eSpeak NG is an open source speech synthesizer that supports more than hundred languages and accents.
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

espeak-ng.c 22KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797
  1. /*
  2. * Copyright (C) 2006 to 2013 by Jonathan Duddington
  3. * email: [email protected]
  4. * Copyright (C) 2015-2016 Reece H. Dunn
  5. *
  6. * This program is free software; you can redistribute it and/or modify
  7. * it under the terms of the GNU General Public License as published by
  8. * the Free Software Foundation; either version 3 of the License, or
  9. * (at your option) any later version.
  10. *
  11. * This program is distributed in the hope that it will be useful,
  12. * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13. * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  14. * GNU General Public License for more details.
  15. *
  16. * You should have received a copy of the GNU General Public License
  17. * along with this program; if not, see: <http://www.gnu.org/licenses/>.
  18. */
  19. #include "config.h"
  20. #include <ctype.h>
  21. #include <errno.h>
  22. #include <getopt.h>
  23. #include <stdbool.h>
  24. #include <stdint.h>
  25. #include <stdio.h>
  26. #include <stdlib.h>
  27. #include <string.h>
  28. #include <sys/stat.h>
  29. #include <fcntl.h>
  30. #include <time.h>
  31. #include <espeak-ng/espeak_ng.h>
  32. #include <espeak-ng/speak_lib.h>
  33. #ifndef PROGRAM_NAME
  34. #define PROGRAM_NAME "espeak-ng"
  35. #endif
  36. #ifndef PLAYBACK_MODE
  37. #define PLAYBACK_MODE ENOUTPUT_MODE_SPEAK_AUDIO
  38. #endif
  39. extern ESPEAK_NG_API void strncpy0(char *to, const char *from, int size);
  40. extern ESPEAK_NG_API int utf8_in(int *c, const char *buf);
  41. extern ESPEAK_NG_API int GetFileLength(const char *filename);
  42. static const char *help_text =
  43. "\n" PROGRAM_NAME " [options] [\"<words>\"]\n\n"
  44. "-f <text file> Text file to speak\n"
  45. "--stdin Read text input from stdin at once till to the end of a stream.\n\n"
  46. "If neither -f nor --stdin are provided, then <words> from arguments are spoken,\n"
  47. "or text is spoken from stdin, read separately one line by line at a time.\n\n"
  48. "-a <integer>\n"
  49. "\t Amplitude, 0 to 200, default is 100\n"
  50. "-d <device>\n"
  51. "\t Use the specified device to speak the audio on. If not specified, the\n"
  52. "\t default audio device is used.\n"
  53. "-g <integer>\n"
  54. "\t Word gap. Pause between words, units of 10mS at the default speed\n"
  55. "-k <integer>\n"
  56. "\t Indicate capital letters with: 1=sound, 2=the word \"capitals\",\n"
  57. "\t higher values indicate a pitch increase (try -k20).\n"
  58. "-l <integer>\n"
  59. "\t Line length. If not zero (which is the default), consider\n"
  60. "\t lines less than this length as end-of-clause\n"
  61. "-p <integer>\n"
  62. "\t Pitch adjustment, 0 to 99, default is 50\n"
  63. "-P <integer>\n"
  64. "\t Pitch range adjustment, 0 to 99, default is 50\n"
  65. "-s <integer>\n"
  66. "\t Speed in approximate words per minute. The default is 175\n"
  67. "-v <voice name>\n"
  68. "\t Use voice file of this name from espeak-ng-data/voices\n"
  69. "-w <wave file name>\n"
  70. "\t Write speech to this WAV file, rather than speaking it directly\n"
  71. "-b\t Input text encoding, 1=UTF8, 2=8 bit, 4=16 bit \n"
