuf._ProtoNameProviding { |
|||
static let protoMessageName: String = _protobuf_package + ".TranslateRequest" |
|||
static let _protobuf_nameMap = SwiftProtobuf._NameMap(bytecode: "\0\u{3}request_meta\0\u{1}event\0\u{1}user\0\u{3}source_audio\0\u{3}target_audio\0\u{1}request\0\u{1}denoise\0") |
|||
|
|||
fileprivate class _StorageClass { |
|||
var _requestMeta: Data_Speech_Common_RequestMeta? = nil |
|||
var _event: Data_Speech_Event_Type = .none |
|||
var _user: Data_Speech_Understanding_User? = nil |
|||
var _sourceAudio: Data_Speech_Understanding_Audio? = nil |
|||
var _targetAudio: Data_Speech_Understanding_Audio? = nil |
|||
var _request: Data_Speech_Ast_ReqParams? = nil |
|||
var _denoise: Bool? = nil |
|||
|
|||
// This property is used as the initial default value for new instances of the type. |
|||
// The type itself is protecting the reference to its storage via CoW semantics. |
|||
// This will force a copy to be made of this reference when the first mutation occurs; |
|||
// hence, it is safe to mark this as `nonisolated(unsafe)`. |
|||
static nonisolated(unsafe) let defaultInstance = _StorageClass() |
|||
|
|||
private init() {} |
|||
|
|||
init(copying source: _StorageClass) { |
|||
_requestMeta = source._requestMeta |
|||
_event = source._event |
|||
_user = source._user |
|||
_sourceAudio = source._sourceAudio |
|||
_targetAudio = source._targetAudio |
|||
_request = source._request |
|||
_denoise = source._denoise |
|||
} |
|||
} |
|||
|
|||
fileprivate mutating func _uniqueStorage() -> _StorageClass { |
|||
if !isKnownUniquelyReferenced(&_storage) { |
|||
_storage = _StorageClass(copying: _storage) |
|||
} |
|||
return _storage |
|||
} |
|||
|
|||
mutating func decodeMessage<D: SwiftProtobuf.Decoder>(decoder: inout D) throws { |
|||
_ = _uniqueStorage() |
|||
try withExtendedLifetime(_storage) { (_storage: _StorageClass) in |
|||
while let fieldNumber = try decoder.nextFieldNumber() { |
|||
// The use of inline closures is to circumvent an issue where the compiler |
|||
// allocates stack space for every case branch when no optimizations are |
|||
// enabled. https://github.com/apple/swift-protobuf/issues/1034 |
|||
switch fieldNumber { |
|||
case 1: try { try decoder.decodeSingularMessageField(value: &_storage._requestMeta) }() |
|||
case 2: try { try decoder.decodeSingularEnumField(value: &_storage._event) }() |
|||
case 3: try { try decoder.decodeSingularMessageField(value: &_storage._user) }() |
|||
case 4: try { try decoder.decodeSingularMessageField(value: &_storage._sourceAudio) }() |
|||
case 5: try { try decoder.decodeSingularMessageField(value: &_storage._targetAudio) }() |
|||
case 6: try { try decoder.decodeSingularMessageField(value: &_storage._request) }() |
|||
case 7: try { try decoder.decodeSingularBoolField(value: &_storage._denoise) }() |
|||
default: break |
|||
} |
|||
} |
|||
} |
|||
} |
|||
|
|||
func traverse<V: SwiftProtobuf.Visitor>(visitor: inout V) throws { |
|||
try withExtendedLifetime(_storage) { (_storage: _StorageClass) in |
|||
// The use of inline closures is to circumvent an issue where the compiler |
|||
// allocates stack space for every if/case branch local when no optimizations |
|||
// are enabled. https://github.com/apple/swift-protobuf/issues/1034 and |
|||
// https://github.com/apple/swift-protobuf/issues/1182 |
|||
try { if let v = _storage._requestMeta { |
|||
try visitor.visitSingularMessageField(value: v, fieldNumber: 1) |
|||
} }() |
|||
if _storage._event != .none { |
|||
try visitor.visitSingularEnumField(value: _storage._event, fieldNumber: 2) |
|||
} |
|||
try { if let v = _storage._user { |
|||
try visitor.visitSingularMessageField(value: v, fieldNumber: 3) |
|||
} }() |
|||
try { if let v = _storage._sourceAudio { |
|||
try visitor.visitSingularMessageField(value: v, fieldNumber: 4) |
|||
} }() |
|||
try { if let v = _storage._targetAudio { |
|||
try visitor.visitSingularMessageField(value: v, fieldNumber: 5) |
|||
} }() |
|||
try { if let v = _storage._request { |
|||
try visitor.visitSingularMessageField(value: v, fieldNumber: 6) |
|||
} }() |
|||
try { if let v = _storage._denoise { |
|||
try visitor.visitSingularBoolField(value: v, fieldNumber: 7) |
|||
} }() |
|||
} |
|||
try unknownFields.traverse(visitor: &visitor) |
|||
} |
|||
|
|||
static func ==(lhs: Data_Speech_Ast_TranslateRequest, rhs: Data_Speech_Ast_TranslateRequest) -> Bool { |
|||
if lhs._storage !== rhs._storage { |
|||
let storagesAreEqual: Bool = withExtendedLifetime((lhs._storage, rhs._storage)) { (_args: (_StorageClass, _StorageClass)) in |
|||
let _storage = _args.0 |
|||
let rhs_storage = _args.1 |
|||
if _storage._requestMeta != rhs_storage._requestMeta {return false} |
