1 |
/***************************************************************************
|
2 |
* Copyright (C) 2005 to 2007 by Jonathan Duddington *
|
3 |
* email: jonsd@users.sourceforge.net *
|
4 |
* *
|
5 |
* This program is free software; you can redistribute it and/or modify *
|
6 |
* it under the terms of the GNU General Public License as published by *
|
7 |
* the Free Software Foundation; either version 3 of the License, or *
|
8 |
* (at your option) any later version. *
|
9 |
* *
|
10 |
* This program is distributed in the hope that it will be useful, *
|
11 |
* but WITHOUT ANY WARRANTY; without even the implied warranty of *
|
12 |
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the *
|
13 |
* GNU General Public License for more details. *
|
14 |
* *
|
15 |
* You should have received a copy of the GNU General Public License *
|
16 |
* along with this program; if not, write see: *
|
17 |
* <http://www.gnu.org/licenses/>. *
|
18 |
***************************************************************************/
|
19 |
|
20 |
#include "StdAfx.h"
|
21 |
|
22 |
#include "speech.h"
|
23 |
|
24 |
#include <stdio.h>
|
25 |
#include <stdlib.h>
|
26 |
#include <string.h>
|
27 |
|
28 |
#ifndef PLATFORM_DOS
|
29 |
#ifdef PLATFORM_WINDOWS
|
30 |
#include <windows.h>
|
31 |
#include <winreg.h>
|
32 |
#else
|
33 |
#include <unistd.h>
|
34 |
#endif
|
35 |
#endif
|
36 |
|
37 |
#ifndef NEED_GETOPT
|
38 |
#include <getopt.h>
|
39 |
#endif
|
40 |
#include <time.h>
|
41 |
#include <signal.h>
|
42 |
#include <locale.h>
|
43 |
#include <sys/stat.h>
|
44 |
|
45 |
#include "speak_lib.h"
|
46 |
#include "phoneme.h"
|
47 |
#include "synthesize.h"
|
48 |
#include "voice.h"
|
49 |
#include "translate.h"
|
50 |
|
51 |
|
52 |
|
53 |
extern void Write4Bytes(FILE *f, int value);
|
54 |
char path_home[N_PATH_HOME]; // this is the espeak-data directory
|
55 |
|
56 |
char filetype[5];
|
57 |
char wavefile[200];
|
58 |
int (* uri_callback)(int, const char *, const char *) = NULL;
|
59 |
int (* phoneme_callback)(const char *) = NULL;
|
60 |
|
61 |
FILE *f_wave = NULL;
|
62 |
int quiet = 0;
|
63 |
unsigned int samples_total = 0;
|
64 |
unsigned int samples_split = 0;
|
65 |
unsigned int wavefile_count = 0;
|
66 |
int end_of_sentence = 0;
|
67 |
|
68 |
static const char *help_text =
|
69 |
"\nspeak [options] [\"<words>\"]\n\n"
|
70 |
"-f <text file> Text file to speak\n"
|
71 |
"--stdin Read text input from stdin instead of a file\n\n"
|
72 |
"If neither -f nor --stdin, <words> are spoken, or if none then text is\n"
|
73 |
"spoken from stdin, each line separately.\n\n"
|
74 |
"-a <integer>\n"
|
75 |
"\t Amplitude, 0 to 200, default is 100\n"
|
76 |
"-g <integer>\n"
|
77 |
"\t Word gap. Pause between words, units of 10mS at the default speed\n"
|
78 |
"-l <integer>\n"
|
79 |
"\t Line length. If not zero (which is the default), consider\n"
|
80 |
"\t lines less than this length as end-of-clause\n"
|
81 |
"-p <integer>\n"
|
82 |
"\t Pitch adjustment, 0 to 99, default is 50\n"
|
83 |
"-s <integer>\n"
|
84 |
"\t Speed in words per minute 80 to 390, default is 170\n"
|
85 |
"-v <voice name>\n"
|
86 |
"\t Use voice file of this name from espeak-data/voices\n"
|
87 |
"-w <wave file name>\n"