  72. "-m\t Interpret SSML markup, and ignore other < > tags\n"
  73. "--ssml-break=<percentage>\n"
  74. "\t Set SSML break time multiplier, default is 100\n"
  75. "-q\t Quiet, don't produce any speech (may be useful with -x)\n"
  76. "-x\t Write phoneme mnemonics to stdout\n"
  77. "-X\t Write phonemes mnemonics and translation trace to stdout\n"
  78. "-z\t No final sentence pause at the end of the text\n"
  79. "-D\t Enable deterministic random mode\n"
  80. "--compile=<voice name>\n"
  81. "\t Compile pronunciation rules and dictionary from the current\n"
  82. "\t directory. <voice name> specifies the language\n"
  83. "--compile-debug=<voice name>\n"
  84. "\t Compile pronunciation rules and dictionary from the current\n"
  85. "\t directory, including line numbers for use with -X.\n"
  86. "\t <voice name> specifies the language\n"
  87. #if USE_MBROLA
  88. "--compile-mbrola=<voice name>\n"
  89. "\t Compile an MBROLA voice\n"
  90. #endif
  91. "--compile-intonations\n"
  92. "\t Compile the intonation data\n"
  93. "--compile-phonemes=<phsource-dir>\n"
  94. "\t Compile the phoneme data using <phsource-dir> or the default phsource directory\n"
  95. "--ipa Write phonemes to stdout using International Phonetic Alphabet\n"
  96. "--path=\"<path>\"\n"
  97. "\t Specifies the directory containing the espeak-ng-data directory\n"
  98. "--pho Write mbrola phoneme data (.pho) to stdout or to the file in --phonout\n"
  99. "--phonout=\"<filename>\"\n"
  100. "\t Write phoneme output from -x -X --ipa and --pho to this file\n"
  101. "--punct=\"<characters>\"\n"
  102. "\t Speak the names of punctuation characters during speaking. If\n"
  103. "\t =<characters> is omitted, all punctuation is spoken.\n"
  104. "--sep=<character>\n"
  105. "\t Separate phonemes (from -x --ipa) with <character>.\n"
  106. "\t Default is space, z means ZWJN character.\n"
  107. "--split=<minutes>\n"
  108. "\t Starts a new WAV file every <minutes>. Used with -w\n"
  109. "--stdout Write speech output to stdout\n"
  110. "--tie=<character>\n"
  111. "\t Use a tie character within multi-letter phoneme names.\n"
  112. "\t Default is U+361, z means ZWJ character.\n"
  113. "--version Shows version number and date, and location of espeak-ng-data\n"
  114. "--voices=<language>\n"
  115. "\t List the available voices for the specified language.\n"
  116. "\t If <language> is omitted, then list all voices.\n"
  117. "--load Load voice from a file in current directory by name.\n"
  118. "-h, --help Show this help.\n";
  119. static int samplerate;
  120. bool quiet = false;
  121. unsigned int samples_total = 0;
  122. unsigned int samples_split = 0;
  123. unsigned int samples_split_seconds = 0;
  124. unsigned int wavefile_count = 0;
  125. FILE *f_wavfile = NULL;
  126. char filetype[5];
  127. char wavefile[200];
  128. static void DisplayVoices(FILE *f_out, char *language)
  129. {
  130. int ix;
  131. const char *p;
  132. int len;
  133. int count;
  134. int c;
  135. size_t j;
  136. const espeak_VOICE *v;
  137. const char *lang_name;
  138. char age_buf[12];
  139. char buf[80];
  140. const espeak_VOICE **voices;
  141. espeak_VOICE voice_select;
  142. static const char genders[4] = { '-', 'M', 'F', '-' };
  143. if ((language != NULL) && (language[0] != 0)) {
  144. // display only voices for the specified language, in order of priority
