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

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774
  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. "-s <integer>\n"
  64. "\t Speed in approximate words per minute. The default is 175\n"
  65. "-v <voice name>\n"
  66. "\t Use voice file of this name from espeak-ng-data/voices\n"
  67. "-w <wave file name>\n"
  68. "\t Write speech to this WAV file, rather than speaking it directly\n"
  69. "-b\t Input text encoding, 1=UTF8, 2=8 bit, 4=16 bit \n"
  70. "-m\t Interpret SSML markup, and ignore other < > tags\n"
  71. "-q\t Quiet, don't produce any speech (may be useful with -x)\n"
  72. "-x\t Write phoneme mnemonics to stdout\n"
  73. "-X\t Write phonemes mnemonics and translation trace to stdout\n"
  74. "-z\t No final sentence pause at the end of the text\n"
  75. "-D\t Enable deterministic random mode\n"
  76. "--compile=<voice name>\n"
  77. "\t Compile pronunciation rules and dictionary from the current\n"
  78. "\t directory. <voice name> specifies the language\n"
  79. "--compile-debug=<voice name>\n"
  80. "\t Compile pronunciation rules and dictionary from the current\n"
  81. "\t directory, including line numbers for use with -X.\n"
  82. "\t <voice name> specifies the language\n"
  83. "--compile-mbrola=<voice name>\n"
  84. "\t Compile an MBROLA voice\n"
  85. "--compile-intonations\n"
  86. "\t Compile the intonation data\n"
  87. "--compile-phonemes=<phsource-dir>\n"
  88. "\t Compile the phoneme data using <phsource-dir> or the default phsource directory\n"
  89. "--ipa Write phonemes to stdout using International Phonetic Alphabet\n"
  90. "--path=\"<path>\"\n"
  91. "\t Specifies the directory containing the espeak-ng-data directory\n"
  92. "--pho Write mbrola phoneme data (.pho) to stdout or to the file in --phonout\n"
  93. "--phonout=\"<filename>\"\n"
  94. "\t Write phoneme output from -x -X --ipa and --pho to this file\n"
  95. "--punct=\"<characters>\"\n"
  96. "\t Speak the names of punctuation characters during speaking. If\n"
  97. "\t =<characters> is omitted, all punctuation is spoken.\n"
  98. "--sep=<character>\n"
  99. "\t Separate phonemes (from -x --ipa) with <character>.\n"
  100. "\t Default is space, z means ZWJN character.\n"
  101. "--split=<minutes>\n"
  102. "\t Starts a new WAV file every <minutes>. Used with -w\n"
  103. "--stdout Write speech output to stdout\n"
  104. "--tie=<character>\n"
  105. "\t Use a tie character within multi-letter phoneme names.\n"
  106. "\t Default is U+361, z means ZWJ character.\n"
  107. "--version Shows version number and date, and location of espeak-ng-data\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. "--load Load voice from a file in current directory by name.\n"
  112. "-h, --help Show this help.\n";
  113. static int samplerate;
  114. bool quiet = false;
  115. unsigned int samples_total = 0;
  116. unsigned int samples_split = 0;
  117. unsigned int samples_split_seconds = 0;
  118. unsigned int wavefile_count = 0;
  119. FILE *f_wavfile = NULL;
  120. char filetype[5];
  121. char wavefile[200];
  122. static void DisplayVoices(FILE *f_out, char *language)
  123. {
  124. int ix;
  125. const char *p;
  126. int len;
  127. int count;
  128. int c;
  129. size_t j;
  130. const espeak_VOICE *v;
  131. const char *lang_name;
  132. char age_buf[12];
  133. char buf[80];
  134. const espeak_VOICE **voices;