|
88 |
"\t Write output to this WAV file, rather than speaking it directly\n"
|
89 |
"-b\t Input text encoding, 1=UTF8, 2=8 bit, 4=16 bit \n"
|
90 |
"-m\t Interpret SSML markup, and ignore other < > tags\n"
|
91 |
"-q\t Quiet, don't produce any speech (may be useful with -x)\n"
|
92 |
"-x\t Write phoneme mnemonics to stdout\n"
|
93 |
"-X\t Write phonemes mnemonics and translation trace to stdout\n"
|
94 |
"-z\t No final sentence pause at the end of the text\n"
|
95 |
"--stdout Write speech output to stdout\n"
|
96 |
"--compile=<voice name>\n"
|
97 |
"\t Compile the pronunciation rules and dictionary in the current\n"
|
98 |
"\t directory. =<voice name> is optional and specifies which language\n"
|
99 |
"--path=\"<path>\"\n"
|
100 |
"\t Specifies the directory containing the espeak-data directory\n"
|
101 |
"--phonout=\"<filename>\"\n"
|
102 |
"\t Write output from -x -X commands and mbrola phoneme data to this file\n"
|
103 |
"--punct=\"<characters>\"\n"
|
104 |
"\t Speak the names of punctuation characters during speaking. If\n"
|
105 |
"\t =<characters> is omitted, all punctuation is spoken.\n"
|
106 |
"--split=\"<minutes>\"\n"
|
107 |
"\t Starts a new WAV file every <minutes>. Used with -w\n"
|
108 |
"--voices=<language>\n"
|
109 |
"\t List the available voices for the specified language.\n"
|
110 |
"\t If <language> is omitted, then list all voices.\n"
|
111 |
"-k <integer>\n"
|
112 |
"\t Indicate capital letters with: 1=sound, 2=the word \"capitals\",\n"
|
113 |
"\t higher values = a pitch increase (try -k20).\n";
|
114 |
|
115 |
|
116 |
void DisplayVoices(FILE *f_out, char *language);
|
117 |
|
118 |
USHORT voice_pcnt[N_PEAKS+1][3];
|
119 |
|
120 |
|
121 |
|
122 |
int GetFileLength(const char *filename)
|
123 |
{//====================================
|
124 |
struct stat statbuf;
|
125 |
|
126 |
if(stat(filename,&statbuf) != 0)
|
127 |
return(0);
|
128 |
|
129 |
if((statbuf.st_mode & S_IFMT) == S_IFDIR)
|
130 |
// if(S_ISDIR(statbuf.st_mode))
|
131 |
return(-2); // a directory
|
132 |
|
133 |
return(statbuf.st_size);
|
134 |
} // end of GetFileLength
|
135 |
|
136 |
|
137 |
char *Alloc(int size)
|
138 |
{//==================
|
139 |
char *p;
|
140 |
if((p = (char *)malloc(size)) == NULL)
|
141 |
fprintf(stderr,"Can't allocate memory\n");
|
142 |
return(p);
|
143 |
}
|
144 |
|
145 |
void Free(void *ptr)
|
146 |
{//=================
|
147 |
if(ptr != NULL)
|
148 |
free(ptr);
|
149 |
}
|
150 |
|
151 |
|
152 |
void DisplayVoices(FILE *f_out, char *language)
|
153 |
{//============================================
|
154 |
int ix;
|
155 |
const char *p;
|
156 |
int len;
|
157 |
int count;
|
158 |
int scores = 0;
|
159 |
const espeak_VOICE *v;
|
160 |
const char *lang_name;
|
161 |
char age_buf[12];
|
162 |
const espeak_VOICE **voices;
|
163 |
espeak_VOICE voice_select;
|
164 |
|
165 |
static char genders[4] = {' ','M','F',' '};
|
166 |
|
167 |
if((language != NULL) && (language[0] != 0))
|
168 |
{
|
169 |
// display only voices for the specified language, in order of priority
|
170 |
voice_select.languages = language;
|
171 |
voice_select.age = 0;
|
172 |
voice_select.gender = 0;
|
173 |
voice_select.name = NULL;
|
174 |
