/***************************************************************************
* Copyright (C) 2005 to 2007 by Jonathan Duddington *
* email: jonsd@users.sourceforge.net *
* *
* This program is free software; you can redistribute it and/or modify *
* it under the terms of the GNU General Public License as published by *
* the Free Software Foundation; either version 3 of the License, or *
* (at your option) any later version. *
* *
* This program is distributed in the hope that it will be useful, *
* but WITHOUT ANY WARRANTY; without even the implied warranty of *
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the *
* GNU General Public License for more details. *
* *
* You should have received a copy of the GNU General Public License *
* along with this program; if not, write see: *
* . *
***************************************************************************/
#define USE_MODULE
#include
#include
#include
#include
#include
#include "kernel.h"
#include "speech.h"
#include "speak_lib.h"
#include "phoneme.h"
#include "synthesize.h"
#include "voice.h"
#include "translate.h"
#define os_X 0x20000
// interface to the assembler section
extern "C" {
extern void DMA_Handler(void);
// used from the cmhgfile
extern _kernel_oserror *user_init(char *cmd_fail, int podule_base, void *pw);
extern _kernel_oserror *swi_handler(int swi_no, int *r, void *pw);
extern _kernel_oserror *command_handler(char *arg_string, int argc, int cmd_no, void *pw);
extern int callback_handler(_kernel_swi_regs *r, void *pw);
extern int callback_entry(_kernel_swi_regs *r, void *pw);
extern int sound_handler(_kernel_swi_regs *r, void *pw);
extern int sound_entry(_kernel_swi_regs *r, void *pw);
}
extern int Generate(PHONEME_LIST *phoneme_list, int *n_ph, int resume);
extern void RiscosOpenSound();
extern int WcmdqUsed();
extern void FreePhData();
extern void FreeDictionary();
extern int wcmdq_head;
extern int wcmdq_tail;
extern int current_source_index;
FILE *f_text;
int (* uri_callback)(int, const char *, const char *) = NULL;
int (* phoneme_callback)(const char *) = NULL;
int amp = 8; // default
char path_home[N_PATH_HOME] = "";
char wavefile[120];
char textbuffile[L_tmpnam];
int sample_rate_index; // current value
int current_voice_num=0;
int n_voice_files = 0;
// output sound buffer, 2 bytes per sample
static unsigned short SoundBuf[4096];
static void *module_data;
static int callback_inhibit = 0;
static int more_text=0;
#define N_VOICE_NAMES 60
static char *voice_names[40];
#define N_STATIC_BUF 8000
static char static_buf[N_STATIC_BUF];
static _kernel_oserror errblk;
USHORT voice_pcnt[N_PEAKS+1][3];
static const char *help_text =
"\nspeak [options] [\"\"]\n\n"
"-f Text file to speak\n"
//"--stdin Read text input from stdin instead of a file\n\n"
"If neither -f nor --stdin, are spoken, or if none then text is\n"
"spoken from stdin, each line separately.\n\n"
"-a \n"
"\t Amplitude, 0 to 200, default is 100\n"
"-g \n"
"\t Word gap. Pause between words, units of 10mS at the default speed\n"
"-l \n"
"\t Line length. If not zero (which is the default), consider\n"
"\t lines less than this length as end-of-clause\n"
"-p \n"
"\t Pitch adjustment, 0 to 99, default is 50\n"
"-s \n"
"\t Speed in words per minute 80 to 370, default is 170\n"
"-v \n"
"\t Use voice file of this name from espeak-data/voices\n"
"-w \n"
