mjohnson/espeak-ng
1
0
Fork 0
mirror of https://github.com/rhasspy/espeak-ng.git synced 2026-11-04 06:32:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

[1.45.11]

Browse Source 
lang=Turkish, fix pronunciation of suffixes, fix suffixes after names.

Windows: command option --stdout caused spurious 0x0d bytes to be included in speech output (CR/LF problem).

Fixed crash which could occur when a clause starts with a hyphen.

espeakedit: New input format for Bulgarian lexicon.

lang=de: Fix important pronunciation error (editing error) for "sage".


git-svn-id: https://espeak.svn.sourceforge.net/svnroot/espeak/trunk@259 d46cf337-b52f-0410-862d-fd96e6ae7743
This commit is contained in:
jonsd
2011-03-31 16:10:36 +00:00
parent 478fde1a20
commit 82b01c0336
29 changed files with 507 additions and 161 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/translate.cpp
+36 -12
View File
@@ -386,6 +386,9 @@ int IsAlpha(unsigned int c)
return(1);
if(lookupwchar(extra_indic_alphas, c) != 0)
return(1);
if((c >= 0xd7a) && (c <= 0xd7f))
return(1); // malaytalam chillu characters
return(0);
}
@@ -577,6 +580,19 @@ char *strchr_w(const char *s, int c)
}
int IsAllUpper(const char *word)
{//=============================
int c;
while((*word != 0) && !isspace2(*word))
{
word += utf8_in(&c, word);
if(!iswupper(c))
return(0);
}
return(1);
}
static char *SpeakIndividualLetters(Translator *tr, char *word, char *phonemes, int spell_word)
{//============================================================================================
int posn = 0;
@@ -1022,7 +1038,7 @@ if((wmark > 0) && (wmark < 8))
length = 999;
wordx = word1;
while(((length < 3) && (length > 0))|| (word_length > 1 && Unpronouncable(tr,wordx)))
while(((length < 3) && (length > 0))|| (word_length > 1 && Unpronouncable(tr, wordx, posn)))
{
// This word looks "unpronouncable", so speak letters individually until we
// find a remainder that we can pronounce.
@@ -1176,20 +1192,26 @@ if((wmark > 0) && (wmark < 8))
if(prefix_type & SUFX_B)
{
// SUFX_B is used for Turkish, tr_rules contains "(Pb£
// retranslate the prefix part
// SUFX_B is used for Turkish, tr_rules contains " ' (Pb"
// examine the prefix part
char *wordpf;
char prefix_phonemes2[12];
strncpy0(prefix_phonemes2,end_phonemes,sizeof(prefix_phonemes2));
wordpf = &prefix_chars[1];
found = LookupDictList(tr, &wordpf, phonemes, dictionary_flags, SUFX_P, wtab); // without prefix
if(found == 0)
strcpy(prefix_phonemes, phonemes);
// look for stress marker or $abbrev
found = LookupDictList(tr, &wordpf, phonemes, dictionary_flags, 0, wtab);
if(found)
{
end_type = TranslateRules(tr, wordpf, phonemes, N_WORD_PHONEMES, end_phonemes, 0, dictionary_flags);
sprintf(prefix_phonemes,"%s%s%s",phonemes,end_phonemes,prefix_phonemes2);
strcpy(prefix_phonemes, phonemes);
}
if(dictionary_flags[0] & FLAG_ABBREV)
{
prefix_phonemes[0] = 0;
SpeakIndividualLetters(tr, wordpf, prefix_phonemes, 1);
}
prefix_flags = 1;
}
else
{
@@ -2772,7 +2794,8 @@ if((c == '/') && (tr->langopts.testing & 2) && IsDigit09(next_in) && IsAlpha(pre
{
// '-' between two letters is a hyphen, treat as a space
word_flags |= FLAG_HYPHEN;
words[word_count-1].flags |= FLAG_HYPHEN_AFTER;
if(word_count > 0)
words[word_count-1].flags |= FLAG_HYPHEN_AFTER;
c = ' ';
}
}
@@ -2809,7 +2832,7 @@ if((c == '/') && (tr->langopts.testing & 2) && IsDigit09(next_in) && IsAlpha(pre
space_inserted = 1;
}
else
if(!(words[word_count-1].flags & FLAG_NOSPACE) && IsAlpha(prev_in))
if((word_count > 0) && !(words[word_count-1].flags & FLAG_NOSPACE) && IsAlpha(prev_in))
{
// dot after a word, with space following, probably an abbreviation
words[word_count-1].flags |= FLAG_HAS_DOT;
@@ -2930,7 +2953,7 @@ if((c == '/') && (tr->langopts.testing & 2) && IsDigit09(next_in) && IsAlpha(pre
// end of 'word'
sbuf[ix++] = ' ';
if((ix > words[word_count].start) && (word_count < N_CLAUSE_WORDS-1))
if((word_count < N_CLAUSE_WORDS-1) && (ix > words[word_count].start))
{
if(embedded_count > 0)
{
@@ -3056,7 +3079,8 @@ if((c == '/') && (tr->langopts.testing & 2) && IsDigit09(next_in) && IsAlpha(pre
*pn++ = *pw++;
}
else
if((*pw == tr->langopts.thousands_sep) && (pw[1] == ' ') && iswdigit(pw[2]))
if((*pw == tr->langopts.thousands_sep) && (pw[1] == ' ')
&& iswdigit(pw[2]) && (pw[3] != ' ') && (pw[4] != ' ')) // don't allow only 1 or 2 digits in the final part
{
pw += 2;
ix++; // skip "word"