  145. voice_select.languages = language;
  146. voice_select.age = 0;
  147. voice_select.gender = 0;
  148. voice_select.name = NULL;
  149. voices = espeak_ListVoices(&voice_select);
  150. } else
  151. voices = espeak_ListVoices(NULL);
  152. fprintf(f_out, "Pty Language Age/Gender VoiceName File Other Languages\n");
  153. for (ix = 0; (v = voices[ix]) != NULL; ix++) {
  154. count = 0;
  155. p = v->languages;
  156. while (*p != 0) {
  157. len = strlen(p+1);
  158. lang_name = p+1;
  159. if (v->age == 0)
  160. strcpy(age_buf, " --");
  161. else
  162. sprintf(age_buf, "%3d", v->age);
  163. if (count == 0) {
  164. for (j = 0; j < sizeof(buf); j++) {
  165. // replace spaces in the name
  166. if ((c = v->name[j]) == ' ')
  167. c = '_';
  168. if ((buf[j] = c) == 0)
  169. break;
  170. }
  171. fprintf(f_out, "%2d %-15s%s/%c %-18s %-20s ",
  172. p[0], lang_name, age_buf, genders[v->gender], buf, v->identifier);
  173. } else
  174. fprintf(f_out, "(%s %d)", lang_name, p[0]);
  175. count++;
  176. p += len+2;
  177. }
  178. fputc('\n', f_out);
  179. }
  180. }
  181. static void Write4Bytes(FILE *f, int value)
  182. {
  183. // Write 4 bytes to a file, least significant first
  184. int ix;
  185. for (ix = 0; ix < 4; ix++) {
  186. fputc(value & 0xff, f);
  187. value = value >> 8;
  188. }
  189. }
  190. static int OpenWavFile(char *path, int rate)
  191. {
  192. static const unsigned char wave_hdr[44] = {
  193. 'R', 'I', 'F', 'F', 0x24, 0xf0, 0xff, 0x7f, 'W', 'A', 'V', 'E', 'f', 'm', 't', ' ',
  194. 0x10, 0, 0, 0, 1, 0, 1, 0, 9, 0x3d, 0, 0, 0x12, 0x7a, 0, 0,
  195. 2, 0, 0x10, 0, 'd', 'a', 't', 'a', 0x00, 0xf0, 0xff, 0x7f
  196. };
  197. if (path == NULL)
  198. return 2;
  199. while (isspace(*path)) path++;
  200. f_wavfile = NULL;
  201. if (path[0] != 0) {
  202. if (strcmp(path, "stdout") == 0) {
  203. #ifdef _WIN32
  204. // prevent Windows adding 0x0d before 0x0a bytes
  205. _setmode(_fileno(stdout), _O_BINARY);
  206. #endif
  207. f_wavfile = stdout;
  208. } else
  209. f_wavfile = fopen(path, "wb");
  210. }
  211. if (f_wavfile == NULL) {
  212. fprintf(stderr, "Can't write to: '%s'\n", path);
  213. return 1;
  214. }
  215. fwrite(wave_hdr, 1, 24, f_wavfile);
  216. Write4Bytes(f_wavfile, rate);
  217. Write4Bytes(f_wavfile, rate * 2);
  218. fwrite(&wave_hdr[32], 1, 12, f_wavfile);
  219. return 0;
  220. }
  221. static void CloseWavFile()
  222. {
  223. unsigned int pos;
  224. if ((f_wavfile == NULL) || (f_wavfile == stdout))
  225. return;
  226. fflush(f_wavfile);
  227. pos = ftell(f_wavfile);
  228. if (fseek(f_wavfile, 4, SEEK_SET) != -1)
  229. Write4Bytes(f_wavfile, pos - 8);
  230. if (fseek(f_wavfile, 40, SEEK_SET) != -1)
  231. Write4Bytes(f_wavfile, pos - 44);
  232. fclose(f_wavfile);
  233. f_wavfile = NULL;
  234. }
  235. static int SynthCallback(short *wav, int numsamples, espeak_EVENT *events)
  236. {
  237. char fname[210];
  238. if (quiet || wav == NULL) return 0;
  239. while (events->type != 0) {
  240. if (events->type == espeakEVENT_SAMPLERATE) {
  241. samplerate = events->id.number;
  242. samples_split = samples_split_seconds * samplerate;
  243. } else if (events->type == espeakEVENT_SENTENCE) {
  244. // start a new WAV file when the limit is reached, at this sentence boundary
  245. if ((samples_split > 0) && (samples_total > samples_split)) {