"\t Write output to this WAV file, rather than speaking it directly\n"
"-b\t Input text is 8-bit encoding\n"
"-m\t Interpret SSML markup, and ignore other < > tags\n"
"-q\t Quiet, don't produce any speech (may be useful with -x)\n"
"-x\t Write phoneme mnemonics to stdout\n"
"-X\t Write phonemes mnemonics and translation trace to stdout\n"
//"--stdout Write speech output to stdout\n"
"--compile=\n"
"\t Compile the pronunciation rules and dictionary in the current\n"
"\t directory. = is optional and specifies which language\n"
"--punct=\"\"\n"
"\t Speak the names of punctuation characters during speaking. If\n"
"\t = is omitted, all punctuation is spoken.\n"
//"--voices=\n"
//"\t List the available voices for the specified language.\n"
//"\t If is omitted, then list all voices.\n"
"-k \n"
"\t Indicate capital letters with: 1=sound, 2=the word \"capitals\",\n"
"\t higher values = a pitch increase (try -k20).\n";
int GetFileLength(const char *filename)
{//====================================
int length=0;
int type;
_kernel_swi_regs regs;
_kernel_oserror *error;
regs.r[0] = 5;
regs.r[1] = (int)filename;
regs.r[2] = 0;
regs.r[3] = 0;
regs.r[4] = 0;
regs.r[5] = 0;
error = _kernel_swi(0x20008,®s,®s);
if(error)
return(0);
type = regs.r[0];
length = regs.r[4];
if(type==2)
return(-2); // a directory
if((type!=1) && (type!=3))
return(0); /* not a file */
return(length);
} /* end of GetFileLength */
void ReadVoiceNames2(char *directory)
{//==================================
int len;
int *type;
char *p;
_kernel_swi_regs regs;
_kernel_oserror *error;
char buf[80];
char directory2[sizeof(path_home)+100];
regs.r[0] = 10;
regs.r[1] = (int)directory;
regs.r[2] = (int)buf;
regs.r[3] = 1;
regs.r[4] = 0;
regs.r[5] = sizeof(buf);
regs.r[6] = 0;
while(regs.r[3] > 0)
{
error = _kernel_swi(0x0c+os_X,®s,®s); /* OS_GBPB 10, read directory entries */
if((error != NULL) || (regs.r[3] == 0))
{
break;
}
type = (int *)(&buf[16]);
len = strlen(&buf[20]);
if(*type == 2)
{
// a sub-directory
sprintf(directory2,"%s.%s",directory,&buf[20]);
ReadVoiceNames2(directory2);
}
else
{
p = Alloc(len+1);
strcpy(p,&buf[20]);
voice_names[n_voice_files++] = p;
if(n_voice_files >= (N_VOICE_NAMES-1))
break;
}
}
}
void ReadVoiceNames()
{//===================
char directory[sizeof(path_home)+10];
for(n_voice_files=0; n_voice_files sizeof(SoundBuf))
size = sizeof(SoundBuf);
out_ptr = (unsigned char *)(&SoundBuf[0]);
out_end = (unsigned char *)(&SoundBuf[size]);
WavegenFill(1);
}
int initialise(void)
{//=================
sprintf(path_home,"%s.espeak-data","");
if(GetFileLength(path_home) != -2)
{
// not found, try the 10 character version of the directory name
sprintf(path_home,"%s.espeak-dat","");
}
if(GetFileLength(path_home) != -2)
{
// still can't find data directory
sprintf(errblk.errmess,"Speak: Can't find data directory: '%s'\n",path_home);
return(-1);
}
WavegenInit(22050,0);
LoadPhData();
SetVoiceStack(NULL);
SynthesizeInit();
return(0);
}
void speak_text_string(char *data, int terminator, int len, int wait, int voice_num)
/**********************************************************************************/
/* 'wait' indictes wait until speaking is finished before returning */
{
int c;
int ix;
static static_length=0;
static int user_token=0; /* increment for each call of translate() */
_kernel_swi_regs regs;