voices = espeak_ListVoices(&voice_select);
|
175 |
scores = 1;
|
176 |
}
|
177 |
else
|
178 |
{
|
179 |
voices = espeak_ListVoices(NULL);
|
180 |
}
|
181 |
|
182 |
fprintf(f_out,"Pty Language Age/Gender VoiceName File Other Langs\n");
|
183 |
|
184 |
for(ix=0; (v = voices[ix]) != NULL; ix++)
|
185 |
{
|
186 |
count = 0;
|
187 |
p = v->languages;
|
188 |
while(*p != 0)
|
189 |
{
|
190 |
len = strlen(p+1);
|
191 |
lang_name = p+1;
|
192 |
|
193 |
if(v->age == 0)
|
194 |
strcpy(age_buf," ");
|
195 |
else
|
196 |
sprintf(age_buf,"%3d",v->age);
|
197 |
|
198 |
if(count==0)
|
199 |
{
|
200 |
fprintf(f_out,"%2d %-12s%s%c %-17s %-11s ",
|
201 |
p[0],lang_name,age_buf,genders[v->gender],v->name,v->identifier);
|
202 |
}
|
203 |
else
|
204 |
{
|
205 |
fprintf(f_out,"(%s %d)",lang_name,p[0]);
|
206 |
}
|
207 |
count++;
|
208 |
p += len+2;
|
209 |
}
|
210 |
// if(scores)
|
211 |
// fprintf(f_out,"%3d ",v->score);
|
212 |
fputc('\n',f_out);
|
213 |
}
|
214 |
} // end of DisplayVoices
|
215 |
|
216 |
|
217 |
|
218 |
|
219 |
static int OpenWaveFile(const char *path, int rate)
|
220 |
//=================================================
|
221 |
{
|
222 |
// Set the length of 0x7ffff000 for --stdout
|
223 |
// This will be changed to the correct length for -w (write to file)
|
224 |
static unsigned char wave_hdr[44] = {
|
225 |
'R','I','F','F',0x24,0xf0,0xff,0x7f,'W','A','V','E','f','m','t',' ',
|
226 |
0x10,0,0,0,1,0,1,0, 9,0x3d,0,0,0x12,0x7a,0,0,
|
227 |
2,0,0x10,0,'d','a','t','a', 0x00,0xf0,0xff,0x7f};
|
228 |
|
229 |
if(path == NULL)
|
230 |
return(2);
|
231 |
|
232 |
if(strcmp(path,"stdout")==0)
|
233 |
f_wave = stdout;
|
234 |
else
|
235 |
f_wave = fopen(path,"wb");
|
236 |
|
237 |
if(f_wave != NULL)
|
238 |
{
|
239 |
fwrite(wave_hdr,1,24,f_wave);
|
240 |
Write4Bytes(f_wave,rate);
|
241 |
Write4Bytes(f_wave,rate * 2);
|
242 |
fwrite(&wave_hdr[32],1,12,f_wave);
|
243 |
return(0);
|
244 |
}
|
245 |
return(1);
|
246 |
} // end of OpenWaveFile
|
247 |
|
248 |
|
249 |
|
250 |
|
251 |
static void CloseWaveFile()
|
252 |
//=========================
|
253 |
{
|
254 |
unsigned int pos;
|
255 |
|
256 |
if((f_wave == NULL) || (f_wave == stdout))
|
257 |
return;
|
258 |
|
259 |
fflush(f_wave);
|
260 |
pos = ftell(f_wave);
|
261 |
|
262 |
fseek(f_wave,4,SEEK_SET);
|
263 |
Write4Bytes(f_wave,pos - 8);
|
264 |
|
265 |
fseek(f_wave,40,SEEK_SET);
|
266 |
Write4Bytes(f_wave,pos - 44);
|
267 |
|
268 |
|
269 |
fclose(f_wave);
|
270 |
f_wave = NULL;
|
271 |
|
272 |
} // end of CloseWaveFile
|
273 |
|
274 |
|
275 |
|
276 |
|
277 |
void MarkerEvent(int type, unsigned int char_position, int value, unsigned char *out_ptr)
|
278 |
{//======================================================================================
|
279 |
// Do nothing in the command-line version.
|
280 |
if(type == 2)
|
281 |
end_of_sentence = 1;
|
282 |
} // end of MarkerEvent
|
283 |
|
284 |
|
285 |
static int WavegenFile(void)
|
286 |
{//=========================
|
287 |
int finished;
|
288 |
unsigned char wav_outbuf[512];
|
289 |
char fname[210];
|
290 |
|
291 |
out_ptr = out_start = wav_outbuf;
|
292 |
out_end = wav_outbuf + sizeof(wav_outbuf);
|
293 |
|
294 |
finished = WavegenFill(0);
|
295 |
|
296 |
if(quiet)
|
