C++ Espeak SAPI/dll在Windows上的使用情况？

问题：我正在尝试使用espeak文本到语音引擎。 因此，我让它在linux上正常工作（代码如下）。 现在我也想把这个基本程序移植到windows上，但这几乎是不可能的

部分问题是windows dll只允许音频输出同步，这意味着它需要回调，但我不知道如何从回调播放音频。。。首先它崩溃了，然后我意识到，我需要一个回调函数，现在我得到了回调函数中的数据，但我不知道如何播放它。。。因为它既不是wav文件，也不像在Linux上那样自动播放

sourceforge站点是相当无用的，因为它基本上说使用SAPI版本，但是没有关于如何使用SAPI espeak dll的示例

不管怎样，这是我的密码，有人能帮忙吗

#ifdef __cplusplus
#include <cstdio>
#include <cstdlib>
#include <cstring>
#else
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#endif

#include <assert.h>
#include <ctype.h>

//#include "speak_lib.h"
#include "espeak/speak_lib.h"

// libespeak-dev: /usr/include/espeak/speak_lib.h
// apt-get install libespeak-dev
// apt-get install libportaudio-dev

// g++ -o mine mine.cpp -lespeak
// g++ -o mine mine.cpp -I/usr/include/espeak/ -lespeak
// gcc -o mine mine.cpp -I/usr/include/espeak/ -lespeak


char voicename[40];
int samplerate;
int quiet = 0;
static char genders[4] = {' ','M','F',' '};

//const char *data_path = "/usr/share/";   // /usr/share/espeak-data/
const char *data_path = NULL;   // use default path for espeak-data


int strrcmp(const char *s, const char *sub)
{
int slen = strlen(s);
int sublen = strlen(sub);
return memcmp(s + slen - sublen, sub, sublen);
}


char * strrcpy(char *dest, const char *source)
{
// Pre assertions
assert(dest != NULL);
assert(source != NULL);
assert(dest != source);

// tk: parentheses
while((*dest++ = *source++))
    ;
return(--dest);
}

const char* GetLanguageVoiceName(const char* pszShortSign)
{
#define LANGUAGE_LENGTH 30
static char szReturnValue[LANGUAGE_LENGTH] ;
memset(szReturnValue, 0, LANGUAGE_LENGTH);

for (int i = 0; pszShortSign[i] != '\0'; ++i)
    szReturnValue[i] = (char) tolower(pszShortSign[i]);

const espeak_VOICE **voices;
espeak_VOICE voice_select;
voices = espeak_ListVoices(NULL);

const espeak_VOICE *v;
for(int ix=0; (v = voices[ix]) != NULL; ix++)
{
    if( !strrcmp( v->languages, szReturnValue) )
    {
        strcpy(szReturnValue, v->name);
        return szReturnValue;
    }
} // End for

strcpy(szReturnValue, "default");
return szReturnValue;
} // End function getvoicename


void ListVoices()
{
const espeak_VOICE **voices;
espeak_VOICE voice_select;
voices = espeak_ListVoices(NULL);

const espeak_VOICE *v;
for(int ix=0; (v = voices[ix]) != NULL; ix++)
{
    printf("Shortsign: %s\n", v->languages);
    printf("age: %d\n", v->age);
    printf("gender: %c\n", genders[v->gender]);
    printf("name: %s\n", v->name);
    printf("\n\n");
} // End for
} // End function getvoicename


int main()
{
printf("Hello World!\n");
const char* szVersionInfo = espeak_Info(NULL);

printf("Espeak version: %s\n", szVersionInfo);
samplerate = espeak_Initialize(AUDIO_OUTPUT_PLAYBACK,0,data_path,0);

strcpy(voicename, "default");
// espeak --voices
strcpy(voicename, "german");
strcpy(voicename, GetLanguageVoiceName("DE"));

if(espeak_SetVoiceByName(voicename) != EE_OK)
{
    printf("Espeak setvoice error...\n");
}

static char word[200] = "Hello World" ;
strcpy(word, "TV-fäns aufgepasst, es ist 20 Uhr 15. Zeit für Rambo 3");
strcpy(word, "Unnamed Player wurde zum Opfer von GSG9");
int speed = 220;
int volume = 500; // volume in range 0-100    0=silence
int pitch = 50; //  base pitch, range 0-100.  50=normal