  246. CloseWavFile();
  247. samples_total = 0;
  248. wavefile_count++;
  249. }
  250. }
  251. events++;
  252. }
  253. if (f_wavfile == NULL) {
  254. if (samples_split > 0) {
  255. sprintf(fname, "%s_%.2d%s", wavefile, wavefile_count+1, filetype);
  256. if (OpenWavFile(fname, samplerate) != 0)
  257. return 1;
  258. } else if (OpenWavFile(wavefile, samplerate) != 0)
  259. return 1;
  260. }
  261. if (numsamples > 0) {
  262. samples_total += numsamples;
  263. fwrite(wav, numsamples*2, 1, f_wavfile);
  264. }
  265. return 0;
  266. }
  267. static void PrintVersion()
  268. {
  269. const char *version;
  270. const char *path_data;
  271. espeak_Initialize(AUDIO_OUTPUT_SYNCHRONOUS, 0, NULL, espeakINITIALIZE_DONT_EXIT);
  272. version = espeak_Info(&path_data);
  273. printf("eSpeak NG text-to-speech: %s Data at: %s\n", version, path_data);
  274. }
  275. int main(int argc, char **argv)
  276. {
  277. static const struct option long_options[] = {
  278. { "help", no_argument, 0, 'h' },
  279. { "stdin", no_argument, 0, 0x100 },
  280. { "compile-debug", optional_argument, 0, 0x101 },
  281. { "compile", optional_argument, 0, 0x102 },
  282. { "punct", optional_argument, 0, 0x103 },
  283. { "voices", optional_argument, 0, 0x104 },
  284. { "stdout", no_argument, 0, 0x105 },
  285. { "split", optional_argument, 0, 0x106 },
  286. { "path", required_argument, 0, 0x107 },
  287. { "phonout", required_argument, 0, 0x108 },
  288. { "pho", no_argument, 0, 0x109 },
  289. { "ipa", optional_argument, 0, 0x10a },
  290. { "version", no_argument, 0, 0x10b },
  291. { "sep", optional_argument, 0, 0x10c },
  292. { "tie", optional_argument, 0, 0x10d },
  293. #if USE_MBROLA
  294. { "compile-mbrola", optional_argument, 0, 0x10e },
  295. #endif
  296. { "compile-intonations", no_argument, 0, 0x10f },
  297. { "compile-phonemes", optional_argument, 0, 0x110 },
  298. { "load", no_argument, 0, 0x111 },
  299. { "ssml-break", required_argument, 0, 0x112 },
  300. { 0, 0, 0, 0 }
  301. };
  302. FILE *f_text = NULL;
  303. char *p_text = NULL;
  304. FILE *f_phonemes_out = stdout;
  305. char *data_path = NULL; // use default path for espeak-ng-data
  306. int option_index = 0;
  307. int c;
  308. int ix;
  309. char *optarg2;
  310. int value;
  311. int flag_stdin = 0;
  312. int flag_compile = 0;
  313. int flag_load = 0;
  314. int filesize = 0;
  315. int synth_flags = espeakCHARS_AUTO | espeakPHONEMES | espeakENDPAUSE;
  316. int volume = -1;
  317. int speed = -1;
  318. int pitch = -1;
  319. int pitch_range = -1;
  320. int wordgap = -1;
  321. int option_capitals = -1;
  322. int option_punctuation = -1;
  323. int phonemes_separator = 0;
  324. int phoneme_options = 0;
  325. int option_linelength = 0;
  326. int option_waveout = 0;
  327. int ssml_break = -1;
  328. bool deterministic = 0;
  329. espeak_VOICE voice_select;
  330. char filename[200];
  331. char voicename[40];
  332. char devicename[200];
  333. #define N_PUNCTLIST 100
  334. wchar_t option_punctlist[N_PUNCTLIST];
  335. voicename[0] = 0;
  336. wavefile[0] = 0;
  337. filename[0] = 0;
  338. devicename[0] = 0;
  339. option_punctlist[0] = 0;
  340. while (true) {
  341. c = getopt_long(argc, argv, "a:b:Dd:f:g:hk:l:mp:P:qs:v:w:xXz",
  342. long_options, &option_index);
  343. // Detect the end of the options.
