From 084584ee1606808e6f8ffc857959408e1d509539 Mon Sep 17 00:00:00 2001 From: jonsd Date: Wed, 27 Jun 2012 18:02:26 +0000 Subject: [PATCH] [1.46.19] Speak names of characters U+2800 to U+28FF, Braille dot symbols. Language option to speak numbers which use characters '0' to '9' with the English voice. Add some more characters to the vowel and consonant lists for Indian languages. git-svn-id: https://espeak.svn.sourceforge.net/svnroot/espeak/trunk@284 d46cf337-b52f-0410-862d-fd96e6ae7743 --- dictsource/de_list | 1 + dictsource/dict_phonemes | 16 ++++++++-- dictsource/en_list | 1 + dictsource/ta_list | 7 ++++- dictsource/ta_rules | 1 + phsource/compile_report | 15 +++++----- src/dictionary.cpp | 14 +++++---- src/numbers.cpp | 63 +++++++++++++++++++++++++++++----------- src/phoneme.h | 2 +- src/synthdata.cpp | 2 +- src/tr_languages.cpp | 15 +++++++--- src/translate.cpp | 20 ++++++++++--- src/translate.h | 3 ++ 13 files changed, 118 insertions(+), 42 deletions(-) diff --git a/dictsource/de_list b/dictsource/de_list index 58f6f6a3..96405e13 100644 --- a/dictsource/de_list +++ b/dictsource/de_list @@ -665,6 +665,7 @@ numero $1 oboe o:b'o:@ orgie $alt ok o:k'e: +ödem Y:dEm paradies $3 passage pas'aZ@ diff --git a/dictsource/dict_phonemes b/dictsource/dict_phonemes index 4b9c045b..ccb082ad 100644 --- a/dictsource/dict_phonemes +++ b/dictsource/dict_phonemes @@ -102,7 +102,7 @@ p Q r R s s; t T ts v x z -Dictionary en_dict 2012-04-07 +Dictionary en_dict 2012-06-15 0 3 3: @ @- @2 @5 @L a a# A: A@ aa aI aI3 aI@ @@ -202,7 +202,7 @@ q r r. s S s. t T t. th th. v w x z -Dictionary ta_dict 2012-06-06 +Dictionary ta_dict 2012-06-27 a a: aI aU e E e: i I i: o o: u U u2 u: @@ -797,3 +797,15 @@ o u U u# Y k l m n N p q r s S t tS v w X z Z + + +Dictionary ga_dict 2012-06-27 + +0 @ a A: e E e: i +I i: o O o: u U u1 +u: V + +: ; b c C d d[ dZ +f g h j k l m n +p r R s S t t[ tS +v w x X z diff --git a/dictsource/en_list b/dictsource/en_list index 0203bd1f..14e32f42 100644 --- a/dictsource/en_list +++ b/dictsource/en_list @@ -103,6 +103,7 @@ _hok h'Uk _cap k,ap@-t@L _?A l,Et3 _?? sImb@L +_braille br'eIl _#9 tab _#32 speIs //… _::d%0td,0t // for elipsis or ... while reading diff --git a/dictsource/ta_list b/dictsource/ta_list index 41a02f87..8fd805a0 100644 --- a/dictsource/ta_list +++ b/dictsource/ta_list @@ -118,12 +118,17 @@ U+bf9 ru:ba:j (பி . ஏ) bije: $dot (கி . மீ) kilo:mi:t.t.Vr $dot aud ostr'eIlIVn||d'olVz +cny tS'aIni:z||j'uVn +eur ju:ro:z gbp br'itiS||p'aUndz -inr 'IndIVn||r'u:ba:j +inr 'IndIVn||r'u:pi:z +jpy dZ'a:pVni:z||j'en +rub r'VSVn||r'u:bVlz usd j'ues||d'olVz + // numbers _0 suz.ijVm // சுழியம் _1 onRU diff --git a/dictsource/ta_rules b/dictsource/ta_rules index b111c105..549bd881 100644 --- a/dictsource/ta_rules +++ b/dictsource/ta_rules @@ -502,6 +502,7 @@ _மன்மதக்) க (ுகை g //endsort //sort + ஷங்) க (ர kV _கோபி) க kV# _ஸ்டா) க kV# _ஜமாய்) க kV# diff --git a/phsource/compile_report b/phsource/compile_report index 616ff8b2..04a40f15 100644 --- a/phsource/compile_report +++ b/phsource/compile_report @@ -79,7 +79,7 @@ consonants 10 116 bo 10 152 kk 20 117 fa 9 110 - ga 24 124 + ga 25 125 Data file Used by b/b [b] base @@ -407,7 +407,7 @@ l/l_ [l] base [l/] fr l/l_@ [l/3] base [l/] fr -l/l@ [h] base +l/l@ [hq] base [l#] base [l] fr [l/2] fr @@ -441,7 +441,7 @@ l/L2_oL [l/2] base l/L2_uL [l/2] base l/l_3 [l/] de l/l_4 [ll] sq -l/la [h] base +l/la [hq] base [l#] base [l] fr [l/2] fr @@ -449,7 +449,7 @@ l/la [h [K] tn l/l_a [l/3] base [l/] fr -l/le [h] base +l/le [hq] base [l#] base [l] fr [l/2] fr @@ -461,7 +461,7 @@ l/L_eL_af [&] af [&:] af l/l_front [L] sq l/l_front_ [l/4] sq -l/li [h] base +l/li [hq] base [l#] base [l] fr [l/2] fr @@ -475,7 +475,7 @@ ll/ll [L] bg ll/_ll [L] bg l/l_long [l] base [l] fr -l/lo [h] base +l/lo [hq] base [l#] base [l/2] fr [K] nso @@ -485,7 +485,7 @@ l/l_o [l/3] base l^/l_rfx [l.] base [l] ru [l^] ru -l/lu [h] base +l/lu [hq] base [l#] base [l] fr [l/2] fr @@ -2143,6 +2143,7 @@ vowel/u [u] base [u] ak [u:] wo [u] bo + [u1] ga [u:] ga vowel/u# [u:] en-sc [Y] tr diff --git a/src/dictionary.cpp b/src/dictionary.cpp index b4b4d323..1ecf649b 100644 --- a/src/dictionary.cpp +++ b/src/dictionary.cpp @@ -347,8 +347,8 @@ int HashDictionary(const char *string) -char *EncodePhonemes(char *p, char *outptr, unsigned char *bad_phoneme) -/*********************************************************************/ +const char *EncodePhonemes(const char *p, char *outptr, unsigned char *bad_phoneme) +/***************************************************************************/ /* Translate a phoneme string from ascii mnemonics to internal phoneme numbers, from 'p' up to next blank . Returns advanced 'p' @@ -364,7 +364,8 @@ char *EncodePhonemes(char *p, char *outptr, unsigned char *bad_phoneme) int consumed; unsigned int mnemonic_word; - bad_phoneme[0] = 0; + if(bad_phoneme != NULL) + bad_phoneme[0] = 0; // skip initial blanks while(isspace(*p)) @@ -424,8 +425,11 @@ char *EncodePhonemes(char *p, char *outptr, unsigned char *bad_phoneme) if(max_ph == 0) { // not recognised, report and ignore - bad_phoneme[0] = *p; - bad_phoneme[1] = 0; + if(bad_phoneme != NULL) + { + bad_phoneme[0] = *p; + bad_phoneme[1] = 0; + } *outptr++ = 0; return(p+1); } diff --git a/src/numbers.cpp b/src/numbers.cpp index 72df2df3..48ad7c75 100644 --- a/src/numbers.cpp +++ b/src/numbers.cpp @@ -590,12 +590,13 @@ int TranslateLetter(Translator *tr, char *word, char *phonemes, int control) int n_bytes; int letter; int len; + int ix; int save_option_phonemes; char *p2; char *pbuf; char capital[20]; - char ph_buf[60]; - char ph_buf2[60]; + char ph_buf[80]; + char ph_buf2[80]; char hexbuf[6]; ph_buf[0] = 0; @@ -646,22 +647,49 @@ int TranslateLetter(Translator *tr, char *word, char *phonemes, int control) if(ph_buf[0] == 0) { // character name not found - if(iswalpha(letter)) - Lookup(tr, "_?A", ph_buf); - - if((ph_buf[0]==0) && !iswspace(letter)) - Lookup(tr, "_??", ph_buf); - - if(ph_buf[0] != 0) + if((letter >= 0x2800) && (letter <= 0x28ff)) { - // speak the hexadecimal number of the character code - sprintf(hexbuf,"%x",letter); - pbuf = ph_buf; - for(p2 = hexbuf; *p2 != 0; p2++) + // braille dots symbol + Lookup(tr, "_braille", ph_buf); + if(ph_buf[0] == 0) { - pbuf += strlen(pbuf); - *pbuf++ = phonPAUSE_VSHORT; - LookupLetter(tr, *p2, 0, pbuf, 1); + EncodePhonemes("br'e:l", ph_buf, NULL); + } + + if(ph_buf[0] != 0) + { + pbuf = ph_buf + strlen(ph_buf); + for(ix=0; ix<8; ix++) + { + if(letter & (1 << ix)) + { + *pbuf++ = phonPAUSE_VSHORT; + LookupLetter(tr, '1'+ix, 0, pbuf, 1); + pbuf += strlen(pbuf); + } + } + } + } + + if(ph_buf[0]== 0) + { + if(iswalpha(letter)) + Lookup(tr, "_?A", ph_buf); + + if((ph_buf[0]==0) && !iswspace(letter)) + Lookup(tr, "_??", ph_buf); + + if(ph_buf[0] != 0) + { + // speak the hexadecimal number of the character code + sprintf(hexbuf,"%x",letter); + pbuf = ph_buf; + for(p2 = hexbuf; *p2 != 0; p2++) + { + pbuf += strlen(pbuf); + *pbuf++ = phonPAUSE_VSHORT; + LookupLetter(tr, *p2, 0, pbuf, 1); + } } } } @@ -1951,8 +1979,9 @@ int TranslateNumber(Translator *tr, char *word1, char *ph_out, unsigned int *fla return(0); // speak digits individually if(tr->langopts.numbers != 0) + { return(TranslateNumber_1(tr, word1, ph_out, flags, wtab, control)); - + } return(0); } // end of TranslateNumber diff --git a/src/phoneme.h b/src/phoneme.h index d834d4b3..cf8f98c9 100644 --- a/src/phoneme.h +++ b/src/phoneme.h @@ -161,7 +161,7 @@ typedef struct { int LookupPhonemeString(const char *string); int PhonemeCode(unsigned int mnem); -char *EncodePhonemes(char *p, char *outptr, unsigned char *bad_phoneme); +const char *EncodePhonemes(const char *p, char *outptr, unsigned char *bad_phoneme); void DecodePhonemes(const char *inptr, char *outptr); extern const char *WordToString(unsigned int word); diff --git a/src/synthdata.cpp b/src/synthdata.cpp index e4766485..2cbade72 100644 --- a/src/synthdata.cpp +++ b/src/synthdata.cpp @@ -35,7 +35,7 @@ #include "translate.h" #include "wave.h" -const char *version_string = "1.46.17 05.May.12"; +const char *version_string = "1.46.19 27.Jun.12"; const int version_phdata = 0x014600; int option_device_number = -1; diff --git a/src/tr_languages.cpp b/src/tr_languages.cpp index 01c4cfc1..c4eeade4 100644 --- a/src/tr_languages.cpp +++ b/src/tr_languages.cpp @@ -308,22 +308,23 @@ static void SetCyrillicLetters(Translator *tr) 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 SetLetterBitsRange(tr,LETTERGP_A,0x3e,0x4d); // + vowel signs, and virama - SetLetterBitsRange(tr,LETTERGP_A,0x55,0x57); // + vowel signs + SetLetterBits(tr,LETTERGP_A, dev_vowels2); // + extra vowels and vowel signs SetLetterBitsRange(tr,LETTERGP_B,0x3e,0x4d); // vowel signs, and virama - SetLetterBitsRange(tr,LETTERGP_B,0x55,0x57); // + vowel signs + 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 - SetLetterBitsRange(tr,LETTERGP_Y,0x55,0x57); // + 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 @@ -711,6 +712,12 @@ Translator *SelectTranslator(const char *name) { tr->letter_bits_offset = OFFSET_GURMUKHI; } + if(name2 == L('n','e')) + { + tr->langopts.break_numbers = 0x2aaaa8; + tr->langopts.max_digits = 22; + tr->langopts.numbers2 |= NUM2_ENGLISH_NUMERALS; + } SetIndicLetters(tr); } break; diff --git a/src/translate.cpp b/src/translate.cpp index 3ac2de4b..52020c10 100644 --- a/src/translate.cpp +++ b/src/translate.cpp @@ -407,6 +407,9 @@ int IsAlpha(unsigned int c) if((c >= 0x1100) && (c <= 0x11ff)) return(1); //Korean jamo + if((c >= 0x2800) && (c <= 0x28ff)) + return(1); // braille + if((c > 0x3040) && (c <= 0xa700)) return(1); // Chinese/Japanese. Should never get here, but Mac OS 10.4's iswalpha seems to be broken, so just make sure @@ -981,6 +984,13 @@ if((wmark > 0) && (wmark < 8)) if(word_phonemes[0] == phonSWITCH) return(0); + if((tr->langopts.numbers2 & NUM2_ENGLISH_NUMERALS) && !