// espeak.cpp 625
espeak_SetParameter(espeakRATE, speed, 0);
espeak_SetParameter(espeakVOLUME,volume,0);
espeak_SetParameter(espeakPITCH,pitch,0);
// espeakRANGE:   pitch range, range 0-100. 0-monotone, 50=normal
// espeakPUNCTUATION:  which punctuation characters to announce:
    // value in espeak_PUNCT_TYPE (none, all, some), 
espeak_VOICE *voice_spec = espeak_GetCurrentVoice();
voice_spec->gender=2; // 0=none 1=male, 2=female,
//voice_spec->age = age;

espeak_SetVoiceByProperties(voice_spec);


espeak_Synth( (char*) word, strlen(word)+1, 0, POS_CHARACTER, 0, espeakCHARS_AUTO, NULL, NULL);
espeak_Synchronize();

strcpy(voicename, GetLanguageVoiceName("EN"));
espeak_SetVoiceByName(voicename);
strcpy(word, "Geany was fragged by GSG9 Googlebot");
strcpy(word, "Googlebot");

espeak_Synth( (char*) word, strlen(word)+1, 0, POS_CHARACTER, 0, espeakCHARS_AUTO, NULL, NULL);
espeak_Synchronize();


espeak_Terminate();
printf("Espeak terminated\n");
return EXIT_SUCCESS; 
}

/*
if(espeak_SetVoiceByName(voicename) != EE_OK)
{
    memset(&voice_select,0,sizeof(voice_select));
    voice_select.languages = voicename;
    if(espeak_SetVoiceByProperties(&voice_select) != EE_OK)
    {
        fprintf(stderr,"%svoice '%s'\n",err_load,voicename);
        exit(2);
    }
}
*/