297 |
return(finished);
|
298 |
|
299 |
if(f_wave == NULL)
|
300 |
{
|
301 |
sprintf(fname,"%s_%.2d%s",wavefile,++wavefile_count,filetype);
|
302 |
if(OpenWaveFile(fname, samplerate) != 0)
|
303 |
return(1);
|
304 |
}
|
305 |
|
306 |
if(end_of_sentence)
|
307 |
{
|
308 |
end_of_sentence = 0;
|
309 |
if((samples_split > 0 ) && (samples_total > samples_split))
|
310 |
{
|
311 |
CloseWaveFile();
|
312 |
samples_total = 0;
|
313 |
}
|
314 |
}
|
315 |
|
316 |
if(f_wave != NULL)
|
317 |
{
|
318 |
samples_total += (out_ptr - wav_outbuf)/2;
|
319 |
fwrite(wav_outbuf, 1, out_ptr - wav_outbuf, f_wave);
|
320 |
}
|
321 |
return(finished);
|
322 |
} // end of WavegenFile
|
323 |
|
324 |
|
325 |
|
326 |
static void init_path(char *argv0, char *path_specified)
|
327 |
{//=====================================================
|
328 |
|
329 |
if(path_specified)
|
330 |
{
|
331 |
sprintf(path_home,"%s/espeak-data",path_specified);
|
332 |
return;
|
333 |
}
|
334 |
|
335 |
#ifdef PLATFORM_WINDOWS
|
336 |
HKEY RegKey;
|
337 |
unsigned long size;
|
338 |
unsigned long var_type;
|
339 |
char *p;
|
340 |
char *env;
|
341 |
unsigned char buf[sizeof(path_home)-12];
|
342 |
|
343 |
#if 0
|
344 |
if(((env = getenv("ESPEAK_DATA_PATH")) != NULL) && ((strlen(env)+12) < sizeof(path_home)))
|
345 |
{
|
346 |
sprintf(path_home,"%s\\espeak-data",env);
|
347 |
if(GetFileLength(path_home) == -2)
|
348 |
return; // an espeak-data directory exists in the directory specified by environment variable
|
349 |
}
|
350 |
#endif
|
351 |
|
352 |
strcpy(path_home,argv0);
|
353 |
if((p = strrchr(path_home,'\\')) != NULL)
|
354 |
{
|
355 |
strcpy(&p[1],"espeak-data");
|
356 |
if(GetFileLength(path_home) == -2)
|
357 |
return; // an espeak-data directory exists in the same directory as the espeak program
|
358 |
}
|
359 |
|
360 |
// otherwise, look in the Windows Registry
|
361 |
buf[0] = 0;
|
362 |
RegOpenKeyEx(HKEY_LOCAL_MACHINE, "Software\\Microsoft\\Speech\\Voices\\Tokens\\eSpeak", 0, KEY_READ, &RegKey);
|
363 |
size = sizeof(buf);
|
364 |
var_type = REG_SZ;
|
365 |
RegQueryValueEx(RegKey, "path", 0, &var_type, buf, &size);
|
366 |
|
367 |
sprintf(path_home,"%s\\espeak-data",buf);
|
368 |
#else
|
369 |
#ifdef PLATFORM_DOS
|
370 |
strcpy(path_home,PATH_ESPEAK_DATA);
|
371 |
#else
|
372 |
char *env;
|
373 |
#if 0
|
374 |
if((env = getenv("ESPEAK_DATA_PATH")) != NULL)
|
375 |
{
|
376 |
snprintf(path_home,sizeof(path_home),"%s/espeak-data",env);
|
377 |
if(GetFileLength(path_home) == -2)
|
378 |
return; // an espeak-data directory exists
|
379 |
}
|
380 |
#endif
|
381 |
|
382 |
#if 0
|
383 |
snprintf(path_home,sizeof(path_home),"%s/espeak-data",getenv("HOME"));
|
384 |
if(access(path_home,R_OK) != 0)
|
385 |
{
|
386 |
strcpy(path_home,PATH_ESPEAK_DATA);
|
387 |
}
|
388 |
#endif
|
389 |
#endif
|
390 |
#endif
|
391 |
}
|
392 |
|
393 |
|
394 |
static int initialise(void)
|
395 |
{//========================
|
396 |
int param;
|
397 |
int result;
|
398 |
|
399 |
// It seems that the wctype functions don't work until the locale has been set
|
400 |
// to something other than the default "C". Then, not only Latin1 but also the
|
401 |
// other characters give the correct results with iswalpha() etc.
