微软SDK文字转语音基本使用

1、导入SDK

OC支持版本

pod 'MicrosoftCognitiveServicesSpeech-iOS', '~> 1.33.0'

2、导入头文件

import <MicrosoftCognitiveServicesSpeech/SPXSpeechApi.h>

3、初始化

从自己服务器或者微软开发后台自己拿到ttsRegion和ttsToken

SPXSpeechConfiguration * config = [[SPXSpeechConfiguration alloc] initWithAuthorizationToken:@"ttsToken"] region:[@"ttsRegion"] error:nil];

4、转化

/**  
  1、设置语音文件保存地址
*/
            NSString *uuid = [[NSUUID UUID] UUIDString];
            NSArray *paths = NSSearchPathForDirectoriesInDomains(NSDocumentDirectory,
                                                                 NSUserDomainMask, YES);
            NSString *documentDirectory = [paths objectAtIndex:0];
            NSString *path = [documentDirectory stringByAppendingPathComponent:@"Voice"];
            if (![[NSFileManager defaultManager] fileExistsAtPath:path]) {
                [[NSFileManager defaultManager] createDirectoryAtPath:path withIntermediateDirectories:YES attributes:nil error:nil];
            }
            path = [NSString stringWithFormat:@"%@/%@.wav",path,uuid];
/**
  2、配置地址
*/
            SPXAudioConfiguration * audioConfig = [[SPXAudioConfiguration alloc] initWithWavFileOutput:path];
            SPXSpeechSynthesizer * synthesize = [[SPXSpeechSynthesizer alloc] initWithSpeechConfiguration:config audioConfiguration:audioConfig];
/**
  3、
配置角色、人物、风格等参数，并开始转化我使用的是ssml方式，
结果监听、结果描述
还有其他方式，可在官方文档查阅
*/
            NSString * ssmlString = [TTVConfigService.shareService ssmlString:self.typeTextView.text];
            SPXSpeechSynthesisResult *result = [synthesize speakSsml:ssmlString];
            NSLog(@"result %.f",result.audioDuration);
            SPXSpeechSynthesisCancellationDetails * detail = [[SPXSpeechSynthesisCancellationDetails alloc] initFromCanceledSynthesisResult:result];

/**
替换掉对应位置的参数。
MainJson：是在官方下载的支持的所有语言和角色、风格等。
这里只简单设置了三个参数，还有很多可根据自己需求设置
*/
- (NSString *) ssmlString:(NSString *) content {
    NSString * path = [[NSBundle mainBundle] pathForResource:@"MainJson" ofType:@"geojson"];
    if (!path) {
        return nil;
    }
    NSString *parseJason = [[NSString alloc] initWithContentsOfFile:path encoding:NSUTF8StringEncoding error:nil];
    NSData *jaonData = [parseJason dataUsingEncoding:NSUTF8StringEncoding];
    NSDictionary *dict = [NSJSONSerialization JSONObjectWithData:jaonData options:NSJSONReadingMutableContainers error:nil];
    __block NSString * language = SafeString(self.language);
    
    NSArray * lanuagesArray =  self.totalConfigDictionary[@"voiceVoList"];
    
    [lanuagesArray enumerateObjectsUsingBlock:^(NSDictionary *  _Nonnull obj_, NSUInteger idx, BOOL * _Nonnull stop) {
        if ([[obj_ objectForKey:@"language"] isEqualToString:self.language]) {
            language = [obj_ objectForKey:@"locale"];
        }
    }];
    
    NSString * ssmlString = [NSString stringWithFormat:dict[@"mainJson"],SafeString(language) ,SafeString(self.lanuageShortName),SafeString(self.selectedVoice),SafeString(self.selectedStyle),content];
    NSLog(@"%@",ssmlString);

    return ssmlString;
}

{
    "mainJson" : "<speak xmlns:mstts=\"https://www.w3.org/2001/mstts\" version=\"1.0\"\n       xmlns=\"http://www.w3.org/2001/10/synthesis\" xml:lang=\"%@\">\n    <voice name=\"%@\">\n        <mstts:express-as style=\"%@\" role=\"%@\" styledegree=\"1\">\n            %@\n        </mstts:express-as>\n    </voice>\n</speak>\n\n"
}

官方文档地址

然后就可以将本地语音文件使用到自己的业务了。