regs.r[0] = (int)callback_entry;
regs.r[1] = (int)module_data;
_kernel_swi(0x5f, ®s, ®s);
if((voice_num >= 0) && (voice_num != current_voice_num) && (voice_num < N_VOICE_NAMES))
{
SetVoiceByName(voice_names[voice_num]);
WavegenSetVoice(voice);
}
current_voice_num = voice_num;
/* don't count CR as terminator if length is specified */
if(len > 0) terminator = 0;
ix = 0;
if(more_text == 0)
static_length = 0;
else
{
strcat(&static_buf[static_length]," : ");
static_length+=3;
}
if(terminator == 0)
{
while(((c = data[ix++]) != 0) && (static_length < N_STATIC_BUF-4))
{
static_buf[static_length++] = c;
if(ix == len)
break;
}
}
else
{
while(((c = data[ix++]) != '\r') && (c != '\n') && (c != 0) && (static_length < N_STATIC_BUF-3))
{
static_buf[static_length++] = c;
if(ix == len)
break;
}
}
static_buf[static_length] = 0;
if(option_waveout==0)
{
if(more_text == 0)
{
InitText(0);
RiscosOpenSound();
more_text = SpeakNextClause(NULL,(void *)static_buf,0);
}
while(wait)
{
if((more_text==0) && (wcmdq_head == wcmdq_tail))
break;
//we need to block to allow the callback handler to run
regs.r[0] = 129; // wait for key press
regs.r[1] = 10;
regs.r[2] = 0;
_kernel_swi(0x06, ®s, ®s); // OS_Byte
}
}
else
{
more_text = 0;
SpeakNextClause(NULL,(void *)static_buf,0);
for(;;)
{
if(WavegenFile() != 0)
break; // finished, wavegen command queue is empty
if(Generate(phoneme_list,&n_phoneme_list,1)==0)
SpeakNextClause(NULL,NULL,1);
}
CloseWaveFile(samplerate);
}
} /* end of speak_text_string */
void speak_file(char *fname)
{//=========================
FILE *f_in;
char buf[120];
f_in = fopen(fname,"r");
if(f_in == NULL)
{
fprintf(stderr,"Can't read file: '%s'",fname);
return;
}
more_text = 1;
if(option_waveout == 0)
{
RiscosOpenSound();
SpeakNextClause(f_in,NULL,0);
}
else
{
more_text = 0;
SpeakNextClause(f_in,NULL,0);
for(;;)
{
if(WavegenFile() != 0)
break; // finished, wavegen command queue is empty
if(Generate(phoneme_list,&n_phoneme_list,1)==0)
SpeakNextClause(NULL,NULL,1);
}
CloseWaveFile(samplerate);
}
}
void set_say_options(int reg2, int reg3)
/**************************************/
/* Sets options from information in 'say' SWI */
/* R3 bits 0-7 stress indicator character
bit 8 inhibit unpronouncable check */
{
option_linelength = 0;
option_phonemes = 0;
option_waveout = 0;
option_harmonic1 = 8;
option_multibyte = 0; // auto
option_capitals = 0;
option_punctuation = 0;
option_punctlist[0] = 0;
} /* end of set_say_options */
void jsd_swi_functions(int *r)
/****************************/
{
switch(r[0])
{
case 0: /* major version */
r[0] = 3;
break;
case 1: /* register user */
break;
case 2: /* deregister user */
break;
case 3:
// r[0] = (int)speech_to_phonemes((char *)r[1]);
break;
case 4:
// r[0] = reload_word_dict(NULL);
break;
case 5: /* get table of voice names */
r[0] = (int)voice_names;
break;
case 6: /* update voice data, r1 = voice_number */
if(r[1] < N_VOICE_NAMES)
{
SetVoiceByName(voice_names[r[1]]);
current_voice_num = r[1];
WavegenSetVoice(voice);
}
break;
case 7: /* load voice data */
// init_voice((char *)r[1]);
break;
default:
r[0] = 0;
r[1] = 0;
break;
}
} /* end of jsd_swi_functions */
_kernel_oserror *swi_handler(int swi_no, int *r, void *pw)
/*********************************************************/
{
int value;
int q_length;
int speed;
int amp;
value = r[0];
switch(swi_no)
{
case 0: // ready ?