\ifdef\uuucplusplus
#包括
#包括
#包括
#否则
#包括
#包括
#包括
#恩迪夫
#包括
#包括
//#包括“speak_lib.h”
#包括“espeak/speak_lib.h”
//libespeak dev:/usr/include/espeak/speak_lib.h
//apt get安装libespeak dev
//apt get安装libportaudio dev
//g++-o mine mine.cpp-lespeak
//g++-o mine-mine.cpp-I/usr/include/espeak/-lespeak
//gcc-o mine mine.cpp-I/usr/include/espeak/-lespeak
char-voicename[40]；
int采样器；
int安静=0；
静态字符性别[4]={''，'M'，'F'，''}；
//const char*data_path=“/usr/share/”；///usr/share/espeak数据/
const char*data_path=NULL；//使用espeak数据的默认路径
int strrcmp（常量字符*s，常量字符*sub）
{
int slen=strlen（s）；
int SUBELN=strlen（sub）；
返回memcmp（s+slen-子项，子项，子项）；
}
char*strrcpy（char*dest，const char*source）
{
//预断言
断言（dest！=NULL）；
断言（源！=NULL）；
断言（dest！=源）；
//tk：括号
而（（*dest++=*source++）
;
返回（--dest）；
}
常量字符*GetLanguageVoiceName（常量字符*pszShortSign）
{
#定义语言长度30
静态字符返回值[语言长度]；
memset（szReturnValue，0，语言长度）；
for（int i=0；pszShortSign[i]！='\0'；++i）
szReturnValue[i]=（char）tolower（pszShortSign[i]）；
const espeak_VOICE**voices；
espeak_VOICE_选择；
voices=espeak_ListVoices（空）；
const espeak_VOICE*v；
for（int ix=0；（v=voices[ix]）！=NULL；ix++）
{
如果（！strrcmp（v->languages，szReturnValue））
{
strcpy（szReturnValue，v->name）；
返回值；
}
}//结束
strcpy（szReturnValue，默认值）；
返回值；
}//结束函数getvoicename
void ListVoices（）
{
const espeak_VOICE**voices；
espeak_VOICE_选择；
voices=espeak_ListVoices（空）；
const espeak_VOICE*v；
for（int ix=0；（v=voices[ix]）！=NULL；ix++）
{
printf（“短符号：%s\n”，v->语言）；
printf（“年龄：%d\n”，v->age）；
printf（“性别：%c\n”，性别[v->gender]）；
printf（“名称：%s\n”，v->name）；
printf（“\n\n”）；
}//结束
}//结束函数getvoicename
int main（）
{
printf（“你好，世界！\n”）；
const char*szVersionInfo=espeak_Info（空）；
printf（“Espeak版本：%s\n”，szVersionInfo）；
samplerate=espeak\u Initialize（音频输出播放，0，数据路径，0）；
strcpy（voicename，“默认”）；
//埃斯皮克——声音
strcpy（语音名称，“德语”）；
strcpy（voicename，GetLanguageVoiceName（“DE”））；
如果（espeak_SetVoiceByName（voicename）！=EE_OK）
{
printf（“Espeak setvoice错误…\n”）；
}
静态字符字[200]=“你好世界”；
strcpy（单词“TV-fäns aufgepasst，es ist 20 Uhr 15.Zeit für Rambo 3”）；
strcpy（单词，“无名球员wurde zum Opfer von GSG9”）；
内速度=220；
int volume=500；//0-100范围内的音量0=静音
int-pitch=50；//基音节，范围0-100.50=正常
//espeak.cpp 625
espeak_设置参数（espeakRATE，速度，0）；
espeak_设置参数（espeakVOLUME，volume，0）；
espeak_设置参数（espeakPITCH，pitch，0）；
//espeakRANGE：音高范围，范围0-100。0-单调，50=正常
//Espeak标点符号：要宣布的标点符号：
//espeak_PUNCT_类型中的值（无、全部、部分），
espeak_VOICE*VOICE_spec=espeak_GetCurrentVoice（）；
声音规格->性别=2；//0=无1=男性，2=女性，
//语音规格->年龄=年龄；
espeak_SetVoiceByProperties（语音规范）；
espeak_Synth（（char*）字，strlen（word）+1,0，POS_字符，0，espeak_自动，NULL，NULL）；
espeak_Synchronize（）；
strcpy（voicename，GetLanguageVoiceName（“EN”））；
espeak_SetVoiceByName（语音名称）；
strcpy（单词“Geany被GSG9谷歌机器人折磨”）；
strcpy（单词“谷歌机器人”）；
espeak_Synth（（char*）字，strlen（word）+1,0，POS_字符，0，espeak_自动，NULL，NULL）；
espeak_Synchronize（）；
espeak_Terminate（）；
printf（“Espeak终止\n”）；
返回退出成功；
}
/*
如果（espeak_SetVoiceByName（voicename）！=EE_OK）
{
memset（&voice_-select，0，sizeof（voice_-select））；
voice_select.languages=语音名称；
如果（espeak_设置VoiceByProperties（&voice_select）！=EE_OK）
{
fprintf（标准，“%s声音“%s”\n”，错误加载，语音名称）；
出口（2）；
}
}
*/