|
402 |
#ifdef PLATFORM_RISCOS
|
403 |
setlocale(LC_CTYPE,"ISO8859-1");
|
404 |
#else
|
405 |
#if 0
|
406 |
if(setlocale(LC_CTYPE,"en_US.UTF-8") == NULL)
|
407 |
{
|
408 |
if(setlocale(LC_CTYPE,"UTF-8") == NULL)
|
409 |
setlocale(LC_CTYPE,"");
|
410 |
}
|
411 |
#endif
|
412 |
#endif
|
413 |
|
414 |
|
415 |
WavegenInit(22050,0); // 22050
|
416 |
if((result = LoadPhData()) != 1)
|
417 |
{
|
418 |
if(result == -1)
|
419 |
{
|
420 |
fprintf(stderr,"Failed to load espeak-data\n");
|
421 |
exit(1);
|
422 |
}
|
423 |
else
|
424 |
fprintf(stderr,"Wrong version of espeak-data 0x%x (expects 0x%x) at %s\n",result,version_phdata,path_home);
|
425 |
}
|
426 |
LoadConfig();
|
427 |
SetVoiceStack(NULL);
|
428 |
SynthesizeInit();
|
429 |
|
430 |
for(param=0; param<N_SPEECH_PARAM; param++)
|
431 |
param_stack[0].parameter[param] = param_defaults[param];
|
432 |
|
433 |
return(0);
|
434 |
}
|
435 |
|
436 |
|
437 |
static void StopSpeak(int unused)
|
438 |
{//==============================
|
439 |
// signal(SIGINT,SIG_IGN);
|
440 |
// DEBUG
|
441 |
// printf("\n*** Interrupting speech output (use Ctrl-D to actually quit).\n");
|
442 |
fflush(stdout);
|
443 |
SpeakNextClause(NULL,NULL,5);
|
444 |
// signal(SIGINT,StopSpeak);
|
445 |
} // end of StopSpeak()
|
446 |
|
447 |
#ifdef NEED_GETOPT
|
448 |
struct option {
|
449 |
char *name;
|
450 |
int has_arg;
|
451 |
int *flag;
|
452 |
int val;
|
453 |
};
|
454 |
int optind;
|
455 |
static int optional_argument;
|
456 |
static const char *arg_opts = "abfgklpsvw"; // which options have arguments
|
457 |
static char *opt_string="";
|
458 |
#define no_argument 0
|
459 |
#define required_argument 1
|
460 |
#define optional_argument 2
|
461 |
#endif
|
462 |
|
463 |
int main (int argc, char **argv)
|
464 |
//==============================
|
465 |
{
|
466 |
static struct option long_options[] =
|
467 |
{
|
468 |
/* These options set a flag. */
|
469 |
// {"verbose", no_argument, &verbose_flag, 1},
|
470 |
// {"brief", no_argument, &verbose_flag, 0},
|
471 |
|
472 |
/* These options don't set a flag.
|
473 |
We distinguish them by their indices. */
|
474 |
{"help", no_argument, 0, 'h'},
|
475 |
{"stdin", no_argument, 0, 0x100},
|
476 |
{"compile-debug", optional_argument, 0, 0x101},
|
477 |
{"compile", optional_argument, 0, 0x102},
|
478 |
{"punct", optional_argument, 0, 0x103},
|
479 |
{"voices", optional_argument, 0, 0x104},
|
480 |
{"stdout", no_argument, 0, 0x105},
|
481 |
{"split", optional_argument, 0, 0x106},
|
482 |
{"path", required_argument, 0, 0x107},
|
483 |
{"phonout", required_argument, 0, 0x108},
|
484 |
{0, 0, 0, 0}
|
485 |
};
|
486 |
|
487 |
static const char *err_load = "Failed to read ";
|
488 |
|
489 |
FILE *f_text=NULL;
|
490 |
const char *p_text=NULL;
|
491 |
char *data_path = NULL; // use default path for espeak-data
|
492 |
|
493 |
int option_index = 0;
|
494 |
int c;
|
495 |
int value;
|
496 |
int speed=170;
|
497 |
int ix;
|
498 |
char *optarg2;
|
499 |
int amp = 100; // default
|
500 |
int wordgap = 0;
|
501 |
int speaking = 0;
|
502 |
int flag_stdin = 0;
|
503 |
int flag_compile = 0;
|
504 |
int pitch_adjustment = 50;
|
505 |
espeak_VOICE voice_select;
|
506 |
char filename[200];
|
507 |
char voicename[40];
|
508 |
char dictname[40];
|
509 |
|
510 |
voicename[0] = 0;
