mjohnson/espeak-ng
1
0
Fork 0
mirror of https://github.com/rhasspy/espeak-ng.git synced 2026-11-01 05:46:19 +00:00
Code Issues Packages Projects Releases Wiki Activity

[1.42.03 release]

Browse Source 
Allow .txt release for *_rules and *_list files.
Rules lookup, consider multi-byte characters when calculating matchign points.
Fix bug in Roman Number translation.


git-svn-id: https://espeak.svn.sourceforge.net/svnroot/espeak/trunk@232 d46cf337-b52f-0410-862d-fd96e6ae7743
This commit is contained in:
jonsd
2009-12-08 17:59:06 +00:00
parent e404310b8f
commit de52d78665
19 changed files with 1362 additions and 1018 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/dictionary.cpp
+35 -6
View File
@@ -1648,7 +1648,7 @@ static char *DecodeRule(const char *group_chars, int group_length, char *rule)
static char output[60];
static char symbols[] = {' ',' ',' ',' ',' ',' ',' ',' ',' ',
'@','&','%','+','#','S','D','Z','A','L',' ',' ',' ',' ',' ','N','K','V',' ','T','X','?','W'};
'@','&','%','+','#','S','D','Z','A','L','!',' ','?','?','J','N','K','V','?','T','X','?','W'};
static char symbols_lg[] = {'A','B','C','H','F','G','Y'};
@@ -1998,7 +1998,8 @@ static void MatchRule(Translator *tr, char *word[], int group_length, char *rule
if((letter == rb) || ((letter==(unsigned char)REPLACED_E) && (rb=='e')))
{
add_points = 21;
if((letter & 0xc0) != 0x80)
add_points = 21; // don't add point for non-initial UTF-8 bytes
consumed++;
}
else
@@ -2148,6 +2149,27 @@ static void MatchRule(Translator *tr, char *word[], int group_length, char *rule
}
break;
case RULE_SKIPCHARS:
{
// Used for lang=Tamil, used to match on the next word after an unknown word ending
// only look until the end of the word (including the end-of-word marker)
// Jx means 'skip characters until x', where 'x' may be '_' for 'end of word'
char *p = post_ptr + letter_xbytes;
char *p2 = p;
int rule_w; // skip characters until this
utf8_in(&rule_w,rule);
while((letter_w != rule_w) && (letter_w != RULE_SPACE))
{
p2 = p;
p += utf8_in(&letter_w,p);
}
if(letter_w == rule_w)
{
post_ptr = p2;
}
}
break;
case RULE_INC_SCORE:
add_points = 20; // force an increase in points
break;
@@ -2180,10 +2202,11 @@ static void MatchRule(Translator *tr, char *word[], int group_length, char *rule
default:
if(letter == rb)
{
if(letter == RULE_SPACE)
add_points = (21-distance_right);
else
if((letter & 0xc0) != 0x80)
{
// not for non-initial UTF-8 bytes
add_points = (21-distance_right);
}
}
else
failed = 1;
@@ -2348,7 +2371,13 @@ static void MatchRule(Translator *tr, char *word[], int group_length, char *rule
if(letter == RULE_SPACE)
add_points = 4;
else
add_points = (21-distance_left);
{
if((letter & 0xc0) != 0x80)
{
// not for non-initial UTF-8 bytes
add_points = (21-distance_left);
}
}
}
else
failed = 1;