上面的代码是针对Linux的。 下面的代码是我在Vista x64（32位emu）上得到的代码：

\ifdef\uuucplusplus
#包括
#包括
#包括
#否则
#包括
#包括
#包括
#恩迪夫
#包括
#包括
#包括“speak_lib.h”
//#包括“espeak/speak_lib.h”
//libespeak dev:/usr/include/espeak/speak_lib.h
//apt get安装libespeak dev
//apt get安装libportaudio dev
//g++-o mine mine.cpp-lespeak
//g++-o mine-mine.cpp-I/usr/include/espeak/-lespeak
//gcc-o mine mine.cpp-I/usr/include/espeak/-lespeak
char-voicename[40]；
不完整；
int安静=0；
静态字符性别[4]={''，'M'，'F'，''}；
//const char*data_path=“/usr/share/”；///usr/share/espeak数据/
//const char*data_path=NULL；//使用espeak数据的默认路径
const char*data_path=“C:\\Users\\Username\\Desktop\\espeak-1.43-source\\espeak-1.43-source\\”；
int strrcmp（常量字符*s，常量字符*sub）
{
int slen=strlen（s）；
int SUBELN=strlen（sub）；
返回memcmp（s+slen-子项，子项，子项）；
}
char*strrcpy（char*dest，const
#ifdef __cplusplus
#include <cstdio>
#include <cstdlib>
#include <cstring>
#else
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#endif

#include <assert.h>
#include <ctype.h>

#include "speak_lib.h"
//#include "espeak/speak_lib.h"

// libespeak-dev: /usr/include/espeak/speak_lib.h
// apt-get install libespeak-dev
// apt-get install libportaudio-dev

// g++ -o mine mine.cpp -lespeak
// g++ -o mine mine.cpp -I/usr/include/espeak/ -lespeak
// gcc -o mine mine.cpp -I/usr/include/espeak/ -lespeak


char voicename[40];
int iSampleRate;
int quiet = 0;
static char genders[4] = {' ','M','F',' '};

//const char *data_path = "/usr/share/";   // /usr/share/espeak-data/
//const char *data_path = NULL;   // use default path for espeak-data
const char *data_path = "C:\\Users\\Username\\Desktop\\espeak-1.43-source\\espeak-1.43-source\\";


int strrcmp(const char *s, const char *sub)
{
int slen = strlen(s);
int sublen = strlen(sub);
return memcmp(s + slen - sublen, sub, sublen);
}


char * strrcpy(char *dest, const char *source)
{
// Pre assertions
assert(dest != NULL);
assert(source != NULL);
assert(dest != source);

// tk: parentheses
while((*dest++ = *source++))
    ;
return(--dest);
}

const char* GetLanguageVoiceName(const char* pszShortSign)
{
#define LANGUAGE_LENGTH 30
static char szReturnValue[LANGUAGE_LENGTH] ;
memset(szReturnValue, 0, LANGUAGE_LENGTH);

for (int i = 0; pszShortSign[i] != '\0'; ++i)
    szReturnValue[i] = (char) tolower(pszShortSign[i]);

const espeak_VOICE **voices;
espeak_VOICE voice_select;
voices = espeak_ListVoices(NULL);

const espeak_VOICE *v;
for(int ix=0; (v = voices[ix]) != NULL; ix++)
{
    if( !strrcmp( v->languages, szReturnValue) )
    {
        strcpy(szReturnValue, v->name);
        return szReturnValue;
    }
} // End for

strcpy(szReturnValue, "default");
return szReturnValue;
} // End function getvoicename


void ListVoices()
{
const espeak_VOICE **voices;
espeak_VOICE voice_select;
voices = espeak_ListVoices(NULL);

const espeak_VOICE *v;
for(int ix=0; (v = voices[ix]) != NULL; ix++)
{
    printf("Shortsign: %s\n", v->languages);
    printf("age: %d\n", v->age);
    printf("gender: %c\n", genders[v->gender]);
    printf("name: %s\n", v->name);
    printf("\n\n");
} // End for
} // End function getvoicename


/* Callback from espeak.  Directly speaks using AudioTrack. */
#define LOGI(x) printf("%s\n", x)
static int AndroidEspeakDirectSpeechCallback(short *wav, int numsamples, espeak_EVENT *events) 
{
    char buf[100];
    sprintf(buf, "AndroidEspeakDirectSpeechCallback: %d samples", numsamples);
    LOGI(buf);

    if (wav == NULL) 
{
        LOGI("Null: speech has completed");
    }

    if (numsamples > 0)
{
        //audout->write(wav, sizeof(short) * numsamples);
        sprintf(buf, "AudioTrack wrote: %d bytes", sizeof(short) * numsamples);
        LOGI(buf);
    }