  135. espeak_VOICE voice_select;
  136. static const char genders[4] = { '-', 'M', 'F', '-' };
  137. if ((language != NULL) && (language[0] != 0)) {
  138. // display only voices for the specified language, in order of priority
  139. voice_select.languages = language;
  140. voice_select.age = 0;
  141. voice_select.gender = 0;
  142. voice_select.name = NULL;
  143. voices = espeak_ListVoices(&voice_select);
  144. } else
  145. voices = espeak_ListVoices(NULL);
  146. fprintf(f_out, "Pty Language Age/Gender VoiceName File Other Languages\n");
  147. for (ix = 0; (v = voices[ix]) != NULL; ix++) {
  148. count = 0;
  149. p = v->languages;
  150. while (*p != 0) {
  151. len = strlen(p+1);
  152. lang_name = p+1;
  153. if (v->age == 0)
  154. strcpy(age_buf, " --");
  155. else
  156. sprintf(age_buf, "%3d", v->age);
  157. if (count == 0) {
  158. for (j = 0; j < sizeof(buf); j++) {
  159. // replace spaces in the name
  160. if ((c = v->name[j]) == ' ')
  161. c = '_';
  162. if ((buf[j] = c) == 0)
  163. break;
  164. }
  165. fprintf(f_out, "%2d %-15s%s/%c %-18s %-20s ",
  166. p[0], lang_name, age_buf, genders[v->gender], buf, v->identifier);
  167. } else
  168. fprintf(f_out, "(%s %d)", lang_name, p[0]);
  169. count++;
  170. p += len+2;
  171. }
  172. fputc('\n', f_out);
  173. }
  174. }
  175. static void Write4Bytes(FILE *f, int value)
  176. {
  177. // Write 4 bytes to a file, least significant first
  178. int ix;
  179. for (ix = 0; ix < 4; ix++) {
  180. fputc(value & 0xff, f);
  181. value = value >> 8;
  182. }
  183. }
  184. static int OpenWavFile(char *path, int rate)
  185. {
  186. static const unsigned char wave_hdr[44] = {
  187. 'R', 'I', 'F', 'F', 0x24, 0xf0, 0xff, 0x7f, 'W', 'A', 'V', 'E', 'f', 'm', 't', ' ',
  188. 0x10, 0, 0, 0, 1, 0, 1, 0, 9, 0x3d, 0, 0, 0x12, 0x7a, 0, 0,
  189. 2, 0, 0x10, 0, 'd', 'a', 't', 'a', 0x00, 0xf0, 0xff, 0x7f
  190. };
  191. if (path == NULL)
  192. return 2;
  193. while (isspace(*path)) path++;
  194. f_wavfile = NULL;
  195. if (path[0] != 0) {
  196. if (strcmp(path, "stdout") == 0) {
  197. #ifdef PLATFORM_WINDOWS
  198. // prevent Windows adding 0x0d before 0x0a bytes
  199. _setmode(_fileno(stdout), _O_BINARY);
  200. #endif
  201. f_wavfile = stdout;
  202. } else
  203. f_wavfile = fopen(path, "wb");
  204. }
  205. if (f_wavfile == NULL) {
  206. fprintf(stderr, "Can't write to: '%s'\n", path);
  207. return 1;
  208. }
  209. fwrite(wave_hdr, 1, 24, f_wavfile);
  210. Write4Bytes(f_wavfile, rate);
  211. Write4Bytes(f_wavfile, rate * 2);
  212. fwrite(&wave_hdr[32], 1, 12, f_wavfile);
  213. return 0;
  214. }
  215. static void CloseWavFile()
  216. {
  217. unsigned int pos;
  218. if ((f_wavfile == NULL) || (f_wavfile == stdout))
  219. return;
  220. fflush(f_wavfile);
  221. pos = ftell(f_wavfile);
  222. if (fseek(f_wavfile, 4, SEEK_SET) != -1)
  223. Write4Bytes(f_wavfile, pos - 8);
  224. if (fseek(f_wavfile, 40, SEEK_SET) != -1)
  225. Write4Bytes(f_wavfile, pos - 44);
  226. fclose(f_wavfile);
  227. f_wavfile = NULL;
  228. }
  229. static int SynthCallback(short *wav, int numsamples, espeak_EVENT *events)
  230. {
  231. char fname[210];
  232. if (quiet || wav == NULL) return 0;
  233. while (events->type != 0) {
  234. if (events->type == espeakEVENT_SAMPLERATE) {
  235. samplerate = events->id.number;
  236. samples_split = samples_split_seconds * samplerate;