  344. if (c == -1)
  345. break;
  346. optarg2 = optarg;
  347. switch (c)
  348. {
  349. case 'b':
  350. // input character encoding, 8bit, 16bit, UTF8
  351. if ((sscanf(optarg2, "%d", &value) == 1) && (value <= 4))
  352. synth_flags |= value;
  353. else
  354. synth_flags |= espeakCHARS_8BIT;
  355. break;
  356. case 'd':
  357. strncpy0(devicename, optarg2, sizeof(devicename));
  358. break;
  359. case 'D':
  360. deterministic = 1;
  361. break;
  362. case 'h':
  363. printf("\n");
  364. PrintVersion();
  365. printf("%s", help_text);
  366. return 0;
  367. case 'k':
  368. option_capitals = atoi(optarg2);
  369. break;
  370. case 'x':
  371. phoneme_options |= espeakPHONEMES_SHOW;
  372. break;
  373. case 'X':
  374. phoneme_options |= espeakPHONEMES_TRACE;
  375. break;
  376. case 'm':
  377. synth_flags |= espeakSSML;
  378. break;
  379. case 'p':
  380. pitch = atoi(optarg2);
  381. break;
  382. case 'P':
  383. pitch_range = atoi(optarg2);
  384. break;
  385. case 'q':
  386. quiet = true;
  387. break;
  388. case 'f':
  389. strncpy0(filename, optarg2, sizeof(filename));
  390. break;
  391. case 'l':
  392. option_linelength = atoi(optarg2);
  393. break;
  394. case 'a':
  395. volume = atoi(optarg2);
  396. break;
  397. case 's':
  398. speed = atoi(optarg2);
  399. break;
  400. case 'g':
  401. wordgap = atoi(optarg2);
  402. break;
  403. case 'v':
  404. strncpy0(voicename, optarg2, sizeof(voicename));
  405. break;
  406. case 'w':
  407. option_waveout = 1;
  408. strncpy0(wavefile, optarg2, sizeof(filename));
  409. break;
  410. case 'z': // remove pause from the end of a sentence
  411. synth_flags &= ~espeakENDPAUSE;
  412. break;
  413. case 0x100: // --stdin
  414. flag_stdin = 1;
  415. break;
  416. case 0x105: // --stdout
  417. option_waveout = 1;
  418. strcpy(wavefile, "stdout");
  419. break;
  420. case 0x101: // --compile-debug
  421. case 0x102: // --compile
  422. if (optarg2 != NULL && *optarg2) {
  423. strncpy0(voicename, optarg2, sizeof(voicename));
  424. flag_compile = c;
  425. quiet = true;
  426. break;
  427. } else {
  428. fprintf(stderr, "Voice name to '%s' not specified.\n", c == 0x101 ? "--compile-debug" : "--compile");
  429. exit(EXIT_FAILURE);
  430. }
  431. case 0x103: // --punct
  432. option_punctuation = 1;
  433. if (optarg2 != NULL) {
  434. ix = 0;
  435. while ((ix < N_PUNCTLIST) && ((option_punctlist[ix] = optarg2[ix]) != 0)) ix++;
  436. option_punctlist[N_PUNCTLIST-1] = 0;
  437. option_punctuation = 2;
  438. }
  439. break;
  440. case 0x104: // --voices
  441. espeak_Initialize(AUDIO_OUTPUT_SYNCHRONOUS, 0, data_path, 0);
  442. DisplayVoices(stdout, optarg2);
  443. exit(0);
  444. case 0x106: // -- split
  445. if (optarg2 == NULL)
  446. samples_split_seconds = 30 * 60; // default 30 minutes
  447. else
  448. samples_split_seconds = atoi(optarg2) * 60;
  449. break;
  450. case 0x107: // --path
  451. data_path = optarg2;
  452. break;
  453. case 0x108: // --phonout
  454. if ((f_phonemes_out = fopen(optarg2, "w")) == NULL)
  455. fprintf(stderr, "Can't write to: %s\n", optarg2);
  456. break;