|
511 |
mbrola_name[0] = 0;
|
512 |
dictname[0] = 0;
|
513 |
wavefile[0] = 0;
|
514 |
filename[0] = 0;
|
515 |
option_linelength = 0;
|
516 |
option_phonemes = 0;
|
517 |
option_waveout = 0;
|
518 |
option_wordgap = 0;
|
519 |
option_endpause = 1;
|
520 |
option_phoneme_input = 1;
|
521 |
option_multibyte = espeakCHARS_AUTO; // auto
|
522 |
f_trans = stdout;
|
523 |
|
524 |
#ifdef NEED_GETOPT
|
525 |
optind = 1;
|
526 |
opt_string = "";
|
527 |
while(optind < argc)
|
528 |
{
|
529 |
int len;
|
530 |
char *p;
|
531 |
|
532 |
if((c = *opt_string) == 0)
|
533 |
{
|
534 |
opt_string = argv[optind];
|
535 |
if(opt_string[0] != '-')
|
536 |
break;
|
537 |
|
538 |
optind++;
|
539 |
opt_string++;
|
540 |
c = *opt_string;
|
541 |
}
|
542 |
opt_string++;
|
543 |
p = optarg2 = opt_string;
|
544 |
|
545 |
if(c == '-')
|
546 |
{
|
547 |
opt_string="";
|
548 |
for(ix=0; ;ix++)
|
549 |
{
|
550 |
if(long_options[ix].name == 0)
|
551 |
break;
|
552 |
len = strlen(long_options[ix].name);
|
553 |
if(memcmp(long_options[ix].name,p,len)==0)
|
554 |
{
|
555 |
c = long_options[ix].val;
|
556 |
optarg2 = NULL;
|
557 |
|
558 |
if((long_options[ix].has_arg != 0) && (p[len]=='='))
|
559 |
{
|
560 |
optarg2 = &p[len+1];
|
561 |
}
|
562 |
break;
|
563 |
}
|
564 |
}
|
565 |
}
|
566 |
else
|
567 |
if(strchr(arg_opts,c) != NULL)
|
568 |
{
|
569 |
opt_string="";
|
570 |
if(optarg2[0]==0)
|
571 |
{
|
572 |
// the option's value is in the next argument
|
573 |
optarg2 = argv[optind++];
|
574 |
}
|
575 |
}
|
576 |
#else
|
577 |
while(1)
|
578 |
{
|
579 |
c = getopt_long (argc, argv, "a:b:f:g:hk:l:p:qs:v:w:xXmz", // NOTE: also change arg_opts to indicate which commands have a numeric value
|
580 |
long_options, &option_index);
|
581 |
|
582 |
/* Detect the end of the options. */
|
583 |
if (c == -1)
|
584 |
break;
|
585 |
optarg2 = optarg;
|
586 |
#endif
|
587 |
|
588 |
switch (c)
|
589 |
{
|
590 |
case 'b':
|
591 |
// input character encoding, 8bit, 16bit, UTF8
|
592 |
option_multibyte = espeakCHARS_8BIT;
|
593 |
if((sscanf(optarg2,"%d",&value) == 1) && (value <= 4))
|
594 |
option_multibyte= value;
|
595 |
break;
|
596 |
|
597 |
case 'h':
|
598 |
printf("\nspeak text-to-speech: %s\n%s",version_string,help_text);
|
599 |
exit(0);
|
600 |
break;
|
601 |
|
602 |
case 'k':
|
603 |
option_capitals = atoi(optarg2);
|
604 |
break;
|
605 |
|
606 |
case 'x':
|
607 |
option_phonemes = 1;
|
608 |
break;
|
609 |
|
610 |
case 'X':
|
611 |
option_phonemes = 2;
|
612 |
break;
|
613 |
|
614 |
case 'm':
|
615 |
option_ssml = 1;
|
616 |
break;
|
617 |
|
618 |
case 'p':
|
619 |
pitch_adjustment = atoi(optarg2);
|
620 |
if(pitch_adjustment > 99) pitch_adjustment = 99;
|
621 |
break;
|
622 |
|
623 |
case 'q':
|
624 |
quiet = 1;
|
625 |
break;
|
626 |
|
627 |
case 'f':
|
628 |
strncpy0(filename,optarg2,sizeof(filename));
|
629 |
break;
|
630 |
|
631 |
case 'l':
|
632 |
value = 0;
|
633 |
value = atoi(optarg2);
|
634 |
option_linelength = value;
|
635 |
break;
|
636 |
|
637 |
case 'a':
|
638 |
amp = atoi(optarg2);
|
639 |
break;
|
640 |
|
641 |
case 's':
|
642 |
speed = atoi(optarg2);
|
643 |
break;
|
644 |
|
645 |
case 'g':
|
646 |
wordgap = atoi(optarg2);
|
647 |
break;
|
648 |
|
649 |
case 'v':
|
650 |
strncpy0(voicename,optarg2,sizeof(voicename));
|
651 |