  237. } else if (events->type == espeakEVENT_SENTENCE) {
  238. // start a new WAV file when the limit is reached, at this sentence boundary
  239. if ((samples_split > 0) && (samples_total > samples_split)) {
  240. CloseWavFile();
  241. samples_total = 0;
  242. wavefile_count++;
  243. }
  244. }
  245. events++;
  246. }
  247. if (f_wavfile == NULL) {
  248. if (samples_split > 0) {
  249. sprintf(fname, "%s_%.2d%s", wavefile, wavefile_count+1, filetype);
  250. if (OpenWavFile(fname, samplerate) != 0)
  251. return 1;
  252. } else if (OpenWavFile(wavefile, samplerate) != 0)
  253. return 1;
  254. }
  255. if (numsamples > 0) {
  256. samples_total += numsamples;
  257. fwrite(wav, numsamples*2, 1, f_wavfile);
  258. }
  259. return 0;
  260. }
  261. static void PrintVersion()
  262. {
  263. const char *version;
  264. const char *path_data;
  265. espeak_Initialize(AUDIO_OUTPUT_SYNCHRONOUS, 0, NULL, espeakINITIALIZE_DONT_EXIT);
  266. version = espeak_Info(&path_data);
  267. printf("eSpeak NG text-to-speech: %s Data at: %s\n", version, path_data);
  268. }
  269. int main(int argc, char **argv)
  270. {
  271. static const struct option long_options[] = {
  272. { "help", no_argument, 0, 'h' },
  273. { "stdin", no_argument, 0, 0x100 },
  274. { "compile-debug", optional_argument, 0, 0x101 },
  275. { "compile", optional_argument, 0, 0x102 },
  276. { "punct", optional_argument, 0, 0x103 },
  277. { "voices", optional_argument, 0, 0x104 },
  278. { "stdout", no_argument, 0, 0x105 },
  279. { "split", optional_argument, 0, 0x106 },
  280. { "path", required_argument, 0, 0x107 },
  281. { "phonout", required_argument, 0, 0x108 },
  282. { "pho", no_argument, 0, 0x109 },
  283. { "ipa", optional_argument, 0, 0x10a },
  284. { "version", no_argument, 0, 0x10b },
  285. { "sep", optional_argument, 0, 0x10c },
  286. { "tie", optional_argument, 0, 0x10d },
  287. { "compile-mbrola", optional_argument, 0, 0x10e },
  288. { "compile-intonations", no_argument, 0, 0x10f },
  289. { "compile-phonemes", optional_argument, 0, 0x110 },
  290. { "load", no_argument, 0, 0x111 },
  291. { 0, 0, 0, 0 }
  292. };
  293. FILE *f_text = NULL;
  294. char *p_text = NULL;
  295. FILE *f_phonemes_out = stdout;
  296. char *data_path = NULL; // use default path for espeak-ng-data
  297. int option_index = 0;
  298. int c;
  299. int ix;
  300. char *optarg2;
  301. int value;
  302. int flag_stdin = 0;
  303. int flag_compile = 0;
  304. int flag_load = 0;
  305. int filesize = 0;
  306. int synth_flags = espeakCHARS_AUTO | espeakPHONEMES | espeakENDPAUSE;
  307. int volume = -1;
  308. int speed = -1;
  309. int pitch = -1;
  310. int wordgap = -1;
  311. int option_capitals = -1;
  312. int option_punctuation = -1;
  313. int phonemes_separator = 0;
  314. int phoneme_options = 0;
  315. int option_linelength = 0;
  316. int option_waveout = 0;
  317. bool deterministic = 0;
  318. espeak_VOICE voice_select;
  319. char filename[200];
  320. char voicename[40];
  321. char devicename[200];
  322. #define N_PUNCTLIST 100
  323. wchar_t option_punctlist[N_PUNCTLIST];
  324. voicename[0] = 0;
  325. wavefile[0] = 0;
  326. filename[0] = 0;
  327. devicename[0] = 0;
  328. option_punctlist[0] = 0;
  329. while (true) {
  330. c = getopt_long(argc, argv, "a:b:Dd:f:g:hk:l:mp:qs:v:w:xXz",
  331. long_options, &option_index);
  332. // Detect the end of the options.