(wtab->flags & FLAG_CHAR_REPLACED)) + { + // for this language, speak English numerals (0-9) with the English voice + sprintf(word_phonemes,"%c",phonSWITCH); + return(0); + } + found = TranslateNumber(tr, word1, phonemes, dictionary_flags, wtab, 0); } @@ -2223,7 +2233,7 @@ static int EmbeddedCommand(unsigned int &source_index) -static int SubstituteChar(Translator *tr, unsigned int c, unsigned int next_in, int *insert) +static int SubstituteChar(Translator *tr, unsigned int c, unsigned int next_in, int *insert, int *wordflags) {//========================================================================================= int ix; unsigned int word; @@ -2285,12 +2295,14 @@ static int SubstituteChar(Translator *tr, unsigned int c, unsigned int next_in, if(upper_case) new_c = towupper(new_c); + + *wordflags |= FLAG_CHAR_REPLACED; return(new_c); } -static int TranslateChar(Translator *tr, char *ptr, int prev_in, unsigned int c, unsigned int next_in, int *insert) +static int TranslateChar(Translator *tr, char *ptr, int prev_in, unsigned int c, unsigned int next_in, int *insert, int *wordflags) {//================================================================================================================ // To allow language specific examination and replacement of characters @@ -2368,7 +2380,7 @@ static int TranslateChar(Translator *tr, char *ptr, int prev_in, unsigned int c, } break; } - return(SubstituteChar(tr,c,next_in,insert)); + return(SubstituteChar(tr, c, next_in, insert, wordflags)); } @@ -2680,7 +2692,7 @@ if((c == '/') && (tr->langopts.testing & 2) && IsDigit09(next_in) && IsAlpha(pre word_flags |= FLAG_COMMA_AFTER; } - c = TranslateChar(tr, &source[source_index], prev_in,c, next_in, &char_inserted); // optional language specific function + c = TranslateChar(tr, &source[source_index], prev_in,c, next_in, &char_inserted, &word_flags); // optional language specific function if(c == 8) continue; // ignore this character diff --git a/src/translate.h b/src/translate.h index 6e310798..e58d13f6 100644 --- a/src/translate.h +++ b/src/translate.h @@ -110,6 +110,7 @@ #define FLAG_MULTIPLE_SPACES 0x40000 // word is preceded by multiple spaces, newline, or tab #define FLAG_INDIVIDUAL_DIGITS 0x80000 // speak number as individual digits #define FLAG_DELETE_WORD 0x100000 // don't speak this word, it has been spoken as part of the previous word +#define FLAG_CHAR_REPLACED 0x200000 // characters have been replaced by .replace in the *_rules #define FLAG_SUFFIX_VOWEL 0x08000000 // remember an initial vowel from the suffix #define FLAG_NO_TRACE 0x10000000 // passed to TranslateRules() to suppress dictionary lookup printout @@ -444,12 +445,14 @@ typedef struct { int numbers; #define NUM2_MULTIPLE_ORDINAL 0x1000 +#define NUM2_ENGLISH_NUMERALS 0x2000 // bits 1-4 use variant form of numbers before thousands,millions,etc. // bit6=(LANG=pl) two forms of plural, M or MA // bit7=(LANG-ru) use MB for 1 thousand, million, etc // bit8=(LANG=cs,sk) two forms of plural, M or MA // bit9=(LANG=rw) say "thousand" and "million" before its number, not after // bit12=(LANG=el,es) use ordinal form of hundreds and tens as well as units + // bit13=(LANG=ne) speak (non-replaced) English numerals in English int numbers2; #define BREAK_THOUSANDS 0x49249248