// returns the index into the source text of the currently speaking word
if(current_source_index > 0)
r[1] = current_source_index-1;
else
r[1] = current_source_index; /* source index */
r[2] = 0; /* source tag */
r[3] = 0; /* for future expansion */
r[4] = 0;
r[5] = 0;
if(wcmdq_head == wcmdq_tail)
{
r[0] = -1; /* ready, or nearly */
}
else
r[0] = 0;
break;
case 1: /* restore old sound channel. DO NOTHING */
break;
case 2: /* miscellaneous functions */
jsd_swi_functions(r);
break;
case 3: /* speak text */
// _kernel_irqs_on();
set_say_options(r[2],r[3]);
speak_text_string((char *)r[0],'\r',r[1],0,r[2]);
break;
case 4: /* speak text and wait */
// _kernel_irqs_on(); /* enable interrupts */
set_say_options(r[2],r[3]);
speak_text_string((char *)r[0],'\r',r[1],1,r[2]);
break;
case 5: /* stop speaking */
SpeakNextClause(NULL,NULL,2);
more_text = 0;
break;
case 7: /* pitch */
// not implemented
break;
case 8: /* speed */
speed = (value * 160)/140;
SetParameter(espeakRATE,speed,0);
break;
case 9: /* word_gap */
SetParameter(espeakWORDGAP,value,0);
break;
case 10: /* pitch_range */
// not implemented
break;
case 12: /* reset */
// not implemented
break;
case 13: /* volume */
amp = (value*45)/100;
SetParameter(espeakVOLUME,amp,0);
WavegenSetVoice(voice);
break;
}
return(NULL);
} /* end of swi_handler */
void PitchAdjust(int pitch_adjustment)
{//===================================
int ix, factor;
extern unsigned char pitch_adjust_tab[MAX_PITCH_VALUE+1];
voice->pitch_base = (voice->pitch_base * pitch_adjust_tab[pitch_adjustment])/128;
// adjust formants to give better results for a different voice pitch
factor = 256 + (25 * (pitch_adjustment - 50))/50;
for(ix=0; ix<=5; ix++)
{
voice->freq[ix] = (voice->freq2[ix] * factor)/256;
}
} // end of PitchAdjustment
char *param_string(char **argp)
{//============================
char *p;
int ix=0;
static char buf[80];
p = *argp;
while(*p == ' ') p++;
while(!isspace(*p))
buf[ix++] = *p++;
buf[ix]=0;
*argp = p;
return(buf);
}
int param_number(char **argp)
{//==========================
int value;
char *p;
p = *argp;
while(*p == ' ') p++;
value = atoi(p);
while(!isspace(*p)) p++;
*argp = p;
return(value);
}
void command_line(char *arg_string, int wait)
{//==========================================
int option_index = 0;
int c;
int value;
int speed;
int amp;
int wordgap;
int speaking = 0;
int flag_stdin = 0;
int flag_compile = 0;
int error;
int pitch_adjustment = 50;
char filename[80];
char voicename[40];
char command[80];
char *p;
int ix;
voicename[0] = 0;
wavefile[0] = 0;
filename[0] = 0;
option_linelength = 0;
option_phonemes = 0;
option_waveout = 0;
option_quiet = 0;
option_harmonic1 = 8;
option_multibyte = 0; // auto
option_capitals = 0;
option_punctuation = 0;
option_punctlist[0] = 0;
f_trans = stdout;
p = arg_string;
for(;;)
{
while(*p==' ') p++; // skip spaces
if(*p == '\r') break; // end of line
if(*p == '-')
{
// a command line argument
p++;
switch(*p++)
{
case 'b':
option_multibyte = espeakCHARS_8BIT;
break;
case 'h':
printf("\nspeak text-to-speech: %s\n%s",version_string,help_text);
return;
case 'k':
option_capitals = param_number(&p);
SetParameter(espeakCAPITALS,option_capitals,0);
break;
case 'x':
option_phonemes = 1;
break;
case 'X':
option_phonemes = 2;
break;
case 'm':
option_ssml = 1;
break;
case 'p':
pitch_adjustment = param_number(&p);