  333. if (c == -1)
  334. break;
  335. optarg2 = optarg;
  336. switch (c)
  337. {
  338. case 'b':
  339. // input character encoding, 8bit, 16bit, UTF8
  340. if ((sscanf(optarg2, "%d", &value) == 1) && (value <= 4))
  341. synth_flags |= value;
  342. else
  343. synth_flags |= espeakCHARS_8BIT;
  344. break;
  345. case 'd':
  346. strncpy0(devicename, optarg2, sizeof(devicename));
  347. break;
  348. case 'D':
  349. deterministic = 1;
  350. break;
  351. case 'h':
  352. printf("\n");
  353. PrintVersion();
  354. printf("%s", help_text);
  355. return 0;
  356. case 'k':
  357. option_capitals = atoi(optarg2);
  358. break;
  359. case 'x':
  360. phoneme_options |= espeakPHONEMES_SHOW;
  361. break;
  362. case 'X':
  363. phoneme_options |= espeakPHONEMES_TRACE;
  364. break;
  365. case 'm':
  366. synth_flags |= espeakSSML;
  367. break;
  368. case 'p':
  369. pitch = atoi(optarg2);
  370. break;
  371. case 'q':
  372. quiet = true;
  373. break;
  374. case 'f':
  375. strncpy0(filename, optarg2, sizeof(filename));
  376. break;
  377. case 'l':
  378. option_linelength = atoi(optarg2);
  379. break;
  380. case 'a':
  381. volume = atoi(optarg2);
  382. break;
  383. case 's':
  384. speed = atoi(optarg2);
  385. break;
  386. case 'g':
  387. wordgap = atoi(optarg2);
  388. break;
  389. case 'v':
  390. strncpy0(voicename, optarg2, sizeof(voicename));
  391. break;
  392. case 'w':
  393. option_waveout = 1;
  394. strncpy0(wavefile, optarg2, sizeof(filename));
  395. break;
  396. case 'z': // remove pause from the end of a sentence
  397. synth_flags &= ~espeakENDPAUSE;
  398. break;
  399. case 0x100: // --stdin
  400. flag_stdin = 1;
  401. break;
  402. case 0x105: // --stdout
  403. option_waveout = 1;
  404. strcpy(wavefile, "stdout");
  405. break;
  406. case 0x101: // --compile-debug
  407. case 0x102: // --compile
  408. if (optarg2 != NULL && *optarg2) {
  409. strncpy0(voicename, optarg2, sizeof(voicename));
  410. flag_compile = c;
  411. quiet = true;
  412. break;
  413. } else {
  414. fprintf(stderr, "Voice name to '%s' not specified.\n", c == 0x101 ? "--compile-debug" : "--compile");
  415. exit(EXIT_FAILURE);
  416. }
  417. case 0x103: // --punct
  418. option_punctuation = 1;
  419. if (optarg2 != NULL) {
  420. ix = 0;
  421. while ((ix < N_PUNCTLIST) && ((option_punctlist[ix] = optarg2[ix]) != 0)) ix++;
  422. option_punctlist[N_PUNCTLIST-1] = 0;
  423. option_punctuation = 2;
  424. }
  425. break;
  426. case 0x104: // --voices
  427. espeak_Initialize(AUDIO_OUTPUT_SYNCHRONOUS, 0, data_path, 0);
  428. DisplayVoices(stdout, optarg2);
  429. exit(0);
  430. case 0x106: // -- split
  431. if (optarg2 == NULL)
  432. samples_split_seconds = 30 * 60; // default 30 minutes
  433. else
  434. samples_split_seconds = atoi(optarg2) * 60;
  435. break;
  436. case 0x107: // --path
  437. data_path = optarg2;
  438. break;
  439. case 0x108: // --phonout
  440. if ((f_phonemes_out = fopen(optarg2, "w")) == NULL)
  441. fprintf(stderr, "Can't write to: %s\n", optarg2);
