diff options
Diffstat (limited to 'navit/support/espeak/tr_languages.c')
-rw-r--r-- | navit/support/espeak/tr_languages.c | 1121 |
1 files changed, 807 insertions, 314 deletions
diff --git a/navit/support/espeak/tr_languages.c b/navit/support/espeak/tr_languages.c index 83d1c8041..b1913f205 100644 --- a/navit/support/espeak/tr_languages.c +++ b/navit/support/espeak/tr_languages.c @@ -1,5 +1,5 @@ /*************************************************************************** - * Copyright (C) 2005 to 2007 by Jonathan Duddington * + * Copyright (C) 2005 to 2014 by Jonathan Duddington * * email: jonsd@users.sourceforge.net * * * * This program is free software; you can redistribute it and/or modify * @@ -39,22 +39,109 @@ #define L_qa 0x716100 #define L_grc 0x677263 // grc Ancient Greek #define L_jbo 0x6a626f // jbo Lojban -#define L_pap 0x706170 // pap Papiamento +#define L_mni 0x6d6e69 // mni Manipuri +#define L_pap 0x706170 // pap Papiamento] +#define L_qvi 0x717669 // qvi Kichwa +#define L_shs 0x736873 // shs Shuswap / Secwepemctsin #define L_zhy 0x7a6879 // zhy // start of unicode pages for character sets -#define OFFSET_GREEK 0x380 +#define OFFSET_GREEK 0x380 #define OFFSET_CYRILLIC 0x420 #define OFFSET_ARMENIAN 0x530 +#define OFFSET_HEBREW 0x590 +#define OFFSET_ARABIC 0x600 +#define OFFSET_THAANA 0x780 // Divehi/Maldives #define OFFSET_DEVANAGARI 0x900 -#define OFFSET_BENGALI 0x980 -#define OFFSET_TAMIL 0xb80 -#define OFFSET_KANNADA 0xc80 +#define OFFSET_BENGALI 0x980 +#define OFFSET_GURMUKHI 0xa00 +#define OFFSET_GUJARATI 0xa80 +#define OFFSET_ORIYA 0xb00 +#define OFFSET_TAMIL 0xb80 +#define OFFSET_TELUGU 0xc00 +#define OFFSET_KANNADA 0xc80 #define OFFSET_MALAYALAM 0xd00 -#define OFFSET_KOREAN 0x1100 +#define OFFSET_SINHALA 0x0d80 +#define OFFSET_THAI 0x0e00 +#define OFFSET_LAO 0x0e80 +#define OFFSET_TIBET 0x0f00 +#define OFFSET_MYANMAR 0x1000 +#define OFFSET_GEORGIAN 0x10a0 +#define OFFSET_KOREAN 0x1100 +#define OFFSET_ETHIOPIC 0x1200 + + +// character ranges must be listed in ascending order +ALPHABET alphabets [] = { + {"_el", OFFSET_GREEK, 0x380, 0x3ff, L('e','l'), AL_DONT_NAME | AL_NOT_LETTERS}, + {"_cyr", OFFSET_CYRILLIC, 0x400, 0x52f, 0, 0}, + {"_hy", OFFSET_ARMENIAN, 0x530, 0x58f, L('h','y'), AL_WORDS}, + {"_he", OFFSET_HEBREW, 0x590, 0x5ff, 0, 0}, + {"_ar", OFFSET_ARABIC, 0x600, 0x6ff, 0, 0}, + {"_dv", OFFSET_THAANA, 0x780, 0x7bf, 0, 0}, + {"_hi", OFFSET_DEVANAGARI, 0x900, 0x97f,L('h','i'), AL_WORDS}, + {"_bn", OFFSET_BENGALI, 0x0980, 0x9ff, L('b','n'), AL_WORDS}, + {"_gur", OFFSET_GURMUKHI, 0xa00, 0xa7f, L('p','a'), AL_WORDS}, + {"_gu", OFFSET_GUJARATI, 0xa80, 0xaff, L('g','u'), AL_WORDS}, + {"_or", OFFSET_ORIYA, 0xb00, 0xb7f, 0, 0}, + {"_ta", OFFSET_TAMIL, 0xb80, 0xbff, L('t','a'), AL_WORDS}, + {"_te", OFFSET_TELUGU, 0xc00, 0xc7f, L('t','e'), 0}, + {"_kn", OFFSET_KANNADA, 0xc80, 0xcff, L('k','n'), AL_WORDS}, + {"_ml", OFFSET_MALAYALAM,0xd00, 0xd7f, L('m','l'), AL_WORDS}, + {"_si", OFFSET_SINHALA, 0xd80, 0xdff, L('s','i'), AL_WORDS}, + {"_th", OFFSET_THAI, 0xe00, 0xe7f, 0, 0}, + {"_lo", OFFSET_LAO, 0xe80, 0xeff, 0, 0}, + {"_ti", OFFSET_TIBET, 0xf00, 0xfff, 0, 0}, + {"_my", OFFSET_MYANMAR, 0x1000,0x109f, 0, 0}, + {"_ka", OFFSET_GEORGIAN, 0x10a0,0x10ff, L('k','a'), AL_WORDS}, + {"_ko", OFFSET_KOREAN, 0x1100,0x11ff, L('k','o'), AL_WORDS}, + {"_eth", OFFSET_ETHIOPIC, 0x1200,0x139f, 0, 0}, + {"_braille", 0x2800, 0x2800,0x28ff, 0, AL_NO_SYMBOL}, + {"_ja", 0x3040, 0x3040,0x30ff, 0, AL_NOT_CODE}, + {"_zh", 0x3100, 0x3100,0x9fff, 0, AL_NOT_CODE}, + {"_ko", 0xa700, 0xa700,0xd7ff, L('k','o'), AL_NOT_CODE | AL_WORDS}, + {NULL, 0, 0, 0, 0, 0} +}; + + +ALPHABET *AlphabetFromName(const char *name) +{//========================================== + ALPHABET *alphabet; + + for(alphabet=alphabets; alphabet->name != NULL; alphabet++) + { + if(strcmp(name, &alphabet->name[1]) == 0) + return(alphabet); + } + return(NULL); +} + + +ALPHABET *AlphabetFromChar(int c) +{//=============================== + // Find the alphabet from a character. + ALPHABET *alphabet = alphabets; + + while(alphabet->name != NULL) + { + if(c <= alphabet->range_max) + { + if(c >= alphabet->range_min) + return(alphabet); + else + break; + } + alphabet++; + } + return(NULL); +} + + static void Translator_Russian(Translator *tr); + + static void SetLetterVowel(Translator *tr, int c) {//============================================== tr->letter_bits[c] = (tr->letter_bits[c] & 0x40) | 0x81; // keep value for group 6 (front vowels e,i,y) @@ -78,7 +165,7 @@ static void SetLetterBits(Translator *tr, int group, const char *string) {//===================================================================== int bits; unsigned char c; - + bits = (1L << group); while((c = *string++) != 0) tr->letter_bits[c] |= bits; @@ -96,15 +183,58 @@ static void SetLetterBitsRange(Translator *tr, int group, int first, int last) } } +// ignore these characters +static const unsigned short chars_ignore_default[] = { + 0xad, 1, // soft hyphen + 0x200c, 1, // zero width non-joiner + 0x200d, 1, // zero width joiner + 0, 0 }; + +// alternatively, ignore characters but allow zero-width-non-joiner (lang-fa) +static const unsigned short chars_ignore_zwnj_hyphen[] = { + 0xad, 1, // soft hyphen + 0x640, 1, // igniore Arabic Tatweel (lang=FA) + 0x200c, '-', // zero width non-joiner, replace with hyphen + 0x200d, 1, // zero width joiner + 0, 0 }; + +//const char string_ordinal[] = {0xc2,0xba,0}; // masculine ordinal character, UTF-8 +const char string_ordinal[] = "\xc2\xba"; // masculine ordinal character, UTF-8 + static Translator* NewTranslator(void) {//=================================== Translator *tr; int ix; - static const unsigned char stress_amps2[] = {17,17, 20,20, 20,22, 22,20 }; + static const unsigned char stress_amps2[] = {18,18, 20,20, 20,22, 22,20 }; static const short stress_lengths2[8] = {182,140, 220,220, 220,240, 260,280}; static const wchar_t empty_wstring[1] = {0}; static const wchar_t punct_in_word[2] = {'\'', 0}; // allow hyphen within words + static const unsigned char default_tunes[6] = {0, 1, 2, 3, 0, 0}; + +// Translates character codes in the range transpose_min to transpose_max to +// a number in the range 1 to 63. 0 indicates there is no translation. +// Used up to 57 (max of 63) +static const char transpose_map_latin[] = { + 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, // 0x60 +16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 0, 0, 0, 0, 0, // 0x70 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0x80 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0x90 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0xa0 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0xb0 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0xc0 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0xd0 +27, 28, 29, 0, 0, 30, 31, 32, 33, 34, 35, 36, 0, 37, 38, 0, // 0xe0 + 0, 0, 0, 39, 0, 0, 40, 0, 41, 0, 42, 0, 43, 0, 0, 0, // 0xf0 + 0, 0, 0, 44, 0, 45, 0, 46, 0, 0, 0, 0, 0, 47, 0, 0, // 0x100 + 0, 48, 0, 0, 0, 0, 0, 0, 0, 49, 0, 0, 0, 0, 0, 0, // 0x110 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0x120 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0x130 + 0, 0, 50, 0, 51, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0x140 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 52, 0, 0, 0, 0, // 0x150 + 0, 53, 0, 54, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0x160 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 55, 0, 56, 0, 57, 0, // 0x170 +}; tr = (Translator *)Alloc(sizeof(Translator)); if(tr == NULL) @@ -112,18 +242,23 @@ static Translator* NewTranslator(void) tr->charset_a0 = charsets[1]; // ISO-8859-1, this is for when the input is not utf8 dictionary_name[0] = 0; + tr->dictionary_name[0] = 0; tr->dict_condition=0; + tr->dict_min_size = 0; tr->data_dictrules = NULL; // language_1 translation rules file tr->data_dictlist = NULL; // language_2 dictionary lookup file - tr->transpose_offset = 0; + tr->transpose_min = 0x60; + tr->transpose_max = 0x17f; + tr->transpose_map = transpose_map_latin; + tr->frequent_pairs = NULL; // only need lower case tr->letter_bits_offset = 0; memset(tr->letter_bits,0,sizeof(tr->letter_bits)); memset(tr->letter_groups,0,sizeof(tr->letter_groups)); - // 0-5 sets of characters matched by A B C E F G in pronunciation rules + // 0-6 sets of characters matched by A B C H F G Y in pronunciation rules // these may be set differently for different languages SetLetterBits(tr,0,"aeiou"); // A vowels, except y SetLetterBits(tr,1,"bcdfgjklmnpqstvxz"); // B hard consonants, excluding h,r,w @@ -137,6 +272,7 @@ static Translator* NewTranslator(void) tr->char_plus_apostrophe = empty_wstring; tr->punct_within_word = punct_in_word; + tr->chars_ignore = chars_ignore_default; for(ix=0; ix<8; ix++) { @@ -145,16 +281,23 @@ static Translator* NewTranslator(void) tr->stress_lengths[ix] = stress_lengths2[ix]; } memset(&(tr->langopts),0,sizeof(tr->langopts)); + tr->langopts.max_lengthmod = 500; + tr->langopts.lengthen_tonic = 20; - tr->langopts.stress_rule = 2; + tr->langopts.stress_rule = STRESSPOSN_2R; tr->langopts.unstressed_wd1 = 1; tr->langopts.unstressed_wd2 = 3; tr->langopts.param[LOPT_SONORANT_MIN] = 95; + tr->langopts.param[LOPT_LONG_VOWEL_THRESHOLD] = 190/2; tr->langopts.param[LOPT_MAXAMP_EOC] = 19; tr->langopts.param[LOPT_UNPRONOUNCABLE] = 's'; // don't count this character at start of word + tr->langopts.param[LOPT_BRACKET_PAUSE] = 