    return 0;  // continue synthesis (1 is to abort)
}


static int AndroidEspeakSynthToFileCallback(short *wav, int numsamples,espeak_EVENT *events) 
{
    char buf[100];
    sprintf(buf, "AndroidEspeakSynthToFileCallback: %d samples", numsamples);
    LOGI(buf);

    if (wav == NULL) 
{
        LOGI("Null: speech has completed");
    }

    // The user data should contain the file pointer of the file to write to
    //void* user_data = events->user_data;
FILE* user_data = fopen ( "myfile1.wav" , "ab" );

    FILE* fp = static_cast<FILE *>(user_data);

    // Write all of the samples
    fwrite(wav, sizeof(short), numsamples, fp);
    return 0;  // continue synthesis (1 is to abort)
}



int main()
{
printf("Hello World!\n");
const char* szVersionInfo = espeak_Info(NULL);

printf("Espeak version: %s\n", szVersionInfo);

iSampleRate = espeak_Initialize(AUDIO_OUTPUT_SYNCHRONOUS, 4096, data_path, 0);
if (iSampleRate <= 0) 
{
    printf("Unable to initialize espeak");
    return EXIT_FAILURE;
}

//samplerate = espeak_Initialize(AUDIO_OUTPUT_PLAYBACK,0,data_path,0);

//ListVoices();

strcpy(voicename, "default");
// espeak --voices
//strcpy(voicename, "german");
//strcpy(voicename, GetLanguageVoiceName("DE"));

if(espeak_SetVoiceByName(voicename) != EE_OK)
{
    printf("Espeak setvoice error...\n");
}

static char word[200] = "Hello World" ;
strcpy(word, "TV-fäns aufgepasst, es ist 20 Uhr 15. Zeit für Rambo 3");
strcpy(word, "Unnamed Player wurde zum Opfer von GSG9");
int speed = 220;
int volume = 500; // volume in range 0-100    0=silence
int pitch = 50; //  base pitch, range 0-100.  50=normal


// espeak.cpp 625
espeak_SetParameter(espeakRATE, speed, 0);
espeak_SetParameter(espeakVOLUME,volume,0);
espeak_SetParameter(espeakPITCH,pitch,0);
// espeakRANGE:   pitch range, range 0-100. 0-monotone, 50=normal
// espeakPUNCTUATION:  which punctuation characters to announce:
    // value in espeak_PUNCT_TYPE (none, all, some), 
//espeak_VOICE *voice_spec = espeak_GetCurrentVoice();
//voice_spec->gender=2; // 0=none 1=male, 2=female,
//voice_spec->age = age;

//espeak_SetVoiceByProperties(voice_spec);

//espeak_SetSynthCallback(AndroidEspeakDirectSpeechCallback);
espeak_SetSynthCallback(AndroidEspeakSynthToFileCallback);

unsigned int unique_identifier;
espeak_ERROR err = espeak_Synth( (char*) word, strlen(word)+1, 0, POS_CHARACTER, 0, espeakCHARS_AUTO, &unique_identifier, NULL);

err = espeak_Synchronize();



/*
strcpy(voicename, GetLanguageVoiceName("EN"));
espeak_SetVoiceByName(voicename);
strcpy(word, "Geany was fragged by GSG9 Googlebot");
strcpy(word, "Googlebot");

espeak_Synth( (char*) word, strlen(word)+1, 0, POS_CHARACTER, 0, espeakCHARS_AUTO, NULL, NULL);
espeak_Synchronize();
*/

// espeak_Cancel();
espeak_Terminate();
printf("Espeak terminated\n");
system("pause");
return EXIT_SUCCESS; 
}

Declare Function sndPlaySound Lib "winmm.dll" Alias "sndPlaySoundA" (ByVal lpszSoundName As String, ByVal uFlags As Long) As Long

  sndPlaySound(buffer[0], SND_ASYNC | SND_MEMORY)

  sndPlaySound(filename, SND_ASYNC)