break;
case 'q':
option_quiet = 1;
break;
case 'f':
strncpy0(filename,param_string(&p),sizeof(filename));
break;
case 'l':
option_linelength = param_number(&p);
break;
case 'a':
amp = param_number(&p);
SetParameter(espeakVOLUME,amp,0);
break;
case 's':
speed = param_number(&p);
SetParameter(espeakRATE,speed,0);
break;
case 'g':
wordgap = param_number(&p);
SetParameter(espeakWORDGAP,wordgap,0);
break;
case 'v':
strncpy0(voicename,param_string(&p),sizeof(voicename));
break;
case 'w':
option_waveout=1;
strncpy0(wavefile,param_string(&p),sizeof(wavefile));
break;
case '-':
strncpy0(command,param_string(&p),sizeof(command));
if(memcmp(command,"compile=",8)==0)
{
CompileDictionary(NULL,&command[8],NULL,NULL);
return;
}
else
if(strcmp(command,"help")==0)
{
printf("\nspeak text-to-speech: %s\n%s",version_string,help_text);
return;
}
else
if(memcmp(command,"punct",5)==0)
{
option_punctuation = 1;
if((command[5]=='=') && (command[6]=='"'))
{
ix = 0;
while((ix < N_PUNCTLIST) && ((option_punctlist[ix] = command[ix+7]) != 0)) ix++;
option_punctlist[N_PUNCTLIST-1] = 0;
option_punctuation = 2;
}
SetParameter(espeakPUNCTUATION,option_punctuation,0);
}
else
{
printf("Command not recognised\n");
}
break;
default:
printf("Command not recognised\n");
break;
}
}
else
{
break;
}
}
SetVoiceByName(voicename);
if((filename[0]==0) && (p[0]=='\r'))
{
// nothing to speak
if(option_quiet)
{
SpeakNextClause(NULL,NULL,2); // stop speaking
more_text = 0;
}
}
if(option_waveout || option_quiet)
{
// write speech to a WAV file
if(option_quiet)
{
OpenWaveFile(NULL,samplerate);
option_waveout = 2;
}
else
{
if(OpenWaveFile(wavefile,samplerate) != 0)
{
fprintf(stderr,"Can't write to output file '%s'\n'",wavefile);
return;
}
}
}
if(pitch_adjustment != 50)
{
PitchAdjust(pitch_adjustment);
}
WavegenSetVoice(voice);
if(filename[0]==0)
speak_text_string(p,'\r',0,wait,-1);
else
speak_file(filename);
}
_kernel_oserror *command_handler(char *arg_string, int argc, int cmd_no, void *pw)
/********************************************************************************/
{
switch(cmd_no)
{
case 0: /* Say */
command_line(arg_string,0); // for compatibility with speak V2
break;
case 1: /* Sayw ] */
command_line(arg_string,0);
break;
}
return(NULL);
} /* end of cmd_handler */
// sound handler data
int current_sound_handler=0;
int prev_sound_handler=0;
int prev_sound_data=0;
int prev_sound_rate=13;
int sound_handler_changed=0;
void RiscosCloseSound()
{//====================
_kernel_swi_regs regs;
if((sound_handler_changed) && (prev_sound_handler != (int)DMA_Handler))
{
// check whether current handler is ours
regs.r[0]=0;
_kernel_swi(0x40145,®s,®s);
if(regs.r[1] == (int)DMA_Handler)
{
regs.r[0]=1;
regs.r[1]=prev_sound_handler;
regs.r[2]=prev_sound_data;
_kernel_swi(0x40145,®s,®s); // Sound LinearHandler 1
// reset to the previous sample rate
regs.r[0]=3;
regs.r[1]=prev_sound_rate;
_kernel_swi(0x40146,®s,®s); // Sound_SampleRate 3
current_sound_handler = prev_sound_handler;
sound_handler_changed = 0;
}
}
} // end of RiscosCloseSound
void RiscosOpenSound()
{//===================
_kernel_swi_regs regs;
if(current_sound_handler != (int)DMA_Handler)
{
// register the sound handler
regs.r[0]=1;
regs.r[1]=(int)DMA_Handler;
regs.r[2]=(int)module_data;
_kernel_swi(0x40145,®s,®s); // Sound_LinearHandler 1
prev_sound_handler = regs.r[1];