4; // pause at bracket + tr->langopts.param2[LOPT_BRACKET_PAUSE] = 2; // pauses when announcing bracket names tr->langopts.max_initial_consonants = 3; tr->langopts.replace_chars = NULL; - tr->langopts.ascii_language = ""; // Non-Latin alphabet languages, use this language to speak Latin words, default is English + tr->langopts.ascii_language[0] = 0; // Non-Latin alphabet languages, use this language to speak Latin words, default is English + tr->langopts.alt_alphabet_lang = L('e','n'); + tr->langopts.roman_suffix = ""; SetLengthMods(tr,201); // tr->langopts.length_mods = length_mods_en; @@ -163,16 +306,59 @@ static Translator* NewTranslator(void) tr->langopts.long_stop = 100; tr->langopts.max_roman = 49; + tr->langopts.min_roman = 2; tr->langopts.thousands_sep = ','; tr->langopts.decimal_sep = '.'; + tr->langopts.break_numbers = BREAK_THOUSANDS; // 1000, 1000,000 1,000,000 etc + tr->langopts.max_digits = 14; memcpy(tr->punct_to_tone, punctuation_to_tone, sizeof(tr->punct_to_tone)); + memcpy(tr->langopts.tunes, default_tunes, sizeof(tr->langopts.tunes)); + return(tr); } - -static const unsigned int replace_cyrillic_latin[] = +// common letter pairs, encode these as a single byte +// 2 bytes, using the transposed character codes +static const short pairs_ru[] = { +0x010c, // ла 21052 0x23 +0x010e, // на 18400 +0x0113, // та 14254 +0x0301, // ав 31083 +0x030f, // ов 13420 +0x060e, // не 21798 +0x0611, // ре 19458 +0x0903, // ви 16226 +0x0b01, // ак 14456 +0x0b0f, // ок 17836 +0x0c01, // ал 13324 +0x0c09, // ил 16877 +0x0e01, // ан 15359 +0x0e06, // ен 13543 0x30 +0x0e09, // ин 17168 +0x0e0e, // нн 15973 +0x0e0f, // он 22373 +0x0e1c, // ын 15052 +0x0f03, // во 24947 +0x0f11, // ро 13552 +0x0f12, // со 16368 +0x100f, // оп 19054 +0x1011, // рп 17067 +0x1101, // ар 23967 +0x1106, // ер 18795 +0x1109, // ир 13797 +0x110f, // ор 21737 +0x1213, // тс 25076 +0x1220, // яс 14310 +0x7fff}; +//0x040f ог 12976 +//0x1306 ет 12826 +//0x0f0d мо 12688 + + + +static const unsigned int replace_cyrillic_latin[] = {0x430,'a', 0x431,'b', 0x446,'c', @@ -209,25 +395,63 @@ static const unsigned int replace_cyrillic_latin[] = 0}; // ѓ ѕ ќ -void SetIndicLetters(Translator *tr) +static const unsigned char ru_vowels[] = {0x10,0x15,0x31,0x18,0x1e,0x23,0x2b,0x2d,0x2e,0x2f, 0xb9,0xc9,0x91,0x8f,0x36,0}; //also kazakh +static const unsigned char ru_consonants[] = {0x11,0x12,0x13,0x14,0x16,0x17,0x19,0x1a,0x1b,0x1c,0x1d,0x1f,0x20,0x21,0x22,0x24,0x25,0x26,0x27,0x28,0x29,0x2a,0x2c, 0x73,0x7b,0x83,0x9b,0}; + +static void SetCyrillicLetters(Translator *tr) +{//=========================================== + // character codes offset by 0x420 + static const char ru_soft[] = {0x2c,0x19,0x27,0x29,0}; // letter group B [k ts; s;] + static const char ru_hard[] = {0x2a,0x16,0x26,0x28,0}; // letter group H [S Z ts] + static const char ru_nothard[] = {0x11,0x12,0x13,0x14,0x17,0x19,0x1a,0x1b,0x1c,0x1d,0x1f,0x20,0x21,0x22,0x24,0x25,0x27,0x29,0x2c,0}; + static const char ru_voiced[] = {0x11,0x12,0x13,0x14,0x16,0x17,0}; // letter group G (voiced obstruents) + static const char ru_ivowels[] = {0x2c,0x2e,0x2f,0x31,0}; // letter group Y (iotated vowels & soft-sign) + tr->charset_a0 = charsets[18]; // KOI8-R + tr->transpose_min = 0x430; // convert cyrillic from unicode into range 0x01 to 0x22 + tr->transpose_max = 0x451; + tr->transpose_map = NULL; + tr->frequent_pairs = pairs_ru; + + tr->letter_bits_offset = OFFSET_CYRILLIC; + memset(tr->letter_bits,0,sizeof(tr->letter_bits)); + SetLetterBits(tr,LETTERGP_A,(char *)ru_vowels); + SetLetterBits(tr,LETTERGP_B,ru_soft); + SetLetterBits(tr,LETTERGP_C,(char *)ru_consonants); + SetLetterBits(tr,LETTERGP_H,ru_hard); + SetLetterBits(tr,LETTERGP_F,ru_nothard); + SetLetterBits(tr,LETTERGP_G,ru_voiced); + SetLetterBits(tr,LETTERGP_Y,ru_ivowels); + SetLetterBits(tr,LETTERGP_VOWEL2,(char *)ru_vowels); +} // end of SetCyrillicLetters + + +static void SetIndicLetters(Translator *tr) {//================================= // Set letter types for Indic scripts, Devanagari, Tamill, etc - static const char dev_consonants2[] = {0x02,0x03,0x58,0x59,0x5a,0x5b,0x5c,0x5d,0x5e,0x5f}; + static const char dev_consonants2[] = {0x02,0x03,0x58,0x59,0x5a,0x5b,0x5c,0x5d,0x5e,0x5f,0x7b,0x7c,0x7e,0x7f,0}; + static const char dev_vowels2[] = {0x60,0x61, 0x55,0x56,0x57,0x62,0x63,0}; // non-consecutive vowels and vowel-signs memset(tr->letter_bits,0,sizeof(tr->letter_bits)); - SetLetterBitsRange(tr,LETTERGP_A,0x04,0x14); // vowel letters only + SetLetterBitsRange(tr,LETTERGP_A,0x04,0x14); // vowel letters + SetLetterBitsRange(tr,LETTERGP_A,0x3e,0x4d); // + vowel signs, and virama + SetLetterBits(tr,LETTERGP_A, dev_vowels2); // + extra vowels and vowel signs + SetLetterBitsRange(tr,LETTERGP_B,0x3e,0x4d); // vowel signs, and virama + SetLetterBits(tr,LETTERGP_B, dev_vowels2); // + extra vowels and vowel signs SetLetterBitsRange(tr,LETTERGP_C,0x15,0x39); // the main consonant range SetLetterBits(tr,LETTERGP_C,dev_consonants2); // + additional consonants SetLetterBitsRange(tr,LETTERGP_Y,0x04,0x14); // vowel letters SetLetterBitsRange(tr,LETTERGP_Y,0x3e,0x4c); // + vowel signs + SetLetterBits(tr,LETTERGP_Y, dev_vowels2); // + extra vowels and vowel signs tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words + tr->langopts.suffix_add_e = tr->letter_bits_offset + 0x4d; //virama } -void SetupTranslator(Translator *tr, const short *lengths, const unsigned char *amps) + +static void SetupTranslator(Translator *tr, const short *lengths, const unsigned char *amps) {//================================================================================== if(lengths != NULL) memcpy(tr->stress_lengths,lengths,sizeof(tr->stress_lengths)); @@ -241,15 +465,26 @@ Translator *SelectTranslator(const char *name) int name2 = 0; Translator *tr; - static const unsigned char stress_amps_sk[8] = {17,17, 20,20, 20,22, 22,21 }; + static const short stress_lengths_equal[8] = {230, 230, 230, 230, 0, 0, 230, 230}; + static const unsigned char stress_amps_equal[8] = {19,19, 19,19, 19,19, 19,19 }; + + static const short stress_lengths_fr[8] = {190, 170, 190, 200, 0, 0, 190, 240}; + static const unsigned char stress_amps_fr[8] = {18,16, 18,18, 18,18, 18,18 }; + + static const unsigned char stress_amps_sk[8] = {17,16, 20,20, 20,22, 22,21 }; static const short stress_lengths_sk[8] = {190,190, 210,210, 0,0, 210,210}; + static const short stress_lengths_ta[8] = {200, 200, 210, 210, 0, 0, 230, 230}; + static const short stress_lengths_ta2[8] = {230, 230, 240, 240, 0, 0, 260, 260}; + static const unsigned char stress_amps_ta[8] = {18,18, 18,18, 20,20, 22,22 }; + + tr = NewTranslator(); + strcpy(tr->dictionary_name, name); + // convert name string into a word of up to 4 characters, for the switch() while(*name != 0) name2 = (name2 << 8) + *name++; - tr = NewTranslator(); - switch(name2) { case L('a','f'): @@ -257,18 +492,53 @@ Translator *SelectTranslator(const char *name) static const short stress_lengths_af[8] = {170,140, 220,220, 0, 0, 250,270}; SetupTranslator(tr,stress_lengths_af,NULL); - tr->langopts.stress_rule = 0; + tr->langopts.stress_rule = STRESSPOSN_1L; tr->langopts.vowel_pause = 0x30; tr->langopts.param[LOPT_DIERESES] = 1; tr->langopts.param[LOPT_PREFIXES] = 1; SetLetterVowel(tr,'y'); // add 'y' to vowels - - tr->langopts.numbers = 0x8d1 + NUM_ROMAN; + + tr->langopts.numbers = NUM_SWAP_TENS | NUM_HUNDRED_AND | NUM_SINGLE_AND | NUM_ROMAN | NUM_1900; tr->langopts.accents = 1; } break; + case L('a','m'): // Amharic, Ethiopia + { + SetupTranslator(tr,stress_lengths_fr,stress_amps_fr); + tr->letter_bits_offset = OFFSET_ETHIOPIC; + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = S_NO_AUTO_2 | S_FINAL_DIM; // don't use secondary stress + tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words + tr->langopts.numbers = NUM_OMIT_1_HUNDRED; + } + break; + + + case L('a','r'): // Arabic + tr->letter_bits_offset = OFFSET_ARABIC; + tr->langopts.numbers = NUM_SWAP_TENS | NUM_AND_UNITS | NUM_HUNDRED_AND | NUM_OMIT_1_HUNDRED | NUM_AND_HUNDRED | NUM_THOUSAND_AND | NUM_OMIT_1_THOUSAND; + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words + break; + + case L('b','g'): //Bulgarian + { + SetCyrillicLetters(tr); + SetLetterVowel(tr,0x2a); + tr->charset_a0 = charsets[5]; // ISO-8859-5 + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 0x432; // [v] don't count this character at start of word + tr->langopts.param[LOPT_REGRESSIVE_VOICING] = 0x107; // devoice at end of word, and change voicing to match a following consonant (except v) + tr->langopts.param[LOPT_REDUCE] = 2; + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_ALLOW_SPACE | NUM_OMIT_1_HUNDRED | NUM_HUNDRED_AND | NUM_AND_UNITS | NUM_SINGLE_AND | NUM_ROMAN | NUM_ROMAN_ORDINAL | NUM_ROMAN_CAPITALS ; + tr->langopts.thousands_sep = ' '; // don't allow dot as thousands separator + } + break; + case L('b','n'): // Bengali + case L('a','s'): // Assamese + case L_mni: // Manipuri (temporary placement - it's not indo-european) { static const short stress_lengths_bn[8] = {180, 180, 210, 210, 0, 0, 230, 240}; static const unsigned char stress_amps_bn[8] = {18,18, 18,18, 20,20, 22,22 }; @@ -276,14 +546,28 @@ Translator *SelectTranslator(const char *name) SetupTranslator(tr,stress_lengths_bn,stress_amps_bn); tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable - tr->langopts.stress_rule = 0; - tr->langopts.stress_flags = 0x10004; // use 'diminished' for unstressed