  442. break;
  443. case 0x109: // --pho
  444. phoneme_options |= espeakPHONEMES_MBROLA;
  445. break;
  446. case 0x10a: // --ipa
  447. phoneme_options |= espeakPHONEMES_IPA;
  448. if (optarg2 != NULL) {
  449. // deprecated and obsolete
  450. switch (atoi(optarg2))
  451. {
  452. case 1:
  453. phonemes_separator = '_';
  454. break;
  455. case 2:
  456. phonemes_separator = 0x0361;
  457. phoneme_options |= espeakPHONEMES_TIE;
  458. break;
  459. case 3:
  460. phonemes_separator = 0x200d; // ZWJ
  461. phoneme_options |= espeakPHONEMES_TIE;
  462. break;
  463. }
  464. }
  465. break;
  466. case 0x10b: // --version
  467. PrintVersion();
  468. exit(0);
  469. case 0x10c: // --sep
  470. phoneme_options |= espeakPHONEMES_SHOW;
  471. if (optarg2 == 0)
  472. phonemes_separator = ' ';
  473. else
  474. utf8_in(&phonemes_separator, optarg2);
  475. if (phonemes_separator == 'z')
  476. phonemes_separator = 0x200c; // ZWNJ
  477. break;
  478. case 0x10d: // --tie
  479. phoneme_options |= (espeakPHONEMES_SHOW | espeakPHONEMES_TIE);
  480. if (optarg2 == 0)
  481. phonemes_separator = 0x0361; // default: combining-double-inverted-breve
  482. else
  483. utf8_in(&phonemes_separator, optarg2);
  484. if (phonemes_separator == 'z')
  485. phonemes_separator = 0x200d; // ZWJ
  486. break;
  487. case 0x10e: // --compile-mbrola
  488. {
  489. espeak_ng_InitializePath(data_path);
  490. espeak_ng_ERROR_CONTEXT context = NULL;
  491. espeak_ng_STATUS result = espeak_ng_CompileMbrolaVoice(optarg2, stdout, &context);
  492. if (result != ENS_OK) {
  493. espeak_ng_PrintStatusCodeMessage(result, stderr, context);
  494. espeak_ng_ClearErrorContext(&context);
  495. return EXIT_FAILURE;
  496. }
  497. return EXIT_SUCCESS;
  498. }
  499. case 0x10f: // --compile-intonations
  500. {
  501. espeak_ng_InitializePath(data_path);
  502. espeak_ng_ERROR_CONTEXT context = NULL;
  503. espeak_ng_STATUS result = espeak_ng_CompileIntonation(stdout, &context);
  504. if (result != ENS_OK) {
  505. espeak_ng_PrintStatusCodeMessage(result, stderr, context);
  506. espeak_ng_ClearErrorContext(&context);
  507. return EXIT_FAILURE;
  508. }
  509. return EXIT_SUCCESS;
  510. }
  511. case 0x110: // --compile-phonemes
  512. {
  513. espeak_ng_InitializePath(data_path);
  514. espeak_ng_ERROR_CONTEXT context = NULL;
  515. espeak_ng_STATUS result;
  516. if (optarg2) {
  517. result = espeak_ng_CompilePhonemeDataPath(22050, optarg2, NULL, stdout, &context);
  518. } else {
  519. result = espeak_ng_CompilePhonemeData(22050, stdout, &context);
  520. }
  521. if (result != ENS_OK) {
  522. espeak_ng_PrintStatusCodeMessage(result, stderr, context);
  523. espeak_ng_ClearErrorContext(&context);
  524. return EXIT_FAILURE;
  525. }
  526. return EXIT_SUCCESS;
  527. }
  528. case 0x111: // --load
  529. flag_load = 1;
  530. break;
  531. default:
  532. exit(0);
  533. }
  534. }
  535. espeak_ng_InitializePath(data_path);
  536. espeak_ng_ERROR_CONTEXT context = NULL;