  457. case 0x109: // --pho
  458. phoneme_options |= espeakPHONEMES_MBROLA;
  459. break;
  460. case 0x10a: // --ipa
  461. phoneme_options |= espeakPHONEMES_IPA;
  462. if (optarg2 != NULL) {
  463. // deprecated and obsolete
  464. switch (atoi(optarg2))
  465. {
  466. case 1:
  467. phonemes_separator = '_';
  468. break;
  469. case 2:
  470. phonemes_separator = 0x0361;
  471. phoneme_options |= espeakPHONEMES_TIE;
  472. break;
  473. case 3:
  474. phonemes_separator = 0x200d; // ZWJ
  475. phoneme_options |= espeakPHONEMES_TIE;
  476. break;
  477. }
  478. }
  479. break;
  480. case 0x10b: // --version
  481. PrintVersion();
  482. exit(0);
  483. case 0x10c: // --sep
  484. phoneme_options |= espeakPHONEMES_SHOW;
  485. if (optarg2 == 0)
  486. phonemes_separator = ' ';
  487. else
  488. utf8_in(&phonemes_separator, optarg2);
  489. if (phonemes_separator == 'z')
  490. phonemes_separator = 0x200c; // ZWNJ
  491. break;
  492. case 0x10d: // --tie
  493. phoneme_options |= (espeakPHONEMES_SHOW | espeakPHONEMES_TIE);
  494. if (optarg2 == 0)
  495. phonemes_separator = 0x0361; // default: combining-double-inverted-breve
  496. else
  497. utf8_in(&phonemes_separator, optarg2);
  498. if (phonemes_separator == 'z')
  499. phonemes_separator = 0x200d; // ZWJ
  500. break;
  501. #if USE_MBROLA
  502. case 0x10e: // --compile-mbrola
  503. {
  504. espeak_ng_InitializePath(data_path);
  505. espeak_ng_ERROR_CONTEXT context = NULL;
  506. espeak_ng_STATUS result = espeak_ng_CompileMbrolaVoice(optarg2, stdout, &context);
  507. if (result != ENS_OK) {
  508. espeak_ng_PrintStatusCodeMessage(result, stderr, context);
  509. espeak_ng_ClearErrorContext(&context);
  510. return EXIT_FAILURE;
  511. }
  512. return EXIT_SUCCESS;
  513. }
  514. #endif
  515. case 0x10f: // --compile-intonations
  516. {
  517. espeak_ng_InitializePath(data_path);
  518. espeak_ng_ERROR_CONTEXT context = NULL;
  519. espeak_ng_STATUS result = espeak_ng_CompileIntonation(stdout, &context);
  520. if (result != ENS_OK) {
  521. espeak_ng_PrintStatusCodeMessage(result, stderr, context);
  522. espeak_ng_ClearErrorContext(&context);
  523. return EXIT_FAILURE;
  524. }
  525. return EXIT_SUCCESS;
  526. }
  527. case 0x110: // --compile-phonemes
  528. {
  529. espeak_ng_InitializePath(data_path);
  530. espeak_ng_ERROR_CONTEXT context = NULL;
  531. espeak_ng_STATUS result;
  532. if (optarg2) {
  533. result = espeak_ng_CompilePhonemeDataPath(22050, optarg2, NULL, stdout, &context);
  534. } else {
  535. result = espeak_ng_CompilePhonemeData(22050, stdout, &context);
  536. }
  537. if (result != ENS_OK) {
  538. espeak_ng_PrintStatusCodeMessage(result, stderr, context);
  539. espeak_ng_ClearErrorContext(&context);
  540. return EXIT_FAILURE;
  541. }
  542. return EXIT_SUCCESS;
  543. }
  544. case 0x111: // --load
  545. flag_load = 1;
  546. break;
  547. case 0x112: // --ssml-break
  548. ssml_break = atoi(optarg2);
  549. break;
  550. default:
  551. exit(0);
  552. }
  553. }
  554. espeak_ng_InitializePath(data_path);
  555. espeak_ng_ERROR_CONTEXT context = NULL;