final syllable + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = S_MID_DIM | S_FINAL_DIM; // use 'diminished' for unstressed final syllable tr->letter_bits_offset = OFFSET_BENGALI; SetIndicLetters(tr); // call this after setting OFFSET_BENGALI + SetLetterBitsRange(tr,LETTERGP_B,0x01,0x01); // candranindu SetLetterBitsRange(tr,LETTERGP_F,0x3e,0x4c); // vowel signs, but not virama - tr->langopts.numbers = 0x1; - tr->langopts.numbers2 = NUM2_100000; + tr->langopts.numbers = NUM_SWAP_TENS; + tr->langopts.break_numbers = 0x24924aa8; // for languages which have numbers for 100,000 and 100,00,000, eg Hindi + } + break; + + case L('b','o'): // Tibet + { + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->letter_bits_offset = OFFSET_TIBET; + SetLetterBitsRange(tr,LETTERGP_A,0x71,0x7d); // vowel signs + SetLetterBitsRange(tr,LETTERGP_B,0x71,0x81); // vowel signs and subjoined letters + SetLetterBitsRange(tr,LETTERGP_B,0x90,0xbc); + SetLetterBitsRange(tr,LETTERGP_C,0x40,0x6c); // consonant letters (not subjoined) + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words + tr->langopts.numbers = 1; } break; @@ -296,16 +580,16 @@ Translator *SelectTranslator(const char *name) tr->charset_a0 = charsets[14]; // ISO-8859-14 // tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable - tr->langopts.stress_rule = 2; + tr->langopts.stress_rule = STRESSPOSN_2R; // tr->langopts.intonation_group = 4; // 'diminished' is an unstressed final syllable - tr->langopts.stress_flags = 0x6 | 0x10; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2; tr->langopts.unstressed_wd1 = 0; tr->langopts.unstressed_wd2 = 2; tr->langopts.param[LOPT_SONORANT_MIN] = 120; // limit the shortening of sonorants before short vowels - tr->langopts.numbers = 0x401; + tr->langopts.numbers = NUM_OMIT_1_HUNDRED; SetLetterVowel(tr,'w'); // add letter to vowels and remove from consonants SetLetterVowel(tr,'y'); @@ -314,27 +598,47 @@ Translator *SelectTranslator(const char *name) case L('d','a'): // Danish { - static const short stress_lengths_da[8] = {160,140, 200,200, 0,0, 220,210}; + static const short stress_lengths_da[8] = {160,140, 200,200, 0,0, 220,230}; SetupTranslator(tr,stress_lengths_da,NULL); - tr->langopts.stress_rule = 0; + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.param[LOPT_PREFIXES] = 1; SetLetterVowel(tr,'y'); - tr->langopts.numbers = 0x10c59; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_SWAP_TENS | NUM_HUNDRED_AND | NUM_OMIT_1_HUNDRED | NUM_ORDINAL_DOT | NUM_1900 | NUM_ROMAN | NUM_ROMAN_CAPITALS | NUM_ROMAN_ORDINAL; } break; case L('d','e'): { - static const short stress_lengths_de[8] = {150,130, 200,200, 0, 0, 260,275}; - tr->langopts.stress_rule = 0; + static const short stress_lengths_de[8] = {150,130, 200,200, 0, 0, 270,270}; + static const unsigned char stress_amps_de[] = {20,20, 20,20, 20,22, 22,20 }; + SetupTranslator(tr, stress_lengths_de, stress_amps_de); + tr->langopts.stress_rule = STRESSPOSN_1L; tr->langopts.word_gap = 0x8; // don't use linking phonemes tr->langopts.vowel_pause = 0x30; tr->langopts.param[LOPT_PREFIXES] = 1; - memcpy(tr->stress_lengths,stress_lengths_de,sizeof(tr->stress_lengths)); - - tr->langopts.numbers = 0x11419 + NUM_ROMAN; + tr->langopts.param[LOPT_REGRESSIVE_VOICING] = 0x100; // devoice at end of word + tr->langopts.param[LOPT_LONG_VOWEL_THRESHOLD] = 175/2; + + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_SWAP_TENS | NUM_ALLOW_SPACE | NUM_ORDINAL_DOT | NUM_ROMAN; +// tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_SWAP_TENS | NUM_OMIT_1_HUNDRED | NUM_OMIT_1_THOUSAND | NUM_ALLOW_SPACE | NUM_ORDINAL_DOT | NUM_ROMAN; SetLetterVowel(tr,'y'); + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 2; // use de_rules for unpronouncable rules + } + break; + + case L('d','v'): // Divehi (Maldives) + { + SetupTranslator(tr,stress_lengths_ta,stress_amps_ta); + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words + tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable + tr->letter_bits_offset = OFFSET_THAANA; + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = S_MID_DIM | S_FINAL_DIM; // use 'diminished' for unstressed final syllable + SetLetterBitsRange(tr,LETTERGP_B,0x26,0x30); // vowel signs, and virama + tr->langopts.break_numbers = 0x14a8; // 1000, 100,000 10,000,000 + tr->langopts.numbers = 1; } break; @@ -343,9 +647,13 @@ Translator *SelectTranslator(const char *name) static const short stress_lengths_en[8] = {182,140, 220,220, 0,0, 248,275}; SetupTranslator(tr,stress_lengths_en,NULL); - tr->langopts.stress_rule = 0; - tr->langopts.numbers = 0x841 + NUM_ROMAN; + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = 0x08; + tr->langopts.numbers = NUM_HUNDRED_AND | NUM_ROMAN | NUM_1900; tr->langopts.param[LOPT_COMBINE_WORDS] = 2; // allow "mc" to cmbine with the following word + tr->langopts.suffix_add_e = 'e'; + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 2; // use en_rules for unpronouncable rules + SetLetterBits(tr,6,"aeiouy"); // Group Y: vowels, including y } break; @@ -357,10 +665,10 @@ Translator *SelectTranslator(const char *name) // character codes offset by 0x380 static const char el_vowels[] = {0x10,0x2c,0x2d,0x2e,0x2f,0x30,0x31,0x35,0x37,0x39,0x3f,0x45,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f,0}; - static const char el_fvowels[] = {0x2d,0x2e,0x2f,0x35,0x37,0x39,0x45,0x4d,0}; // ε η ι υ έ ή ί ύ - static const char el_voiceless[]= {0x38,0x3a,0x3e,0x40,0x42,0x43,0x44,0x46,0x47,0}; // θ κ ξ π ς σ τ φ χ + static const char el_fvowels[] = {0x2d,0x2e,0x2f,0x35,0x37,0x39,0x45,0x4d,0}; // ε η ι υ έ ή ί ύ _ + static const char el_voiceless[]= {0x38,0x3a,0x3e,0x40,0x42,0x43,0x44,0x46,0x47,0}; // θ κ ξ π ς σ τ φ χ _ static const char el_consonants[]={0x32,0x33,0x34,0x36,0x38,0x3a,0x3b,0x3c,0x3d,0x3e,0x40,0x41,0x42,0x43,0x44,0x46,0x47,0x48,0}; - static const wchar_t el_char_apostrophe[] = {0x3c3,0}; // σ + static const wchar_t el_char_apostrophe[] = {0x3c3,0}; // σ _ SetupTranslator(tr,stress_lengths_el,stress_amps_el); @@ -370,19 +678,20 @@ Translator *SelectTranslator(const char *name) tr->letter_bits_offset = OFFSET_GREEK; memset(tr->letter_bits,0,sizeof(tr->letter_bits)); SetLetterBits(tr,LETTERGP_A,el_vowels); + SetLetterBits(tr,LETTERGP_VOWEL2,el_vowels); SetLetterBits(tr,LETTERGP_B,el_voiceless); SetLetterBits(tr,LETTERGP_C,el_consonants); - SetLetterBits(tr,LETTERGP_Y,el_fvowels); // front vowels: ε η ι υ + SetLetterBits(tr,LETTERGP_Y,el_fvowels); // front vowels: ε η ι υ _ tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable - tr->langopts.stress_rule = 2; - tr->langopts.stress_flags = 0x6; // mark unstressed final syllables as diminished + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY; // mark unstressed final syllables as diminished tr->langopts.unstressed_wd1 = 0; tr->langopts.unstressed_wd2 = 2; tr->langopts.param[LOPT_SONORANT_MIN] = 130; // limit the shortening of sonorants before short vowels - tr->langopts.numbers = 0x109; - tr->langopts.numbers2 = 0x2; // variant form of numbers before thousands + tr->langopts.numbers = NUM_SINGLE_STRESS | NUM_DECIMAL_COMMA; + tr->langopts.numbers2 = 0x2 | NUM2_MULTIPLE_ORDINAL | NUM2_ORDINAL_NO_AND; // variant form of numbers before thousands if(name2 == L_grc) { @@ -394,74 +703,120 @@ Translator *SelectTranslator(const char *name) case L('e','o'): { - static const short stress_lengths_eo[8] = {145, 145, 230, 170, 0, 0, 360, 370}; + static const short stress_lengths_eo[8] = {150, 150, 230, 180, 0, 0, 300, 320}; static const unsigned char stress_amps_eo[] = {16,14, 20,20, 20,22, 22,21 }; static const wchar_t eo_char_apostrophe[2] = {'l',0}; - + SetupTranslator(tr,stress_lengths_eo,stress_amps_eo); tr->charset_a0 = charsets[3]; // ISO-8859-3 tr->char_plus_apostrophe = eo_char_apostrophe; - tr->langopts.word_gap = 1; +// tr->langopts.word_gap = 1; tr->langopts.vowel_pause = 2; - tr->langopts.stress_rule = 2; - tr->langopts.stress_flags = 0x6 | 0x10; - tr->langopts.unstressed_wd1 = 3; + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2; +// tr->langopts.unstressed_wd1 = 3; tr->langopts.unstressed_wd2 = 2; - tr->langopts.numbers = 0x1409 + NUM_ROMAN; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_OMIT_1_HUNDRED | NUM_ALLOW_SPACE | NUM_ROMAN; } break; case L('e','s'): // Spanish + case L('a','n'): // Aragonese case L('c','a'): // Catalan case L_pap: // Papiamento { - static const short stress_lengths_es[8] = {180, 210, 190, 190, 0, 0, 230, 260}; -// static const short stress_lengths_es[8] = {170, 200, 180, 180, 0, 0, 220, 250}; + static const short stress_lengths_es[8] = {180, 190, 230, 180, 0, 0, 240, 270}; static const unsigned char stress_amps_es[8] = {16,12, 18,18, 20,20, 20,20 }; // 'diminished' is used to mark a quieter, final unstressed syllable static const wchar_t ca_punct_within_word[] = {'\'',0xb7,0}; // ca: allow middle-dot within word SetupTranslator(tr,stress_lengths_es,stress_amps_es); tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable - tr->langopts.stress_rule = 2; + tr->langopts.stress_rule = STRESSPOSN_2R; // stress last syllable if it doesn't end in vowel or "s" or "n" // 'diminished' is an unstressed final syllable - tr->langopts.stress_flags = 0x200 | 0x6 | 0x10; + tr->langopts.stress_flags = S_FINAL_SPANISH | S_FINAL_DIM_ONLY | S_FINAL_NO_2; tr->langopts.unstressed_wd1 = 0; tr->langopts.unstressed_wd2 = 2; tr->langopts.param[LOPT_SONORANT_MIN] = 120; // limit