  537. espeak_ng_STATUS result = espeak_ng_Initialize(&context);
  538. if (result != ENS_OK) {
  539. espeak_ng_PrintStatusCodeMessage(result, stderr, context);
  540. espeak_ng_ClearErrorContext(&context);
  541. exit(1);
  542. }
  543. if (deterministic) {
  544. // Set random generator state to well-known
  545. espeak_ng_SetRandSeed(1);
  546. }
  547. if (option_waveout || quiet) {
  548. // writing to a file (or no output), we can use synchronous mode
  549. result = espeak_ng_InitializeOutput(ENOUTPUT_MODE_SYNCHRONOUS, 0, devicename[0] ? devicename : NULL);
  550. samplerate = espeak_ng_GetSampleRate();
  551. samples_split = samplerate * samples_split_seconds;
  552. espeak_SetSynthCallback(SynthCallback);
  553. if (samples_split) {
  554. char *extn;
  555. extn = strrchr(wavefile, '.');
  556. if ((extn != NULL) && ((wavefile + strlen(wavefile) - extn) <= 4)) {
  557. strcpy(filetype, extn);
  558. *extn = 0;
  559. }
  560. }
  561. } else {
  562. // play the sound output
  563. result = espeak_ng_InitializeOutput(PLAYBACK_MODE, 0, devicename[0] ? devicename : NULL);
  564. samplerate = espeak_ng_GetSampleRate();
  565. }
  566. if (result != ENS_OK) {
  567. espeak_ng_PrintStatusCodeMessage(result, stderr, NULL);
  568. exit(EXIT_FAILURE);
  569. }
  570. if (voicename[0] == 0)
  571. strcpy(voicename, ESPEAKNG_DEFAULT_VOICE);
  572. if(flag_load)
  573. result = espeak_ng_SetVoiceByFile(voicename);
  574. else
  575. result = espeak_ng_SetVoiceByName(voicename);
  576. if (result != ENS_OK) {
  577. memset(&voice_select, 0, sizeof(voice_select));
  578. voice_select.languages = voicename;
  579. result = espeak_ng_SetVoiceByProperties(&voice_select);
  580. if (result != ENS_OK) {
  581. espeak_ng_PrintStatusCodeMessage(result, stderr, NULL);
  582. exit(EXIT_FAILURE);
  583. }
  584. }
  585. if (flag_compile) {
  586. // This must be done after the voice is set
  587. espeak_ng_ERROR_CONTEXT context = NULL;
  588. espeak_ng_STATUS result = espeak_ng_CompileDictionary("", NULL, stderr, flag_compile & 0x1, &context);
  589. if (result != ENS_OK) {
  590. espeak_ng_PrintStatusCodeMessage(result, stderr, context);
  591. espeak_ng_ClearErrorContext(&context);
  592. return EXIT_FAILURE;
  593. }
  594. return EXIT_SUCCESS;
  595. }
  596. // set any non-default values of parameters. This must be done after espeak_Initialize()
  597. if (speed > 0)
  598. espeak_SetParameter(espeakRATE, speed, 0);
  599. if (volume >= 0)
  600. espeak_SetParameter(espeakVOLUME, volume, 0);
  601. if (pitch >= 0)
  602. espeak_SetParameter(espeakPITCH, pitch, 0);
  603. if (option_capitals >= 0)
  604. espeak_SetParameter(espeakCAPITALS, option_capitals, 0);
  605. if (option_punctuation >= 0)
  606. espeak_SetParameter(espeakPUNCTUATION, option_punctuation, 0);
  607. if (wordgap >= 0)
  608. espeak_SetParameter(espeakWORDGAP, wordgap, 0);
  609. if (option_linelength > 0)
  610. espeak_SetParameter(espeakLINELENGTH, option_linelength, 0);
  611. if (option_punctuation == 2)
  612. espeak_SetPunctuationList(option_punctlist);