break;
|
652 |
|
653 |
case 'w':
|
654 |
option_waveout = 1;
|
655 |
strncpy0(wavefile,optarg2,sizeof(wavefile));
|
656 |
break;
|
657 |
|
658 |
case 'z':
|
659 |
option_endpause = 0;
|
660 |
break;
|
661 |
|
662 |
case 0x100: // --stdin
|
663 |
flag_stdin = 1;
|
664 |
break;
|
665 |
|
666 |
case 0x105: // --stdout
|
667 |
option_waveout = 1;
|
668 |
strcpy(wavefile,"stdout");
|
669 |
break;
|
670 |
|
671 |
case 0x101: // --compile-debug
|
672 |
case 0x102: // --compile
|
673 |
if(optarg2 != NULL)
|
674 |
strncpy0(voicename,optarg2,sizeof(voicename));
|
675 |
flag_compile = c;
|
676 |
break;
|
677 |
|
678 |
case 0x103: // --punct
|
679 |
option_punctuation = 1;
|
680 |
if(optarg2 != NULL)
|
681 |
{
|
682 |
ix = 0;
|
683 |
while((ix < N_PUNCTLIST) && ((option_punctlist[ix] = optarg2[ix]) != 0)) ix++;
|
684 |
option_punctlist[N_PUNCTLIST-1] = 0;
|
685 |
option_punctuation = 2;
|
686 |
}
|
687 |
break;
|
688 |
|
689 |
case 0x104: // --voices
|
690 |
init_path(argv[0],data_path);
|
691 |
DisplayVoices(stdout,optarg2);
|
692 |
exit(0);
|
693 |
|
694 |
case 0x106: // -- split
|
695 |
if(optarg2 == NULL)
|
696 |
samples_split = 30; // default 30 minutes
|
697 |
else
|
698 |
samples_split = atoi(optarg2);
|
699 |
break;
|
700 |
|
701 |
case 0x107: // --path
|
702 |
data_path = optarg2;
|
703 |
break;
|
704 |
|
705 |
case 0x108: // --phonout
|
706 |
if((f_trans = fopen(optarg2,"w")) == NULL)
|
707 |
{
|
708 |
fprintf(stderr,"Can't write to: %s\n",optarg2);
|
709 |
f_trans = stderr;
|
710 |
}
|
711 |
break;
|
712 |
|
713 |
default:
|
714 |
exit(0);
|
715 |
}
|
716 |
}
|
717 |
|
718 |
init_path(argv[0],data_path);
|
719 |
initialise();
|
720 |
|
721 |
|
722 |
if(flag_compile)
|
723 |
{
|
724 |
LoadVoice(voicename,5);
|
725 |
|
726 |
#ifdef PLATFORM_DOS
|
727 |
char path_dsource[sizeof(path_home)+20];
|
728 |
strcpy(path_dsource,path_home);
|
729 |
path_dsource[strlen(path_home)-11] = 0; // remove "espeak-data" from the end
|
730 |
strcat(path_dsource,"dictsource\\");
|
731 |
CompileDictionary(path_dsource,dictionary_name,NULL,NULL, flag_compile & 0x1);
|
732 |
#else
|
733 |
#ifdef PLATFORM_WINDOWS
|
734 |
char path_dsource[sizeof(path_home)+20];
|
735 |
strcpy(path_dsource,path_home);
|
736 |
path_dsource[strlen(path_home)-11] = 0; // remove "espeak-data" from the end
|
737 |
strcat(path_dsource,"dictsource\\");
|
738 |
CompileDictionary(path_dsource,dictionary_name,NULL,NULL, flag_compile & 0x1);
|
739 |
#else
|
740 |
CompileDictionary(NULL,dictionary_name,NULL,NULL, flag_compile & 0x1);
|
741 |
#endif
|
742 |
#endif
|
743 |
exit(0);
|
744 |
}
|
745 |
|
746 |
|
747 |
if(voicename[0] == 0)
|
748 |
strcpy(voicename,"default");
|
749 |
|
750 |
if(SetVoiceByName(voicename) != EE_OK)
|
751 |
{
|
752 |
memset(&voice_select,0,sizeof(voice_select));
|
753 |
voice_select.languages = voicename;
|
754 |
if(SetVoiceByProperties(&voice_select) != EE_OK)
|
755 |
{
|
756 |
fprintf(stderr,"%svoice '%s'\n",err_load,voicename);
|
757 |
exit(2);
|
758 |
}
|
759 |
}
|
760 |
|
761 |
SetParameter(espeakRATE,speed,0);
|
762 |
SetParameter(espeakVOLUME,amp,0);
|
763 |
SetParameter(espeakCAPITALS,option_capitals,0);
|
764 |
SetParameter(espeakPUNCTUATION,option_punctuation,0);
|
765 |
SetParameter(espeakWORDGAP,wordgap,0);
|
766 |
|
767 |