|||
if _storage._event != rhs_storage._event {return false} |
|||
if _storage._user != rhs_storage._user {return false} |
|||
if _storage._sourceAudio != rhs_storage._sourceAudio {return false} |
|||
if _storage._targetAudio != rhs_storage._targetAudio {return false} |
|||
if _storage._request != rhs_storage._request {return false} |
|||
if _storage._denoise != rhs_storage._denoise {return false} |
|||
return true |
|||
} |
|||
if !storagesAreEqual {return false} |
|||
} |
|||
if lhs.unknownFields != rhs.unknownFields {return false} |
|||
return true |
|||
} |
|||
} |
|||
|
|||
extension Data_Speech_Ast_TranslateResponse: SwiftProtobuf.Message, SwiftProtobuf._MessageImplementationBase, SwiftProtobuf._ProtoNameProviding { |
|||
static let protoMessageName: String = _protobuf_package + ".TranslateResponse" |
|||
static let _protobuf_nameMap = SwiftProtobuf._NameMap(bytecode: "\0\u{3}response_meta\0\u{1}event\0\u{1}data\0\u{1}text\0\u{3}start_time\0\u{3}end_time\0\u{3}spk_chg\0\u{3}muted_duration_ms\0") |
|||
|
|||
mutating func decodeMessage<D: SwiftProtobuf.Decoder>(decoder: inout D) throws { |
|||
while let fieldNumber = try decoder.nextFieldNumber() { |
|||
// The use of inline closures is to circumvent an issue where the compiler |
|||
// allocates stack space for every case branch when no optimizations are |
|||
// enabled. https://github.com/apple/swift-protobuf/issues/1034 |
|||
switch fieldNumber { |
|||
case 1: try { try decoder.decodeSingularMessageField(value: &self._responseMeta) }() |
|||
case 2: try { try decoder.decodeSingularEnumField(value: &self.event) }() |
|||
case 3: try { try decoder.decodeSingularBytesField(value: &self.data) }() |
|||
case 4: try { try decoder.decodeSingularStringField(value: &self.text) }() |
|||
case 5: try { try decoder.decodeSingularInt32Field(value: &self.startTime) }() |
|||
case 6: try { try decoder.decodeSingularInt32Field(value: &self.endTime) }() |
|||
case 7: try { try decoder.decodeSingularBoolField(value: &self.spkChg) }() |
|||
case 8: try { try decoder.decodeSingularInt32Field(value: &self.mutedDurationMs) }() |
|||
default: break |
|||
} |
|||
} |
|||
} |
|||
|
|||
func traverse<V: SwiftProtobuf.Visitor>(visitor: inout V) throws { |
|||
// The use of inline closures is to circumvent an issue where the compiler |
|||
// allocates stack space for every if/case branch local when no optimizations |
|||
// are enabled. https://github.com/apple/swift-protobuf/issues/1034 and |
|||
// https://github.com/apple/swift-protobuf/issues/1182 |
|||
try { if let v = self._responseMeta { |
|||
try visitor.visitSingularMessageField(value: v, fieldNumber: 1) |
|||
} }() |
|||
if self.event != .none { |
|||
try visitor.visitSingularEnumField(value: self.event, fieldNumber: 2) |
|||
} |
|||
if !self.data.isEmpty { |
|||
try visitor.visitSingularBytesField(value: self.data, fieldNumber: 3) |
|||
} |
|||
if !self.text.isEmpty { |
|||
try visitor.visitSingularStringField(value: self.text, fieldNumber: 4) |
|||
} |
|||
if self.startTime != 0 { |
|||
try visitor.visitSingularInt32Field(value: self.startTime, fieldNumber: 5) |
|||
} |
|||
if self.endTime != 0 { |
|||
try visitor.visitSingularInt32Field(value: self.endTime, fieldNumber: 6) |
|||
} |
|||
if self.spkChg != false { |
|||
try visitor.visitSingularBoolField(value: self.spkChg, fieldNumber: 7) |
|||
} |
|||
if self.mutedDurationMs != 0 { |
|||
try visitor.visitSingularInt32Field(value: self.mutedDurationMs, fieldNumber: 8) |
|||
} |
|||
try unknownFields.traverse(visitor: &visitor) |
|||
} |
|||
|
|||
static func ==(lhs: Data_Speech_Ast_TranslateResponse, rhs: Data_Speech_Ast_TranslateResponse) -> Bool { |
|||
if lhs._responseMeta != rhs._responseMeta {return false} |
|||
if lhs.event != rhs.event {return false} |
|||
if lhs.data != rhs.data {return false} |
|||
if lhs.text != rhs.text {return false} |
|||
if lhs.startTime != rhs.startTime {return false} |
|||
if lhs.endTime != rhs.endTime {return false} |
|||
if lhs.spkChg != rhs.spkChg {return false} |
|||
if lhs.mutedDurationMs != rhs.mutedDurationMs {return false} |
|||
if lhs.unknownFields != rhs.unknownFields {return false} |
|||
return true |
|||
} |
|||
} |
|||
File diff suppressed because it is too large
@ -0,0 +1,20 @@ |
|||
import 'package:flutter/material.dart'; |
|||
|
|||
void main() { |
|||
runApp(const MainApp()); |
|||
} |
|||
|
|||
class MainApp extends StatelessWidget { |
|||
const MainApp({super.key}); |
|||
|
|||
@override |
|||
Widget build(BuildContext context) { |
|||
return const MaterialApp( |
|||
home: Scaffold( |
|||
body: Center( |
|||
child: Text('Hello World!'), |
|||
), |
|||
), |
|||
); |
|||
} |
|||
} |
|||
Loading…
Reference in new issue