the shortening of sonorants before short vowels - tr->langopts.numbers = 0x529 + NUM_ROMAN + NUM_ROMAN_AFTER; + tr->langopts.numbers = NUM_SINGLE_STRESS | NUM_DECIMAL_COMMA | NUM_AND_UNITS | NUM_OMIT_1_HUNDRED | NUM_OMIT_1_THOUSAND | NUM_ROMAN | NUM_ROMAN_AFTER; + tr->langopts.numbers2 = NUM2_MULTIPLE_ORDINAL | NUM2_ORDINAL_NO_AND; if(name2 == L('c','a')) { + // stress last syllable unless word ends with a vowel tr->punct_within_word = ca_punct_within_word; - tr->langopts.stress_flags = 0x200 | 0x6 | 0x30; // stress last syllable unless word ends with a vowel + tr->langopts.stress_flags = S_FINAL_SPANISH | S_FINAL_DIM_ONLY | S_FINAL_NO_2 | S_NO_AUTO_2; + } + else + if(name2 == L('a','n')) + { + tr->langopts.stress_flags = S_FINAL_SPANISH | S_FINAL_DIM_ONLY | S_FINAL_NO_2; + tr->langopts.numbers = NUM_SINGLE_STRESS | NUM_DECIMAL_COMMA | NUM_AND_UNITS | NUM_OMIT_1_HUNDRED | NUM_OMIT_1_THOUSAND | NUM_ROMAN | NUM_ROMAN_ORDINAL; + tr->langopts.numbers2 = NUM2_ORDINAL_NO_AND; + tr->langopts.roman_suffix = string_ordinal; } else if(name2 == L_pap) { - tr->langopts.stress_flags = 0x100 | 0x6 | 0x30; // stress last syllable unless word ends with a vowel + // stress last syllable unless word ends with a vowel + tr->langopts.stress_flags = S_FINAL_STRESS_C | S_FINAL_DIM_ONLY | S_FINAL_NO_2 | S_NO_AUTO_2; + } + else + { + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 2; // use es_rules for unpronouncable rules } } break; - case L('e','u'): // basque { static const short stress_lengths_eu[8] = {200, 200, 200, 200, 0, 0, 210, 230}; // very weak stress static const unsigned char stress_amps_eu[8] = {16,16, 18,18, 18,18, 18,18 }; SetupTranslator(tr,stress_lengths_eu,stress_amps_eu); - tr->langopts.stress_rule = 1; // ?? second syllable ?? - tr->langopts.numbers = 0x569 + NUM_VIGESIMAL; + tr->langopts.stress_rule = STRESSPOSN_2L; // ?? second syllable ?? + tr->langopts.numbers = NUM_SINGLE_STRESS | NUM_DECIMAL_COMMA | NUM_HUNDRED_AND | NUM_OMIT_1_HUNDRED | NUM_OMIT_1_THOUSAND | NUM_VIGESIMAL; + } + break; + + + case L('f','a'): // Farsi + { + // Convert characters in the range 0x620 to 0x6cc to the range 1 to 63. + // 0 indicates no translation for this character + static const char transpose_map_fa[] = { + 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, // 0x620 + 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 0, 0, 0, 0, 0, // 0x630 + 0, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, // 0x640 + 42, 43, 0, 0, 44, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0x650 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0x660 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 45, 0, // 0x670 + 0, 0, 0, 0, 0, 0, 46, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0x680 + 0, 0, 0, 0, 0, 0, 0, 0, 47, 0, 0, 0, 0, 0, 0, 0, // 0x690 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 48, 0, 0, 0, 0, 0, 49, // 0x6a0 + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 0x6b0 + 50, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 51 }; // 0x6c0 + tr->transpose_min = 0x620; + tr->transpose_max = 0x6cc; + tr->transpose_map = transpose_map_fa; + tr->letter_bits_offset = OFFSET_ARABIC; + + tr->langopts.numbers = NUM_AND_UNITS | NUM_HUNDRED_AND; + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words + + tr->chars_ignore = chars_ignore_zwnj_hyphen; // replace ZWNJ by hyphen } break; + case L('e','t'): // Estonian + tr->charset_a0 = charsets[4]; // ISO-8859-4 + // drop through to Finnish case L('f','i'): // Finnish { static const unsigned char stress_amps_fi[8] = {18,16, 22,22, 20,22, 22,22 }; @@ -469,12 +824,12 @@ Translator *SelectTranslator(const char *name) SetupTranslator(tr,stress_lengths_fi,stress_amps_fi); - tr->langopts.stress_rule = 0; - tr->langopts.stress_flags = 0x56; // move secondary stress from light to a following heavy syllable + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2 | S_2_TO_HEAVY; // move secondary stress from light to a following heavy syllable tr->langopts.param[LOPT_IT_DOUBLING] = 1; tr->langopts.long_stop = 130; - tr->langopts.numbers = 0x1009; + tr->langopts.numbers = NUM_DECIMAL_COMMA + NUM_ALLOW_SPACE; SetLetterVowel(tr,'y'); // tr->langopts.max_initial_consonants = 2; // BUT foreign words may have 3 tr->langopts.spelling_stress = 1; @@ -482,32 +837,37 @@ Translator *SelectTranslator(const char *name) } break; - case L('f','r'): // french { - static const short stress_lengths_fr[8] = {190, 170, 190, 200, 0, 0, 235, 240}; - static const unsigned char stress_amps_fr[8] = {18,16, 20,20, 20,22, 22,21 }; - SetupTranslator(tr,stress_lengths_fr,stress_amps_fr); - tr->langopts.stress_rule = 3; // stress on final syllable - tr->langopts.stress_flags = 0x0024; // don't use secondary stress + tr->langopts.stress_rule = STRESSPOSN_1R; // stress on final syllable + tr->langopts.stress_flags = S_NO_AUTO_2 | S_FINAL_DIM; // don't use secondary stress tr->langopts.param[LOPT_IT_LENGTHEN] = 1; // remove lengthen indicator from unstressed syllables + tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable + tr->langopts.accents = 2; // Say "Capital" after the letter. - tr->langopts.numbers = 0x1509 + 0x8000 + (NUM_NOPAUSE | NUM_ROMAN | NUM_VIGESIMAL); + tr->langopts.numbers = NUM_SINGLE_STRESS | NUM_DECIMAL_COMMA | NUM_ALLOW_SPACE | NUM_OMIT_1_HUNDRED | NUM_NOPAUSE | NUM_ROMAN | NUM_ROMAN_CAPITALS | NUM_ROMAN_AFTER | NUM_VIGESIMAL | NUM_DFRACTION_4; SetLetterVowel(tr,'y'); } break; -#ifdef deleted - case L('g','a'): // Irish Gaelic + case L('g','a'): // irish + case L('g','d'): // scots gaelic { - tr->langopts.stress_rule = 1; + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = S_NO_AUTO_2; // don't use secondary stress + tr->langopts.numbers = NUM_OMIT_1_HUNDRED | NUM_OMIT_1_THOUSAND; + tr->langopts.accents = 2; // 'capital' after letter name + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 3; // don't count apostrophe + tr->langopts.param[LOPT_IT_LENGTHEN] = 1; // remove [:] phoneme from non-stressed syllables (Lang=gd) } break; -#endif case L('h','i'): // Hindi case L('n','e'): // Nepali + case L('o','r'): // Oriya + case L('p','a'): // Punjabi + case L('g','u'): // Gujarati { static const short stress_lengths_hi[8] = {190, 190, 210, 210, 0, 0, 230, 250}; static const unsigned char stress_amps_hi[8] = {17,14, 20,19, 20,22, 22,21 }; @@ -517,10 +877,35 @@ Translator *SelectTranslator(const char *name) tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable tr->langopts.stress_rule = 6; // stress on last heaviest syllable, excluding final syllable - tr->langopts.stress_flags = 0x10004; // use 'diminished' for unstressed final syllable - tr->langopts.numbers = 0x011; - tr->langopts.numbers2 = NUM2_100000; + tr->langopts.stress_flags = S_MID_DIM | S_FINAL_DIM; // use 'diminished' for unstressed final syllable + tr->langopts.numbers = NUM_SWAP_TENS; + tr->langopts.break_numbers = 0x14aa8; // for languages which have numbers for 100,000 and 100,00,000, eg Hindi tr->letter_bits_offset = OFFSET_DEVANAGARI; + + if(name2 == L('p','a')) + { + tr->letter_bits_offset = OFFSET_GURMUKHI; + } + else + if(name2 == L('g','u')) + { +SetupTranslator(tr,stress_lengths_equal,stress_amps_equal); + tr->letter_bits_offset = OFFSET_GUJARATI; + tr->langopts.stress_rule = STRESSPOSN_2R; + } + else + if(name2 == L('n','e')) + { +SetupTranslator(tr,stress_lengths_equal,stress_amps_equal); + tr->langopts.break_numbers = 0x2aaaa8; + tr->langopts.max_digits = 22; + tr->langopts.numbers2 |= NUM2_ENGLISH_NUMERALS; + } + else + if(name2 == L('o','r')) + { + tr->letter_bits_offset = OFFSET_ORIYA; + } SetIndicLetters(tr); } break; @@ -534,22 +919,25 @@ Translator *SelectTranslator(const char *name) static const short stress_lengths_hr[8] = {180,160, 200,200, 0,0, 220,230}; static const short stress_lengths_sr[8] = {160,150, 200,200, 0,0, 250,260}; + strcpy(tr->dictionary_name, "hbs"); + if(name2 == L('s','r')) SetupTranslator(tr,stress_lengths_sr,stress_amps_hr); else SetupTranslator(tr,stress_lengths_hr,stress_amps_hr); tr->charset_a0 = charsets[2]; // ISO-8859-2 - tr->langopts.stress_rule = 0; - tr->langopts.stress_flags = 0x10; + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = S_FINAL_NO_2; tr->langopts.param[LOPT_REGRESSIVE_VOICING] = 0x3; tr->langopts.max_initial_consonants = 5; tr->langopts.spelling_stress = 1; tr->langopts.accents = 1; - tr->langopts.numbers = 0x140d + 0x4000 + NUM_ROMAN_UC; - tr->langopts.numbers2 = 0x4a; // variant numbers before thousands,milliards + tr->langopts.numbers = NUM_SINGLE_STRESS | NUM_HUNDRED_AND | NUM_OMIT_1_HUNDRED | NUM_DECIMAL_COMMA | NUM_THOUS_SPACE | NUM_DFRACTION_2 | NUM_ROMAN_CAPITALS; + tr->langopts.numbers2 = 0xa + NUM2_THOUSANDS_VAR5; // variant numbers before thousands,milliards tr->langopts.replace_chars = replace_cyrillic_latin; + tr->langopts.our_alphabet = OFFSET_CYRILLIC; // don't say "cyrillic" before letter names SetLetterVowel(tr,'y'); SetLetterVowel(tr,'r'); @@ -557,6 +945,14 @@ Translator *SelectTranslator(const char *name) break; + case L('h','t'): // Haitian Creole +// memcpy(tr->stress_lengths,stress_lengths_fr,sizeof(tr->stress_lengths)); + tr->langopts.stress_rule = STRESSPOSN_1R; // stress on final syllable + tr->langopts.stress_flags = S_NO_AUTO_2 | S_FINAL_DIM; // don't use secondary stress + tr->langopts.numbers = NUM_SINGLE_STRESS | NUM_OMIT_1_HUNDRED | NUM_NOPAUSE | NUM_ROMAN | NUM_VIGESIMAL | NUM_DFRACTION_4; + break; + + case L('h','u'): // Hungarian { static const unsigned char stress_amps_hu[8] = {17,17, 19,19, 20,22, 22,21 }; @@ -566,14 +962,17 @@ Translator *SelectTranslator(const char *name) tr->charset_a0 = charsets[2]; // ISO-8859-2 tr->langopts.vowel_pause = 0x20; - tr->langopts.stress_rule = 0; - tr->langopts.stress_flags = 0x8036; + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2 | S_NO_AUTO_2 | 0x8000 | S_HYPEN_UNSTRESS; tr->langopts.unstressed_wd1 = 2; -// tr->langopts.param[LOPT_REGRESSIVE_VOICING] = 0x4; // don't propagate over word boundaries tr->langopts.param[LOPT_IT_DOUBLING] = 1; - tr->langopts.param[LOPT_COMBINE_WORDS] = 99; // combine some prepositions with the following word + tr->langopts.param[LOPT_ANNOUNCE_PUNCT] = 2; // don't break clause before announcing . ? ! - tr->langopts.numbers = 0x1009 + 0xa000 + NUM_ROMAN + 0x10000; + tr->langopts.numbers = NUM_DFRACTION_5 | NUM_ALLOW_SPACE | NUM_ROMAN | NUM_ROMAN_ORDINAL | NUM_ROMAN_CAPITALS | NUM_ORDINAL_DOT | NUM_OMIT_1_HUNDRED | NUM_OMIT_1_THOUSAND; + tr->langopts.thousands_sep = ' '; // don't allow dot as thousands separator + tr->langopts.decimal_sep = ','; + tr->langopts.max_roman = 899; + tr->langopts.min_roman = 1; SetLetterVowel(tr,'y'); tr->langopts.spelling_stress = 1; SetLengthMods(tr,3); // all equal @@ -585,30 +984,35 @@ SetLengthMods(tr,3); // all equal static const short stress_lengths_hy[8] = {250, 200, 250, 250, 0, 0, 250, 250}; static const char hy_vowels[] = {0x31, 0x35, 0x37, 0x38, 0x3b, 0x48, 0x55, 0}; static const char hy_consonants[] = {0x32,0x33,0x34,0x36,0x39,0x3a,0x3c,0x3d,0x3e,0x3f, - 0x40,0x41,0x42,0x43,0x44,0x45,0x46,0x47,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f,0x50,0x51,0x52,0x53,0x54,0x56,0}; + 0x40,0x41,0x42,0x43,0x44, 0x46,0x47,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f,0x50,0x51,0x52,0x53,0x54,0x56,0}; + static const char hy_consonants2[] = {0x45,0}; SetupTranslator(tr,stress_lengths_hy,NULL); - tr->langopts.stress_rule = 3; // default stress on final syllable + tr->langopts.stress_rule = STRESSPOSN_1R; // default stress on final syllable tr->letter_bits_offset = OFFSET_ARMENIAN; memset(tr->letter_bits,0,sizeof(tr->letter_bits)); SetLetterBits(tr,LETTERGP_A,hy_vowels); + SetLetterBits(tr,LETTERGP_VOWEL2,hy_vowels); + SetLetterBits(tr,LETTERGP_B,hy_consonants); // not including 'j' SetLetterBits(tr,LETTERGP_C,hy_consonants); + SetLetterBits(tr,LETTERGP_C,hy_consonants2); // add 'j' tr->langopts.max_initial_consonants = 6; - tr->langopts.numbers = 0x409; -// tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_ALLOW_SPACE | NUM_OMIT_1_HUNDRED; + // tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words } break; case L('i','d'): // Indonesian + case L('m','s'): // Malay { static const short stress_lengths_id[8] = {160, 200, 180, 180, 0, 0, 220, 240}; static const unsigned char stress_amps_id[8] = {16,18, 18,18, 20,22, 22,21 }; SetupTranslator(tr,stress_lengths_id,stress_amps_id); - tr->langopts.stress_rule = 2; - tr->langopts.numbers = 0x1009 + NUM_ROMAN; - tr->langopts.stress_flags = 0x6 | 0x10; + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_ALLOW_SPACE | NUM_ROMAN; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2; tr->langopts.accents = 2; // "capital" after letter name } break; @@ -619,8 +1023,8 @@ SetLengthMods(tr,3); // all equal static const wchar_t is_lettergroup_B[] = {'c','f','h','k','p','t','x',0xfe,0}; // voiceless conants, including 'þ' ?? 's' SetupTranslator(tr,stress_lengths_is,NULL); - tr->langopts.stress_rule = 0; - tr->langopts.stress_flags = 0x10; + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = S_FINAL_NO_2; tr->langopts.param[LOPT_IT_LENGTHEN] = 0x11; // remove lengthen indicator from unstressed vowels tr->langopts.param[LOPT_REDUCE] = 2; @@ -629,7 +1033,7 @@ SetLengthMods(tr,3); // all equal SetLetterBits(tr,3,"jvr"); // Letter group H tr->letter_groups[1] = is_lettergroup_B; SetLetterVowel(tr,'y'); - tr->langopts.numbers = 0x8e9; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_SINGLE_AND | NUM_HUNDRED_AND | NUM_AND_UNITS | NUM_1900; tr->langopts.numbers2 = 0x2; } break; @@ -642,17 +1046,17 @@ SetLengthMods(tr,3); // all equal SetupTranslator(tr,stress_lengths_it,stress_amps_it); tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable - tr->langopts.stress_rule = 2; - tr->langopts.stress_flags = 0x10 | 0x20000; + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.stress_flags = S_FINAL_NO_2 | S_PRIORITY_STRESS; tr->langopts.vowel_pause = 1; tr->langopts.unstressed_wd1 = 2; tr->langopts.unstressed_wd2 = 2; tr->langopts.param[LOPT_IT_LENGTHEN] = 2; // remove lengthen indicator from unstressed or non-penultimate syllables - tr->langopts.param[LOPT_IT_DOUBLING] = 2; // double the first consonant if the previous word ends in a stressed vowel + tr->langopts.param[LOPT_IT_DOUBLING] = 1; // double the first consonant if the previous word ends in a stressed vowel (changed to =1, 23.01.2014 - only use if prev.word has $double) tr->langopts.param[LOPT_SONORANT_MIN] = 130; // limit the shortening of sonorants before short vowels tr->langopts.param[LOPT_REDUCE] = 1; // reduce vowels even if phonemes are specified in it_list tr->langopts.param[LOPT_ALT] = 2; // call ApplySpecialAttributes2() if a word has $alt or $alt2 - tr->langopts.numbers = 0x2709 + NUM_ROMAN; + tr->langopts.numbers = NUM_SINGLE_VOWEL | NUM_OMIT_1_HUNDRED |NUM_DECIMAL_COMMA | NUM_ROMAN | NUM_DFRACTION_1 | NUM_ROMAN_CAPITALS | NUM_ROMAN_AFTER; tr->langopts.accents = 2; // Say "Capital" after the letter. SetLetterVowel(tr,'y'); } @@ -664,12 +1068,71 @@ SetLengthMods(tr,3); // all equal static const wchar_t jbo_punct_within_word[] = {'.',',','\'',0x2c8,0}; // allow period and comma within a word, also stress marker (from LOPT_CAPS_IN_WORD) SetupTranslator(tr,stress_lengths_jbo,NULL); - tr->langopts.stress_rule = 2; + tr->langopts.stress_rule = STRESSPOSN_2R; tr->langopts.vowel_pause = 0x20c; // pause before a word which starts with a vowel, or after a word which ends in a consonant // tr->langopts.word_gap = 1; tr->punct_within_word = jbo_punct_within_word; tr->langopts.param[LOPT_CAPS_IN_WORD] = 2; // capitals indicate stressed syllables SetLetterVowel(tr,'y'); + tr->langopts.max_lengthmod = 368; + } + break; + + case L('k','a'): // Georgian + { + // character codes offset by 0x1080 + static const char ka_vowels[] = {0x30,0x34,0x38,0x3d,0x43,0x55,0x57,0}; + static const char ka_consonants[] = + {0x31,0x32,0x33,0x35,0x36,0x37,0x39,0x3a,0x3b,0x3c,0x3e,0x3f,0x40,0x41,0x42,0x44, + 0x45,0x46,0x47,0x48,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f,0x50,0x51,0x52,0x53,0x54,0x56,0}; + SetupTranslator(tr,stress_lengths_ta,stress_amps_ta); + memset(tr->letter_bits,0,sizeof(tr->letter_bits)); + SetLetterBits(tr,LETTERGP_A,ka_vowels); + SetLetterBits(tr,LETTERGP_C,ka_consonants); + SetLetterBits(tr,LETTERGP_VOWEL2,ka_vowels); + + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = S_FINAL_NO_2; + tr->letter_bits_offset = OFFSET_GEORGIAN; +// tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words + tr->langopts.max_initial_consonants = 7; + tr->langopts.numbers = NUM_VIGESIMAL | NUM_AND_UNITS | NUM_OMIT_1_HUNDRED |NUM_OMIT_1_THOUSAND | NUM_DFRACTION_5 | NUM_ROMAN; + + tr->langopts.alt_alphabet = OFFSET_CYRILLIC; + tr->langopts.alt_alphabet_lang = L('r','u'); + } + break; + + case L('k','k'): // Kazakh + { + static const unsigned char stress_amps_tr[8] = {18,16, 20,21, 20,21, 21,20 }; + static const short stress_lengths_tr[8] = {190,180, 230,230, 0,0, 250,250}; + + tr->letter_bits_offset = OFFSET_CYRILLIC; + memset(tr->letter_bits,0,sizeof(tr->letter_bits)); + SetLetterBits(tr,LETTERGP_A,(char *)ru_vowels); + SetLetterBits(tr,LETTERGP_C,(char *)ru_consonants); + SetLetterBits(tr,LETTERGP_VOWEL2,(char *)ru_vowels); + + SetupTranslator(tr,stress_lengths_tr,stress_amps_tr); + + tr->langopts.stress_rule = 7; // stress on the last syllable, before any explicitly unstressed syllable + tr->langopts.stress_flags = S_NO_AUTO_2 + S_NO_EOC_LENGTHEN; //no automatic secondary stress, don't lengthen at end-of-clause + tr->langopts.lengthen_tonic = 0; + tr->langopts.param[LOPT_SUFFIX] = 1; + + tr->langopts.numbers = NUM_OMIT_1_HUNDRED | NUM_DFRACTION_6 ; + tr->langopts.max_initial_consonants = 2; +SetLengthMods(tr,3); // all equal + } + break; + + case L('k','l'): // Greenlandic + { + SetupTranslator(tr,stress_lengths_equal,stress_amps_equal); + tr->langopts.stress_rule = 12; + tr->langopts.stress_flags = S_NO_AUTO_2; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_SWAP_TENS | NUM_HUNDRED_AND | NUM_OMIT_1_HUNDRED | NUM_ORDINAL_DOT | NUM_1900 | NUM_ROMAN | NUM_ROMAN_CAPITALS | NUM_ROMAN_ORDINAL; } break; @@ -679,6 +1142,7 @@ SetLengthMods(tr,3); // all equal static const unsigned char ko_voiced[] = {0x02,0x05,0x06,0xab,0xaf,0xb7,0xbc,0}; // voiced consonants, l,m,n,N tr->letter_bits_offset = OFFSET_KOREAN; + tr->langopts.our_alphabet = 0xa700; memset(tr->letter_bits,0,sizeof(tr->letter_bits)); SetLetterBitsRange(tr,LETTERGP_A,0x61,0x75); SetLetterBits(tr,LETTERGP_Y,ko_ivowels); @@ -686,7 +1150,10 @@ SetLengthMods(tr,3); // all equal tr->langopts.stress_rule = 8; // ?? 