  613. espeak_SetPhonemeTrace(phoneme_options | (phonemes_separator << 8), f_phonemes_out);
  614. if (filename[0] == 0) {
  615. if ((optind < argc) && (flag_stdin == 0)) {
  616. // there's a non-option parameter, and no -f or --stdin
  617. // use it as text
  618. p_text = argv[optind];
  619. } else {
  620. f_text = stdin;
  621. if (flag_stdin == 0)
  622. flag_stdin = 2;
  623. }
  624. } else {
  625. struct stat st;
  626. if (stat(filename, &st) != 0) {
  627. fprintf(stderr, "Failed to stat() file '%s'\n", filename);
  628. exit(EXIT_FAILURE);
  629. }
  630. filesize = GetFileLength(filename);
  631. f_text = fopen(filename, "r");
  632. if (f_text == NULL) {
  633. fprintf(stderr, "Failed to read file '%s'\n", filename);
  634. exit(EXIT_FAILURE);
  635. }
  636. if (S_ISFIFO(st.st_mode)) {
  637. flag_stdin = 2;
  638. }
  639. }
  640. if (p_text != NULL) {
  641. int size;
  642. size = strlen(p_text);
  643. espeak_Synth(p_text, size+1, 0, POS_CHARACTER, 0, synth_flags, NULL, NULL);
  644. } else if (flag_stdin) {
  645. size_t max = 1000;
  646. if ((p_text = (char *)malloc(max)) == NULL) {
  647. espeak_ng_PrintStatusCodeMessage(ENOMEM, stderr, NULL);
  648. exit(EXIT_FAILURE);
  649. }
  650. if (flag_stdin == 2) {
  651. // line by line input on stdin or from FIFO
  652. while (fgets(p_text, max, f_text) != NULL) {
  653. p_text[max-1] = 0;
  654. espeak_Synth(p_text, max, 0, POS_CHARACTER, 0, synth_flags, NULL, NULL);
  655. // Allow subprocesses to use the audio data through pipes.
  656. fflush(stdout);
  657. }
  658. if (f_text != stdin) {
  659. fclose(f_text);
  660. }
  661. } else {
  662. // bulk input on stdin
  663. ix = 0;
  664. while (true) {
  665. if ((c = fgetc(stdin)) == EOF)
  666. break;
  667. p_text[ix++] = (char)c;
  668. if (ix >= (max-1)) {
  669. char *new_text = NULL;
  670. if (max <= SIZE_MAX - 1000) {
  671. max += 1000;
  672. new_text = (char *)realloc(p_text, max);
  673. }
  674. if (new_text == NULL) {
  675. free(p_text);
  676. espeak_ng_PrintStatusCodeMessage(ENOMEM, stderr, NULL);
  677. exit(EXIT_FAILURE);
  678. }
  679. p_text = new_text;
  680. }
  681. }
  682. if (ix > 0) {
  683. p_text[ix-1] = 0;
  684. espeak_Synth(p_text, ix+1, 0, POS_CHARACTER, 0, synth_flags, NULL, NULL);
  685. }
  686. }
  687. free(p_text);
  688. } else if (f_text != NULL) {
  689. if ((p_text = (char *)malloc(filesize+1)) == NULL) {
  690. espeak_ng_PrintStatusCodeMessage(ENOMEM, stderr, NULL);
  691. exit(EXIT_FAILURE);
  692. }
  693. fread(p_text, 1, filesize, f_text);
  694. p_text[filesize] = 0;
  695. espeak_Synth(p_text, filesize+1, 0, POS_CHARACTER, 0, synth_flags, NULL, NULL);
  696. fclose(f_text);
  697. free(p_text);
  698. }
  699. result = espeak_ng_Synchronize();
  700. if (result != ENS_OK) {
  701. espeak_ng_PrintStatusCodeMessage(result, stderr, NULL);
  702. exit(EXIT_FAILURE);
  703. }
  704. if (f_phonemes_out != stdout)
  705. fclose(f_phonemes_out);
  706. CloseWavFile();
  707. espeak_ng_Terminate();
  708. return 0;
  709. }