prev_sound_data = regs.r[2];
// set the sample rate
regs.r[0]=3;
regs.r[1]=sample_rate_index;
regs.r[2]=0;
_kernel_swi(0x40146,®s,®s); // Sound_SampleRate
prev_sound_rate = regs.r[1];
current_sound_handler = (int)DMA_Handler;
sound_handler_changed = 1;
}
} // end of RiscosOpenSound
int callback_handler(_kernel_swi_regs *r, void *pw)
/*************************************************/
{
if(Generate(phoneme_list,&n_phoneme_list,1)==0)
{
more_text = SpeakNextClause(NULL,NULL,1);
}
if((WcmdqUsed() == 0) && (more_text == 0))
{
RiscosCloseSound();
}
callback_inhibit = 0;
return(1);
} /* end of callback_handler */
int sound_handler(_kernel_swi_regs *r, void *pw)
/**********************************************/
{
int n_queue;
int size;
int *dma_buf;
int x;
int ix;
module_data = (int *)pw;
dma_buf = (int *)r->r[1];
size = (r->r[2] - r->r[1])/4;
FillSoundBuf(size);
for(ix=0; ixr[0] = 0;
if(callback_inhibit == 0)
{
// set a callback either:
// - queue is low and there is more text to be processed
// - queue is empty and no more text, so callback handler will remove the sound handler
if(((n_queue < 20) && (more_text != 0)) ||
((n_queue==0) && (more_text == 0)))
{
callback_inhibit = 1;
r->r[0] = 1;
r->r[1] = (int)pw;
}
}
return(1);
}
int InitSound16(int sample_rate)
/******************************/
/* Find sample rate index */
{
int current_rate_index; // current value
int sound_mode;
int sound_config;
int srate;
int n_srix;
int ix;
_kernel_swi_regs regs;
_kernel_oserror *error;
sound_mode = 0;
regs.r[0] = 0;
error = _kernel_swi(0x40144+os_X,®s,®s);
sound_mode = regs.r[0];
sound_config = regs.r[1];
if((error == NULL) && (sound_mode == 1))
{
/* 16 bit sound, find sample rate index */
regs.r[0] = 0;
regs.r[1] = 0;
_kernel_swi(0x40146,®s,®s);
n_srix = regs.r[1];
regs.r[0] = 1;
regs.r[1] = 0;
_kernel_swi(0x40146,®s,®s);
current_rate_index = regs.r[1]; // current sample rate index
srate = regs.r[2];
for(ix=1; ix<=n_srix; ix++)
{
regs.r[0] = 2;
regs.r[1] = ix;
_kernel_swi(0x40146,®s,®s);
srate = regs.r[2];
if(srate >= (sample_rate*1024))
{
return(ix);
}
}
}
return(14); // this was the index for 22050
} // end of InitSound16
void RemoveCallback()
/*******************/
{
_kernel_swi_regs regs;
regs.r[0] = (int)callback_entry;
regs.r[1] = (int)module_data;
_kernel_swi(0x5f, ®s, ®s);
}
void terminate_module(void)
/*************************/
{
RiscosCloseSound();
RemoveCallback();
delete translator;
FreePhData();
} /* end of terminate_module */
void kill_module(void)
/********************/
{
_kernel_swi_regs regs;
regs.r[0]=4;
regs.r[1]=(int)"Speak";
_kernel_swi(0x1e,®s,®s); /* RMKill */
}
_kernel_oserror *user_init(char *cmd_fail, int podule_base, void *pw)
/*******************************************************************/
{
_kernel_swi_regs regs;
_kernel_oserror *error;
int param;
// It seems that the wctype functions don't work until the locale has been set
// to something other than the default "C". Then, not only Latin1 but also the
// other characters give the correct results with iswalpha() etc.
static char *locale = "ISO8859-1";
setlocale(LC_CTYPE,locale);
module_data = pw;
sample_rate_index = InitSound16(22050);
if(initialise() < 0)
{
// exit module, errblk.errmess is set by initialise()
errblk.errnum = 0x101;
return(&errblk);
}
ReadVoiceNames();
SetVoiceByName("default");
for(param=0; param