  556. espeak_ng_STATUS result = espeak_ng_Initialize(&context);
  557. if (result != ENS_OK) {
  558. espeak_ng_PrintStatusCodeMessage(result, stderr, context);
  559. espeak_ng_ClearErrorContext(&context);
  560. exit(1);
  561. }
  562. if (deterministic) {
  563. // Set random generator state to well-known
  564. espeak_ng_SetRandSeed(1);
  565. }
  566. if (option_waveout || quiet) {
  567. // writing to a file (or no output), we can use synchronous mode
  568. result = espeak_ng_InitializeOutput(ENOUTPUT_MODE_SYNCHRONOUS, 0, devicename[0] ? devicename : NULL);
  569. samplerate = espeak_ng_GetSampleRate();
  570. samples_split = samplerate * samples_split_seconds;
  571. espeak_SetSynthCallback(SynthCallback);
  572. if (samples_split) {
  573. char *extn;
  574. extn = strrchr(wavefile, '.');
  575. if ((extn != NULL) && ((wavefile + strlen(wavefile) - extn) <= 4)) {
  576. strcpy(filetype, extn);
  577. *extn = 0;
  578. }
  579. }
  580. } else {
  581. // play the sound output
  582. result = espeak_ng_InitializeOutput(PLAYBACK_MODE, 0, devicename[0] ? devicename : NULL);
  583. samplerate = espeak_ng_GetSampleRate();
  584. }
  585. if (result != ENS_OK) {
  586. espeak_ng_PrintStatusCodeMessage(result, stderr, NULL);
  587. exit(EXIT_FAILURE);
  588. }
  589. if (voicename[0] == 0)
  590. strcpy(voicename, ESPEAKNG_DEFAULT_VOICE);
  591. if(flag_load)
  592. result = espeak_ng_SetVoiceByFile(voicename);
  593. else
  594. result = espeak_ng_SetVoiceByName(voicename);
  595. if (result != ENS_OK) {
  596. memset(&voice_select, 0, sizeof(voice_select));
  597. voice_select.languages = voicename;
  598. result = espeak_ng_SetVoiceByProperties(&voice_select);
  599. if (result != ENS_OK) {
  600. espeak_ng_PrintStatusCodeMessage(result, stderr, NULL);
  601. exit(EXIT_FAILURE);
  602. }
  603. }
  604. if (flag_compile) {
  605. // This must be done after the voice is set
  606. espeak_ng_ERROR_CONTEXT context = NULL;
  607. espeak_ng_STATUS result = espeak_ng_CompileDictionary("", NULL, stderr, flag_compile & 0x1, &context);
  608. if (result != ENS_OK) {
  609. espeak_ng_PrintStatusCodeMessage(result, stderr, context);
  610. espeak_ng_ClearErrorContext(&context);
  611. return EXIT_FAILURE;
  612. }
  613. return EXIT_SUCCESS;
  614. }
  615. // set any non-default values of parameters. This must be done after espeak_Initialize()
  616. if (speed > 0)
  617. espeak_SetParameter(espeakRATE, speed, 0);
  618. if (volume >= 0)
  619. espeak_SetParameter(espeakVOLUME, volume, 0);
  620. if (pitch >= 0)
  621. espeak_SetParameter(espeakPITCH, pitch, 0);
  622. if (pitch_range >= 0)
  623. espeak_SetParameter(espeakRANGE, pitch_range, 0);
  624. if (option_capitals >= 0)
  625. espeak_SetParameter(espeakCAPITALS, option_capitals, 0);
  626. if (option_punctuation >= 0)
  627. espeak_SetParameter(espeakPUNCTUATION, option_punctuation, 0);
  628. if (wordgap >= 0)
  629. espeak_SetParameter(espeakWORDGAP, wordgap, 0);
  630. if (option_linelength > 0)
  631. espeak_SetParameter(espeakLINELENGTH, option_linelength, 0);
  632. if (ssml_break > 0)
  633. espeak_SetParameter(espeakSSML_BREAK_MUL, ssml_break, 0);