if(pitch_adjustment != 50)
|
768 |
{
|
769 |
SetParameter(espeakPITCH,pitch_adjustment,0);
|
770 |
}
|
771 |
DoVoiceChange(voice);
|
772 |
|
773 |
if(filename[0]==0)
|
774 |
{
|
775 |
if((optind < argc) && (flag_stdin == 0))
|
776 |
{
|
777 |
// there's a non-option parameter, and no -f or --stdin
|
778 |
// use it as text
|
779 |
p_text = argv[optind];
|
780 |
}
|
781 |
else
|
782 |
{
|
783 |
f_text = stdin;
|
784 |
if(flag_stdin == 0)
|
785 |
option_linelength = -1; // single input lines on stdin
|
786 |
}
|
787 |
}
|
788 |
else
|
789 |
{
|
790 |
f_text = fopen(filename,"r");
|
791 |
}
|
792 |
|
793 |
if((f_text == NULL) && (p_text == NULL))
|
794 |
{
|
795 |
fprintf(stderr,"%sfile '%s'\n",err_load,filename);
|
796 |
exit(1);
|
797 |
}
|
798 |
|
799 |
if(option_waveout || quiet)
|
800 |
{
|
801 |
if(quiet)
|
802 |
{
|
803 |
// no sound output
|
804 |
OpenWaveFile(NULL,samplerate);
|
805 |
option_waveout = 1;
|
806 |
}
|
807 |
else
|
808 |
{
|
809 |
// write sound output to a WAV file
|
810 |
samples_split = (samplerate * samples_split) * 60;
|
811 |
|
812 |
if(samples_split)
|
813 |
{
|
814 |
// don't open the wav file until we start generating speech
|
815 |
char *extn;
|
816 |
extn = strrchr(wavefile,'.');
|
817 |
if((extn != NULL) && ((wavefile + strlen(wavefile) - extn) <= 4))
|
818 |
{
|
819 |
strcpy(filetype,extn);
|
820 |
*extn = 0;
|
821 |
}
|
822 |
}
|
823 |
else
|
824 |
if(OpenWaveFile(wavefile,samplerate) != 0)
|
825 |
{
|
826 |
fprintf(stderr,"Can't write to output file '%s'\n'",wavefile);
|
827 |
exit(3);
|
828 |
}
|
829 |
}
|
830 |
|
831 |
InitText(0);
|
832 |
SpeakNextClause(f_text,p_text,0);
|
833 |
|
834 |
ix = 1;
|
835 |
for(;;)
|
836 |
{
|
837 |
if(WavegenFile() != 0)
|
838 |
{
|
839 |
if(ix == 0)
|
840 |
break; // finished, wavegen command queue is empty
|
841 |
}
|
842 |
|
843 |
if(Generate(phoneme_list,&n_phoneme_list,1)==0)
|
844 |
{
|
845 |
ix = SpeakNextClause(NULL,NULL,1);
|
846 |
}
|
847 |
}
|
848 |
|
849 |
CloseWaveFile();
|
850 |
}
|
851 |
else
|
852 |
{
|
853 |
// Silence on ^C or SIGINT
|
854 |
// signal(SIGINT,StopSpeak);
|
855 |
|
856 |
// output sound using portaudio
|
857 |
WavegenInitSound();
|
858 |
|
859 |
InitText(0);
|
860 |
SpeakNextClause(f_text,p_text,0);
|
861 |
|
862 |
if(option_quiet)
|
863 |
{
|
864 |
while(SpeakNextClause(NULL,NULL,1) != 0);
|
865 |
return(0);
|
866 |
}
|
867 |
|
868 |
#ifdef USE_PORTAUDIO
|
869 |
speaking = 1;
|
870 |
while(speaking)
|
871 |
{
|
872 |
// NOTE: if nanosleep() isn't recognised on your system, try replacing
|
873 |
// this by sleep(1);
|
874 |
#ifdef PLATFORM_WINDOWS
|
875 |
Sleep(300); // 0.3s
|
876 |
#else
|
877 |
#ifdef USE_NANOSLEEP
|
878 |
struct timespec period;
|
879 |
struct timespec remaining;
|
880 |
period.tv_sec = 0;
|
881 |
period.tv_nsec = 300000000; // 0.3 sec
|
882 |
nanosleep(&period,&remaining);
|
883 |
#else
|
884 |
sleep(1);
|
885 |
#endif
|
886 |
#endif
|
887 |
if(SynthOnTimer() != 0)
|
888 |
speaking = 0;
|
889 |
}
|
890 |
#else
|
891 |
fprintf(stderr,"-w option must be used because the program was built without a sound interface\n");
|
892 |
#endif // USE_PORTAUDIO
|
893 |
}
|
894 |
|
895 |
if((f_trans != stdout) && (f_trans != stderr))
|
896 |
fclose(f_trans); // needed for WinCe
|
897 |
return(0);
|
898 |
}
|