1st syllable if it is heavy, else 2nd syllable tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words - tr->langopts.numbers = 0x0401; + tr->langopts.numbers = NUM_OMIT_1_HUNDRED; + tr->langopts.numbers2 = NUM2_MYRIADS; + tr->langopts.break_numbers = 0x1111110; + tr->langopts.max_digits = 20; } break; @@ -700,7 +1167,7 @@ SetLengthMods(tr,3); // all equal tr->langopts.stress_rule = 7; // stress on the last syllable, before any explicitly unstressed syllable - tr->langopts.numbers = 0x100461; + tr->langopts.numbers = NUM_HUNDRED_AND | NUM_AND_UNITS | NUM_OMIT_1_HUNDRED | NUM_AND_HUNDRED; tr->langopts.max_initial_consonants = 2; } break; @@ -708,12 +1175,26 @@ SetLengthMods(tr,3); // all equal case L('l','a'): //Latin { tr->charset_a0 = charsets[4]; // ISO-8859-4, includes a,e,i,o,u-macron - tr->langopts.stress_rule = 2; - tr->langopts.stress_flags = 0x20; + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.stress_flags = S_NO_AUTO_2; tr->langopts.unstressed_wd1 = 0; tr->langopts.unstressed_wd2 = 2; tr->langopts.param[LOPT_DIERESES] = 1; - tr->langopts.numbers = 0x1 + NUM_ROMAN; + tr->langopts.numbers = NUM_ROMAN; + tr->langopts.max_roman = 5000; + } + break; + + case L('l','t'): // Lithuanian + { + tr->charset_a0 = charsets[4]; // ISO-8859-4 + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.stress_flags = S_NO_AUTO_2; + tr->langopts.unstressed_wd1 = 0; + tr->langopts.unstressed_wd2 = 2; + tr->langopts.param[LOPT_DIERESES] = 1; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_OMIT_1_HUNDRED | NUM_DFRACTION_4 | NUM_ORDINAL_DOT; + tr->langopts.numbers2 = NUM2_THOUSANDS_VAR4; tr->langopts.max_roman = 5000; } break; @@ -725,11 +1206,11 @@ SetLengthMods(tr,3); // all equal SetupTranslator(tr,stress_lengths_lv,stress_amps_lv); - tr->langopts.stress_rule = 0; + tr->langopts.stress_rule = STRESSPOSN_1L; tr->langopts.spelling_stress = 1; tr->charset_a0 = charsets[4]; // ISO-8859-4 - tr->langopts.numbers = 0x409 + 0x8000 + 0x10000; - tr->langopts.stress_flags = 0x16 + 0x40000; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_OMIT_1_HUNDRED | NUM_DFRACTION_4 | NUM_ORDINAL_DOT; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2 | S_EO_CLAUSE1; } break; @@ -742,38 +1223,50 @@ SetLengthMods(tr,3); // all equal SetupTranslator(tr,stress_lengths_mk,stress_amps_mk); tr->charset_a0 = charsets[5]; // ISO-8859-5 - tr->letter_groups[0] = vowels_cyrillic; + tr->letter_groups[0] = tr->letter_groups[7] = vowels_cyrillic; + tr->letter_bits_offset = OFFSET_CYRILLIC; - tr->langopts.stress_rule = 4; // antipenultimate - tr->langopts.numbers = 0x0429 + 0x4000; + tr->langopts.stress_rule = STRESSPOSN_3R; // antipenultimate + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_AND_UNITS | NUM_OMIT_1_HUNDRED | NUM_OMIT_1_THOUSAND | NUM_DFRACTION_2; tr->langopts.numbers2 = 0x8a; // variant numbers before thousands,milliards } break; + case L('m','t'): // Maltese + { + tr->charset_a0 = charsets[3]; // ISO-8859-3 + tr->langopts.param[LOPT_REGRESSIVE_VOICING] = 0x100; // devoice at end of word + tr->langopts.stress_rule = STRESSPOSN_2R; // penultimate + tr->langopts.numbers = 1; + } + break; case L('n','l'): // Dutch { static const short stress_lengths_nl[8] = {160,135, 210,210, 0, 0, 260,280}; - tr->langopts.stress_rule = 0; - tr->langopts.vowel_pause = 1; + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.vowel_pause = 0x30; // ?? tr->langopts.param[LOPT_DIERESES] = 1; tr->langopts.param[LOPT_PREFIXES] = 1; + tr->langopts.param[LOPT_REGRESSIVE_VOICING] = 0x100; // devoice at end of word SetLetterVowel(tr,'y'); - - tr->langopts.numbers = 0x11c19; + + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_SWAP_TENS | NUM_OMIT_1_HUNDRED | NUM_OMIT_1_THOUSAND | NUM_ALLOW_SPACE | NUM_1900 | NUM_ORDINAL_DOT; + tr->langopts.ordinal_indicator = "e"; + tr->langopts.stress_flags = S_FIRST_PRIMARY; memcpy(tr->stress_lengths,stress_lengths_nl,sizeof(tr->stress_lengths)); } break; case L('n','o'): // Norwegian { - static const short stress_lengths_no[8] = {160,140, 200,200, 0,0, 220,210}; + static const short stress_lengths_no[8] = {160,140, 200,200, 0,0, 220,230}; SetupTranslator(tr,stress_lengths_no,NULL); - tr->langopts.stress_rule = 0; + tr->langopts.stress_rule = STRESSPOSN_1L; SetLetterVowel(tr,'y'); - tr->langopts.numbers = 0x11849; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_HUNDRED_AND | NUM_ALLOW_SPACE | NUM_1900 | NUM_ORDINAL_DOT; } break; @@ -783,8 +1276,8 @@ SetLengthMods(tr,3); // all equal static const short stress_lengths_om[8] = {200,200, 200,200, 0,0, 200,200}; SetupTranslator(tr,stress_lengths_om,stress_amps_om); - tr->langopts.stress_rule = 2; - tr->langopts.stress_flags = 0x16 + 0x80000; + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2 | 0x80000; } break; @@ -796,12 +1289,12 @@ SetLengthMods(tr,3); // all equal SetupTranslator(tr,stress_lengths_pl,stress_amps_pl); tr->charset_a0 = charsets[2]; // ISO-8859-2 - tr->langopts.stress_rule = 2; - tr->langopts.stress_flags = 0x6; // mark unstressed final syllables as diminished - tr->langopts.param[LOPT_REGRESSIVE_VOICING] = 0x8; + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY; // mark unstressed final syllables as diminished + tr->langopts.param[LOPT_REGRESSIVE_VOICING] = 0x9; tr->langopts.max_initial_consonants = 7; // for example: wchrzczony :) - tr->langopts.numbers=0x1009 + 0x4000; - tr->langopts.numbers2=0x40; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_ALLOW_SPACE | NUM_DFRACTION_2; + tr->langopts.numbers2 = NUM2_THOUSANDS_VAR3; tr->langopts.param[LOPT_COMBINE_WORDS] = 4 + 0x100; // combine 'nie' (marked with $alt2) with some 1-syllable (and 2-syllable) words (marked with $alt) SetLetterVowel(tr,'y'); } @@ -809,18 +1302,22 @@ SetLengthMods(tr,3); // all equal case L('p','t'): // Portuguese { - static const short stress_lengths_pt[8] = {180, 125, 210, 210, 0, 0, 270, 295}; - static const unsigned char stress_amps_pt[8] = {16,13, 19,19, 20,22, 22,21 }; // 'diminished' is used to mark a quieter, final unstressed syllable + static const short stress_lengths_pt[8] = {170, 115, 210, 240, 0, 0, 260, 280}; + static const unsigned char stress_amps_pt[8] = {16,11, 19,21, 20,22, 22,21 }; // 'diminished' is used to mark a quieter, final unstressed syllable SetupTranslator(tr,stress_lengths_pt,stress_amps_pt); tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable - tr->langopts.stress_rule = 3; // stress on final syllable - tr->langopts.stress_flags = 0x6 | 0x10 | 0x20000; - tr->langopts.numbers = 0x069 + 0x4000 + NUM_ROMAN; + tr->langopts.stress_rule = STRESSPOSN_1R; // stress on final syllable + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2 | S_INITIAL_2 | S_PRIORITY_STRESS; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_DFRACTION_2 | NUM_HUNDRED_AND | NUM_AND_UNITS | NUM_ROMAN_CAPITALS; + tr->langopts.numbers2 = NUM2_MULTIPLE_ORDINAL | NUM2_NO_TEEN_ORDINALS | NUM2_ORDINAL_NO_AND; + tr->langopts.max_roman = 5000; SetLetterVowel(tr,'y'); ResetLetterBits(tr,0x2); SetLetterBits(tr,1,"bcdfgjkmnpqstvxz"); // B hard consonants, excluding h,l,r,w,y + tr->langopts.param[LOPT_ALT] = 2; // call ApplySpecialAttributes2() if a word has $alt or $alt2 + tr->langopts.accents = 2; // 'capital' after letter name } break; @@ -831,11 +1328,11 @@ SetLengthMods(tr,3); // all equal SetupTranslator(tr,stress_lengths_ro,stress_amps_ro); - tr->langopts.stress_rule = 2; - tr->langopts.stress_flags = 0x100 + 0x6; + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.stress_flags = S_FINAL_STRESS_C + S_FINAL_DIM_ONLY; tr->charset_a0 = charsets[2]; // ISO-8859-2 - tr->langopts.numbers = 0x1029+0x6000 + NUM_ROMAN; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_ALLOW_SPACE | NUM_DFRACTION_3 | NUM_AND_UNITS | NUM_ROMAN; tr->langopts.numbers2 = 0x1e; // variant numbers before all thousandplex } break; @@ -846,11 +1343,11 @@ SetLengthMods(tr,3); // all equal case L('r','w'): // Kiryarwanda { - tr->langopts.stress_rule = 2; - tr->langopts.stress_flags = 0x16; + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2; tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable - - tr->langopts.numbers = 0x61 + 0x100000 + 0x4000; + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words. Need to allow "bw'" prefix + tr->langopts.numbers = NUM_HUNDRED_AND | NUM_AND_UNITS | NUM_DFRACTION_2 | NUM_AND_HUNDRED; tr->langopts.numbers2 = 0x200; // say "thousands" before its number } break; @@ -863,16 +1360,16 @@ SetLengthMods(tr,3); // all equal SetupTranslator(tr,stress_lengths_sk,stress_amps_sk); tr->charset_a0 = charsets[2]; // ISO-8859-2 - tr->langopts.stress_rule = 0; - tr->langopts.stress_flags = 0x16; + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2; tr->langopts.param[LOPT_REGRESSIVE_VOICING] = 0x3; tr->langopts.max_initial_consonants = 5; tr->langopts.spelling_stress = 1; tr->langopts.param[LOPT_COMBINE_WORDS] = 4; // combine some prepositions with the following word - tr->langopts.numbers = 0x0401 + 0x4000 + NUM_ROMAN; - tr->langopts.numbers2 = 0x100; - tr->langopts.thousands_sep = 0; //no thousands separator + tr->langopts.numbers = NUM_OMIT_1_HUNDRED | NUM_DFRACTION_2 | NUM_ROMAN; + tr->langopts.numbers2 = NUM2_THOUSANDS_VAR2; + tr->langopts.thousands_sep = STRESSPOSN_1L; //no thousands separator tr->langopts.decimal_sep = ','; if(name2 == L('c','s')) @@ -887,6 +1384,47 @@ SetLengthMods(tr,3); // all equal } break; + case L('s','i'): // Sinhala + { + SetupTranslator(tr,stress_lengths_ta,stress_amps_ta); + tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable + + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2; + tr->langopts.spelling_stress = 1; + + tr->letter_bits_offset = OFFSET_SINHALA; + memset(tr->letter_bits,0,sizeof(tr->letter_bits)); + SetLetterBitsRange(tr,LETTERGP_A,0x05,0x16); // vowel letters + SetLetterBitsRange(tr,LETTERGP_A,0x4a,0x73); // + vowel signs, and virama + + SetLetterBitsRange(tr,LETTERGP_B,0x4a,0x73); // vowel signs, and virama + + SetLetterBitsRange(tr,LETTERGP_C,0x1a,0x46); // the main consonant range + + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words + tr->langopts.suffix_add_e = tr->letter_bits_offset + 0x4a; //virama + tr->langopts.numbers = NUM_OMIT_1_THOUSAND | NUM_SINGLE_STRESS_L | NUM_DFRACTION_7; + tr->langopts.numbers2 = NUM2_PERCENT_BEFORE; + tr->langopts.break_numbers = 0x14aa8; // for languages which have numbers for 100,000 and 100,00,000, eg Hindi + } + break; + + case L('s','l'): // Slovenian + tr->charset_a0 = charsets[2]; // ISO-8859-2 + tr->langopts.stress_rule = STRESSPOSN_2R; // Temporary + tr->langopts.stress_flags = S_NO_AUTO_2; + tr->langopts.param[LOPT_REGRESSIVE_VOICING] = 0x103; + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 0x76; // [v] don't count this character at start of word + tr->langopts.param[LOPT_ALT] = 2; // call ApplySpecialAttributes2() if a word has $alt or $alt2 + tr->langopts.param[LOPT_IT_LENGTHEN] = 1; // remove lengthen indicator from unstressed syllables + tr->letter_bits['r'] |= 0x80; // add 'r' to letter group 7, vowels for Unpronouncable test + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_ALLOW_SPACE | NUM_SWAP_TENS | NUM_OMIT_1_HUNDRED | NUM_DFRACTION_2 | NUM_ORDINAL_DOT | NUM_ROMAN; + tr->langopts.numbers2 = 0x100; // plural forms of millions etc + tr->langopts.thousands_sep = ' '; // don't allow dot as thousands separator + tr->langopts.replace_chars = replace_cyrillic_latin; + break; + case L('s','q'): // Albanian { static const short stress_lengths_sq[8] = {150, 150, 180, 180, 0, 0, 300, 300}; @@ -894,10 +1432,10 @@ SetLengthMods(tr,3); // all equal SetupTranslator(tr,stress_lengths_sq,stress_amps_sq); - tr->langopts.stress_rule = 2; - tr->langopts.stress_flags = 0x16 + 0x100; + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2 | S_FINAL_STRESS_C; SetLetterVowel(tr,'y'); - tr->langopts.numbers = 0x69 + 0x8000; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_HUNDRED_AND | NUM_AND_UNITS | NUM_DFRACTION_4; tr->langopts.accents = 2; // "capital" after letter name } break; @@ -909,14 +1447,15 @@ SetLengthMods(tr,3); // all equal static const short stress_lengths_sv[8] = {160,135, 220,220, 0,0, 250,280}; SetupTranslator(tr,stress_lengths_sv,stress_amps_sv); - tr->langopts.stress_rule = 0; + tr->langopts.stress_rule = STRESSPOSN_1L; SetLetterVowel(tr,'y'); - tr->langopts.numbers = 0x1909; + tr->langopts.numbers = NUM_SINGLE_STRESS | NUM_DECIMAL_COMMA | NUM_ALLOW_SPACE | NUM_1900; tr->langopts.accents = 1; } break; case L('s','w'): // Swahili + case L('t','n'): // Setswana { static const short stress_lengths_sw[8] = {160, 170, 200, 200, 0, 0, 320, 340}; static const unsigned char stress_amps_sw[] = {16,12, 19,19, 20,22, 22,21 }; @@ -925,29 +1464,39 @@ SetLengthMods(tr,3); // all equal tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable tr->langopts.vowel_pause = 1; - tr->langopts.stress_rule = 2; - tr->langopts.stress_flags = 0x6 | 0x10; + tr->langopts.stress_rule = STRESSPOSN_2R; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2; + tr->langopts.max_initial_consonants = 4; // for example: mwngi + - tr->langopts.numbers = 0x4e1; - tr->langopts.numbers2 = NUM2_100000a; + tr->langopts.numbers = NUM_AND_UNITS | NUM_HUNDRED_AND | NUM_SINGLE_AND | NUM_OMIT_1_HUNDRED; + tr->langopts.break_numbers = 0x49249268; // for languages which have numbers for 100,000 and 1,000,000 } break; case L('t','a'): // Tamil - case L('m','l'): // Malayalam case L('k','n'): // Kannada + case L('m','l'): // Malayalam case L('m','r'): // Marathi + case L('t','e'): // Telugu { - static const short stress_lengths_ta[8] = {200, 200, 210, 210, 0, 0, 230, 230}; - static const unsigned char stress_amps_ta[8] = {18,18, 18,18, 20,20, 22,22 }; - - SetupTranslator(tr,stress_lengths_ta,stress_amps_ta); + SetupTranslator(tr,stress_lengths_ta2, stress_amps_ta); tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable - tr->langopts.stress_rule = 0; - tr->langopts.stress_flags = 0x10004; // use 'diminished' for unstressed final syllable - tr->letter_bits_offset = OFFSET_TAMIL; + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.stress_flags = S_FINAL_DIM_ONLY | S_FINAL_NO_2; // use 'diminished' for unstressed final syllable + tr->langopts.spelling_stress = 1; + tr->langopts.break_numbers = 0x14a8; // 1000, 100,000 10,000,000 + if(name2 == L('t','a')) + { + SetupTranslator(tr,stress_lengths_ta, NULL); + tr->letter_bits_offset = OFFSET_TAMIL; + tr->langopts.numbers = NUM_OMIT_1_THOUSAND ; + tr->langopts.numbers2 = NUM2_ORDINAL_AND_THOUSANDS; + tr->langopts.param[LOPT_WORD_MERGE] = 1; // don't break vowels betwen words + } + else if(name2 == L('m','r')) { tr->letter_bits_offset = OFFSET_DEVANAGARI; @@ -955,17 +1504,27 @@ SetLengthMods(tr,3); // all equal else if(name2 == L('m','l')) { + static const short stress_lengths_ml[8] = {180, 160, 240, 240, 0, 0, 260, 260}; + SetupTranslator(tr,stress_lengths_ml, stress_amps_equal); tr->letter_bits_offset = OFFSET_MALAYALAM; + tr->langopts.numbers = NUM_OMIT_1_THOUSAND | NUM_OMIT_1_HUNDRED; + tr->langopts.numbers2 = NUM2_OMIT_1_HUNDRED_ONLY; + tr->langopts.stress_rule = 13; // 1st syllable, unless 1st vowel is short and 2nd is long } else if(name2 == L('k','n')) { tr->letter_bits_offset = OFFSET_KANNADA; tr->langopts.numbers = 0x1; - tr->langopts.numbers2 = NUM2_100000; } - tr->langopts.param[LOPT_WORD_MERGE] = 1; // don't break vowels betwen words + else + if(name2 == L('t','e')) + { + tr->letter_bits_offset = OFFSET_TELUGU; + tr->langopts.numbers = 0x1; + } SetIndicLetters(tr); // call this after setting OFFSET_ + SetLetterBitsRange(tr,LETTERGP_B,0x4e,0x4e); // chillu-virama (unofficial) } break; @@ -978,7 +1537,7 @@ SetLengthMods(tr,3); // all equal SetupTranslator(tr,stress_lengths_th,stress_amps_th); tr->langopts.stress_rule = 0; // stress on final syllable of a "word" - tr->langopts.stress_flags = 2; // don't automatically set diminished stress (may be set in the intonation module) + tr->langopts.stress_flags = S_NO_DIM; // don't automatically set diminished stress (may be set in the intonation module) tr->langopts.tone_language = 1; // Tone language, use CalcPitches_Tone() rather than CalcPitches() tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable // tr->langopts.tone_numbers = 1; // a number after letters indicates a tone number (eg. pinyin or jyutping) @@ -988,21 +1547,58 @@ SetLengthMods(tr,3); // all equal #endif case L('t','r'): // Turkish + case L('a','z'): // Azerbaijan { - static const unsigned char stress_amps_tr[8] = {18,18, 20,20, 20,22, 22,21 }; - static const short stress_lengths_tr[8] = {190,190, 190,190, 0,0, 250,270}; + static const unsigned char stress_amps_tr[8] = {18,16, 20,21, 20,21, 21,20 }; + static const short stress_lengths_tr[8] = {190,180, 200,230, 0,0, 240,250}; SetupTranslator(tr,stress_lengths_tr,stress_amps_tr); tr->charset_a0 = charsets[9]; // ISO-8859-9 - Latin5 tr->langopts.stress_rule = 7; // stress on the last syllable, before any explicitly unstressed syllable - tr->langopts.stress_flags = 0x20; //no automatic secondary stress + tr->langopts.stress_flags = S_NO_AUTO_2; //no automatic secondary stress + tr->langopts.dotless_i = 1; + tr->langopts.param[LOPT_SUFFIX] = 1; - tr->langopts.numbers = 0x1509 + 0x4000; + if(name2 == L('a','z')) + { + tr->langopts.numbers = NUM_SINGLE_STRESS | NUM_DECIMAL_COMMA | NUM_ALLOW_SPACE | NUM_OMIT_1_HUNDRED | NUM_OMIT_1_THOUSAND | NUM_DFRACTION_2; + } + else + { + tr->langopts.numbers = NUM_SINGLE_STRESS | NUM_DECIMAL_COMMA | NUM_OMIT_1_HUNDRED | NUM_OMIT_1_THOUSAND | NUM_DFRACTION_2; + } tr->langopts.max_initial_consonants = 2; } break; + case L('t','t'): // Tatar + { + SetCyrillicLetters(tr); + SetupTranslator(tr,stress_lengths_fr,stress_amps_fr); + tr->langopts.stress_rule = STRESSPOSN_1R; // stress on final syllable + tr->langopts.stress_flags = S_NO_AUTO_2; //no automatic secondary stress + tr->langopts.numbers = NUM_SINGLE_STRESS | NUM_DECIMAL_COMMA | NUM_OMIT_1_HUNDRED | NUM_OMIT_1_THOUSAND | NUM_DFRACTION_4; + } + break; + + case L('u','k'): // Ukrainian + { + SetCyrillicLetters(tr); + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 0x432; // [v] don't count this character at start of word + } + break; + + case L('u','r'): // Urdu + case L('s','d'): // Sindhi + { + tr->letter_bits_offset = OFFSET_ARABIC; + tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1; // disable check for unpronouncable words + tr->langopts.numbers = NUM_SWAP_TENS; + tr->langopts.break_numbers = 0x52a8; // for languages which have numbers for 100,000 and 100,00,000, eg Hindi + } + break; + case L('v','i'): // Vietnamese { static const short stress_lengths_vi[8] = {150, 150, 180, 180, 210, 230, 230, 240}; @@ -1019,22 +1615,27 @@ SetLengthMods(tr,3); // all equal 0x1a1, 0x1edd, 0x1edb, 0x1edf, 0x1ee1, 0x1ee3, // ơ 0x75, 0xf9, 0xfa, 0x1ee7, 0x169, 0x1ee5, // u 0x1b0, 0x1eeb, 0x1ee9, 0x1eed, 0x1eef, 0x1ef1, // ư - 0x79, 0x1ef3, 0xfd, 0x1ef7, 0x1ef9, 0x1e, 0 }; // y + 0x79, 0x1ef3, 0xfd, 0x1ef7, 0x1ef9, 0x1ef5, 0 }; // y SetupTranslator(tr,stress_lengths_vi,stress_amps_vi); tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable - tr->langopts.stress_rule = 0; + tr->langopts.stress_rule = STRESSPOSN_1L; tr->langopts.word_gap = 0x21; // length of a final vowel is less dependent on the next consonant, don't merge consonant with next word // tr->langopts.vowel_pause = 4; - tr->letter_groups[0] = vowels_vi; + tr->letter_groups[0] = tr->letter_groups[7] = vowels_vi; tr->langopts.tone_language = 1; // Tone language, use CalcPitches_Tone() rather than CalcPitches() tr->langopts.unstressed_wd1 = 2; - tr->langopts.numbers = 0x0049 + 0x8000; + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_HUNDRED_AND_DIGIT | NUM_DFRACTION_4 | NUM_ZERO_HUNDRED; } break; + case L('w','o'): + tr->langopts.stress_rule = STRESSPOSN_1L; + tr->langopts.numbers = NUM_AND_UNITS | NUM_HUNDRED_AND | NUM_OMIT_1_HUNDRED | NUM_OMIT_1_THOUSAND | NUM_SINGLE_STRESS; + break; + case L('z','h'): case L_zhy: { @@ -1043,13 +1644,14 @@ SetLengthMods(tr,3); // all equal SetupTranslator(tr,stress_lengths_zh,stress_amps_zh); - tr->langopts.stress_rule = 3; // stress on final syllable of a "word" - tr->langopts.stress_flags = 2; // don't automatically set diminished stress (may be set in the intonation module) + tr->langopts.stress_rule = STRESSPOSN_1R; // stress on final syllable of a "word" + tr->langopts.stress_flags = S_NO_DIM; // don't automatically set diminished stress (may be set in the intonation module) tr->langopts.vowel_pause = 0; tr->langopts.tone_language = 1; // Tone language, use CalcPitches_Tone() rather than CalcPitches() tr->langopts.length_mods0 = tr->langopts.length_mods; // don't lengthen vowels in the last syllable tr->langopts.tone_numbers = 1; // a number after letters indicates a tone number (eg. pinyin or jyutping) tr->langopts.ideographs = 1; + tr->langopts.our_alphabet = 0x3100; tr->langopts.word_gap = 0x21; // length of a final vowel is less dependent on the next consonant, don't merge consonant with next word if(name2 == L('z','h')) { @@ -1066,66 +1668,48 @@ SetLengthMods(tr,3); // all equal tr->translator_name = name2; - if(tr->langopts.numbers & 0x8) + ProcessLanguageOptions(&tr->langopts); + return(tr); +} // end of SelectTranslator + + +void ProcessLanguageOptions(LANGUAGE_OPTIONS *langopts) +{//===================================================== + if(langopts->numbers & NUM_DECIMAL_COMMA) { // use . and ; for thousands and decimal separators - tr->langopts.thousands_sep = '.'; - tr->langopts.decimal_sep = ','; + langopts->thousands_sep = '.'; + langopts->decimal_sep = ','; } - if(tr->langopts.numbers & 0x4) + if(langopts->numbers & NUM_THOUS_SPACE) { - tr->langopts.thousands_sep = 0; // don't allow thousands separator, except space + langopts->thousands_sep = 0; // don't allow thousands separator, except space } - return(tr); -} // end of SelectTranslator - - +} //********************************************************************************************************** + static void Translator_Russian(Translator *tr) {//=========================================== static const unsigned char stress_amps_ru[] = {16,16, 18,18, 20,24, 24,22 }; static const short stress_lengths_ru[8] = {150,140, 220,220, 0,0, 260,280}; - - - // character codes offset by 0x420 - static const char ru_vowels[] = {0x10,0x15,0x31,0x18,0x1e,0x23,0x2b,0x2d,0x2e,0x2f,0}; - static const char ru_consonants[] = {0x11,0x12,0x13,0x14,0x16,0x17,0x19,0x1a,0x1b,0x1c,0x1d,0x1f,0x20,0x21,0x22,0x24,0x25,0x26,0x27,0x28,0x29,0x2a,0x2c,0}; - static const char ru_soft[] = {0x2c,0x19,0x27,0x29,0}; // letter group B [k ts; s;] - static const char ru_hard[] = {0x2a,0x16,0x26,0x28,0}; // letter group H [S Z ts] - static const char ru_nothard[] = {0x11,0x12,0x13,0x14,0x17,0x19,0x1a,0x1b,0x1c,0x1d,0x1f,0x20,0x21,0x22,0x24,0x25,0x27,0x29,0x2c,0}; - static const char ru_voiced[] = {0x11,0x12,0x13,0x14,0x16,0x17,0}; // letter group G (voiced obstruents) - static const char ru_ivowels[] = {0x2c,0x15,0x31,0x18,0x2e,0x2f,0}; // letter group Y (iotated vowels & soft-sign) + static const char ru_ivowels2[] = {0x2c,0x15,0x18,0x2e,0x2f,0}; // add more vowels to letter group Y (iotated vowels & soft-sign) SetupTranslator(tr,stress_lengths_ru,stress_amps_ru); - - tr->charset_a0 = charsets[18]; // KOI8-R - tr->transpose_offset = 0x42f; // convert cyrillic from unicode into range 0x01 to 0x22 - tr->transpose_min = 0x430; - tr->transpose_max = 0x451; - - tr->letter_bits_offset = OFFSET_CYRILLIC; - memset(tr->letter_bits,0,sizeof(tr->letter_bits)); - SetLetterBits(tr,0,ru_vowels); - SetLetterBits(tr,1,ru_soft); - SetLetterBits(tr,2,ru_consonants); - SetLetterBits(tr,3,ru_hard); - SetLetterBits(tr,4,ru_nothard); - SetLetterBits(tr,5,ru_voiced); - SetLetterBits(tr,6,ru_ivowels); - SetLetterBits(tr,7,ru_vowels); + SetCyrillicLetters(tr); + SetLetterBits(tr,6,ru_ivowels2); tr->langopts.param[LOPT_UNPRONOUNCABLE] = 0x432; // [v] don't count this character at start of word tr->langopts.param[LOPT_REGRESSIVE_VOICING] = 1; tr->langopts.param[LOPT_REDUCE] = 2; tr->langopts.stress_rule = 5; - tr->langopts.stress_flags = 0x0020; // waas 0x1010 + tr->langopts.stress_flags = S_NO_AUTO_2; - tr->langopts.numbers = 0x0409; - tr->langopts.numbers2 = 0xc2; // variant numbers before thousands + tr->langopts.numbers = NUM_DECIMAL_COMMA | NUM_OMIT_1_HUNDRED; + tr->langopts.numbers2 = 0x2 + NUM2_THOUSANDS_VAR1; // variant numbers before thousands tr->langopts.phoneme_change = 1; tr->langopts.testing = 2; @@ -1145,8 +1729,9 @@ typedef struct { */ -#define RUSSIAN2 + #ifdef RUSSIAN2 +// This is now done in the phoneme data, ph_russian int ChangePhonemes_ru(Translator *tr, PHONEME_LIST2 *phlist, int n_ph, int index, PHONEME_TAB *ph, CHANGEPH *ch) {//============================================================================================================= @@ -1155,8 +1740,6 @@ int ChangePhonemes_ru(Translator *tr, PHONEME_LIST2 *phlist, int n_ph, int index int variant; int vowelix; - int stressed; - int soft; PHONEME_TAB *prev, *next; if(ch->flags & 8) @@ -1179,7 +1762,7 @@ PH('V','#'),PH('I','3'),PH('I','2'),PH('E','3')}; /*2*/ {'o', '8', '8', 'o', '8', '8'}, // O /*3*/ {'i', 'I', 'i', 'a', 'I', 'a'}, // I Uses 3,4,5 columns. /*4*/ {'i', PH('I','#'), 'i', 'i', PH('I','#'), 'i'}, // I# - /*5*/ {'E', PH('E','#'), 'E', 'e', PH('E','#'), 'e'}, // E# + /*5*/ {'E', PH('E','#'), 'E', 'e', PH('E','#'), 'e'}, // E# /*6*/ {'E', PH('E','2'), 'E', 'e', PH('E','2'), 'e'}, // E2 Uses 3,4,5 columns. /*7*/ {PH('j','a'), 'V', PH('j','a'), 'A', 'V', 'A'}, // V# /*8*/ {PH('j','a'), 'I', PH('j','a'), 'e', 'I', 'e'}, // I3 Uses 3,4,5 columns. @@ -1216,8 +1799,8 @@ PH('V','#'),PH('I','3'),PH('I','2'),PH('E','3')}; } // do we need a variant of this vowel, depending on the stress and adjacent phonemes ? variant = -1; - stressed = ch->flags & 2; - soft=prev->phflags & phPALATAL; + int stressed = ch->flags & 2; + int soft=prev->phflags & phPALATAL; if (soft && stressed) variant = 2; else @@ -1240,96 +1823,6 @@ PH('V','#'),PH('I','3'),PH('I','2'),PH('E','3')}; return(0); } -#else - - -int ChangePhonemes_ru(Translator *tr, PHONEME_LIST2 *phlist, int n_ph, int index, PHONEME_TAB *ph, CHANGEPH *ch) -{//============================================================================================================= -// Called for each phoneme in the phoneme list, to allow a language to make changes -// flags: bit 0=1 last phoneme in a word -// bit 1=1 this is the highest stressed vowel in the current word -// bit 2=1 after the highest stressed vowel in the current word -// bit 3=1 the phonemes were specified explicitly, or found from an entry in the xx_list dictionary -// ph The current phoneme - - int variant; - int vowelix; - PHONEME_TAB *prev, *next; - - if(ch->flags & 8) - return(0); // full phoneme translation has already been given - - // Russian vowel softening and reduction rules - if(ph->type == phVOWEL) - { - #define N_VOWELS_RU 7 - static unsigned char vowels_ru[N_VOWELS_RU] = {'a','A','o','E','i','u','y'}; - - // each line gives: soft, reduced, soft-reduced, post-tonic - static unsigned short vowel_replace[N_VOWELS_RU][4] = { - {'&', 'V', 'I', 'V'}, // a - {'&', 'V', 'I', 'V'}, // A - {'8', 'V', 'I', 'V'}, // o - {'e', 'I', 'I', 'I'}, // E - {'i', 'I', 'I', 'I'}, // i - {'u'+('"'<<8), 'U', 'U', 'U'}, // u - {'y', 'Y', 'Y', 'Y'}}; // y - - prev = phoneme_tab[phlist[index-1].phcode]; - next = phoneme_tab[phlist[index+1].phcode]; - -if(prev->mnemonic == 'j') - return(0); - - // lookup the vowel name to get an index into the vowel_replace[] table - for(vowelix=0; vowelix<N_VOWELS_RU; vowelix++) - { - if(vowels_ru[vowelix] == ph->mnemonic) - break; - } - if(vowelix == N_VOWELS_RU) - return(0); - - // do we need a variant of this vowel, depending on the stress and adjacent phonemes ? - variant = -1; - if(ch->flags & 2) - { - // a stressed vowel - if((prev->phflags & phPALATAL) && ((next->phflags & phPALATAL) || phoneme_tab[phlist[index+2].phcode]->mnemonic == ';')) - { - // between two palatal consonants, use the soft variant - variant = 0; - } - } - else - { - // an unstressed vowel - if(prev->phflags & phPALATAL) - { - variant = 2; // unstressed soft - } - else - if((ph->mnemonic == 'o') && ((prev->phflags & phPLACE) == phPLACE_pla)) - { - variant = 2; // unstressed soft ([o] vowel following: ш ж - } - else - if(ch->flags & 4) - { - variant = 3; // post tonic - } - else - { - variant = 1; // unstressed - } - } - if(variant >= 0) - { - phlist[index].phcode = PhonemeCode(vowel_replace[vowelix][variant]); - } - } - return(0); -} #endif |