  634. if (option_punctuation == 2)
  635. espeak_SetPunctuationList(option_punctlist);
  636. espeak_SetPhonemeTrace(phoneme_options | (phonemes_separator << 8), f_phonemes_out);
  637. if (filename[0] == 0) {
  638. if ((optind < argc) && (flag_stdin == 0)) {
  639. // there's a non-option parameter, and no -f or --stdin
  640. // use it as text
  641. p_text = argv[optind];
  642. } else {
  643. f_text = stdin;
  644. if (flag_stdin == 0)
  645. flag_stdin = 2;
  646. }
  647. } else {
  648. struct stat st;
  649. if (stat(filename, &st) != 0) {
  650. fprintf(stderr, "Failed to stat() file '%s'\n", filename);
  651. exit(EXIT_FAILURE);
  652. }
  653. filesize = GetFileLength(filename);
  654. f_text = fopen(filename, "r");
  655. if (f_text == NULL) {
  656. fprintf(stderr, "Failed to read file '%s'\n", filename);
  657. exit(EXIT_FAILURE);
  658. }
  659. if (S_ISFIFO(st.st_mode)) {
  660. flag_stdin = 2;
  661. }
  662. }
  663. if (p_text != NULL) {
  664. int size;
  665. size = strlen(p_text);
  666. espeak_Synth(p_text, size+1, 0, POS_CHARACTER, 0, synth_flags, NULL, NULL);
  667. } else if (flag_stdin) {
  668. size_t max = 1000;
  669. if ((p_text = (char *)malloc(max)) == NULL) {
  670. espeak_ng_PrintStatusCodeMessage(ENOMEM, stderr, NULL);
  671. exit(EXIT_FAILURE);
  672. }
  673. if (flag_stdin == 2) {
  674. // line by line input on stdin or from FIFO
  675. while (fgets(p_text, max, f_text) != NULL) {
  676. p_text[max-1] = 0;
  677. espeak_Synth(p_text, max, 0, POS_CHARACTER, 0, synth_flags, NULL, NULL);
  678. // Allow subprocesses to use the audio data through pipes.
  679. fflush(stdout);
  680. }
  681. if (f_text != stdin) {
  682. fclose(f_text);
  683. }
  684. } else {
  685. // bulk input on stdin
  686. ix = 0;
  687. while (true) {
  688. if ((c = fgetc(stdin)) == EOF)
  689. break;
  690. p_text[ix++] = (char)c;
  691. if (ix >= (max-1)) {
  692. char *new_text = NULL;
  693. if (max <= SIZE_MAX - 1000) {
  694. max += 1000;
  695. new_text = (char *)realloc(p_text, max);
  696. }
  697. if (new_text == NULL) {
  698. free(p_text);
  699. espeak_ng_PrintStatusCodeMessage(ENOMEM, stderr, NULL);
  700. exit(EXIT_FAILURE);
  701. }
  702. p_text = new_text;
  703. }
  704. }
  705. if (ix > 0) {
  706. p_text[ix] = 0;
  707. espeak_Synth(p_text, ix, 0, POS_CHARACTER, 0, synth_flags, NULL, NULL);
  708. }
  709. }
  710. free(p_text);
  711. } else if (f_text != NULL) {
  712. if ((p_text = (char *)malloc(filesize+1)) == NULL) {
  713. espeak_ng_PrintStatusCodeMessage(ENOMEM, stderr, NULL);
  714. exit(EXIT_FAILURE);
  715. }
  716. fread(p_text, 1, filesize, f_text);
  717. p_text[filesize] = 0;
  718. espeak_Synth(p_text, filesize+1, 0, POS_CHARACTER, 0, synth_flags, NULL, NULL);
  719. fclose(f_text);
  720. free(p_text);
  721. }
  722. result = espeak_ng_Synchronize();
  723. if (result != ENS_OK) {
  724. espeak_ng_PrintStatusCodeMessage(result, stderr, NULL);
  725. exit(EXIT_FAILURE);
  726. }
  727. if (f_phonemes_out != stdout)
  728. fclose(f_phonemes_out);
  729. CloseWavFile();
  730. espeak_ng_Terminate();
  731. return 0;
  732. }