[1.45.21]

-x command-line option. Fixed possible buffer overflow when writing phoneme output.

Slavic languages, changed behaviour of [v] during voiced/unvoiced propagation.

When looking for a word boundary when breaking long clauses, allow for greater word size (Southern Indian languages have big words 3-byte UTF-8 character codes).

$combine.  Don't combine with the nerxt word if it is translated using a different language.


git-svn-id: https://espeak.svn.sourceforge.net/svnroot/espeak/trunk@262 d46cf337-b52f-0410-862d-fd96e6ae7743

dictsource/de_rules

      m) a (lt_     A:l
    ker) am (ik     'A:m
      @) an (_      'A:n
-     @) ance (_    'A~s
+     s) ance (_    'A~s
+     s) ancen (_   'A~s@n
     fr) ance (_    'A~s@
    ein) an (der    'an
      @) ant (_     'ant
      _) natur (C@P5  nA:t'u:r
      _) na (tür    n%A:
         nce (_     _^_EN
+        nces (_    _^_EN
      _) ne (b      ne:
         nebeneina (nd n,e:b@n_|aIn'a
      _) neben (@P5 n'e:b@n
 
         schwer     Sve:r
      _) sechs (P5  z'Eks
+        seku (nden ze:k'U
      _) selbst (@@P6 z'Elpst
         service (_S7 z%Ervi:s
         shop       _^_EN
      _) w (lad     v
      _) w (rA      v
 
-     @) wert (_    vErt
+     @) wert (_    ve:rt
      _) wh         _^_EN
      &) ware (_    _^_EN
      _) wasch (@P5 v'aS

dictsource/dict_phonemes

 z    Z    z.   
 
 
-Dictionary bg_dict  2011-04-20
+Dictionary bg_dict  2011-04-25
 
 @    a    e    i    o    u    
 

dictsource/en_rules

             prede (ces   pri:dI
          _) pre (diP3    pr%i
          _) pre (dic     pr%I
-         _) pre (emp     pr%I
+         _) pre (em      pr%i:
          _) pre (fec     pri:
          _) pre (feren   prE
             pre (fix     pri:

dictsource/ta_rules

      _) க் (ரௌண்ட	g
      _) க்வ (A		kw
      _) க் (வாலியர	g
+     _) கி (ராம்ப	k
+     _) க (ஞ்சா		gV
 //endsort
 
 //sort
 என்சை) க்லோப (ீடியா 	klo:p
 	க் (ஸ		k
   _மெ) க்ஸிக (ோ	ksik
+  _ராஜ்) க (ிரண		k
+ இராம) க (ாவிய		k
+ _மெரு) க (ூட்ட		g
 //endsort
 
 
      _) த (ோஹா		d
      _) தௌலதாப (ாத	daUlVd.a:b
      _) த (்ராட்சை	d
+     _) த (வக்கால	dV
+     _) த (ருமங்		dV
+     _) த (ருமப்		dV
 //endsort
 
 //sort
      த) த (்ரூப		t
     பா) த (்ரூம		t
   _மை) த் (ரேய		t
+   L03) த்த (ின		dd
+   L03) த்த (ிசை		dd
+   L03) த்த (ூச		dd
+   L03) த்த (ேவன	dd
+   L03) த்த (ேவர		dd
+   L03) த்த (ேவத		dd
 //endsort
 
 
      _) ப் (ளவுஸ	b
      _) ப் (ளோரிடா	f
      _) ப் (ளௌஸ	b
+     _) பு (ரோகித	p
+     _) ப (ீன்ஸ		b
+     _) ப (ிஜிலி	b
+     _) ப (ேமானி	b
 //endsort
 
 //sort
   ஏரோ) ப் (ளேன		p
      _) ப் (ளோரன்ஸ	f
 	ப் (ஸ		p
+   _அ) ப (ின்_		b
+ _குமார) ப (ாளைய	p
 //endsort
 
 	// If ப+ appearing in the middle of a word is followed by ட+

phsource/compile_report

                 [l/] fr
 l/l_@           [l/3] base
                 [l/] fr
-l/l@            [l#] base
-                [”¦] base
+l/l@            [hös] base
+                [l#] base
                 [l] fr
                 [l/2] fr
                 [K] nso
 l/L2_uL         [l/2] base
 l/l_3           [l/] de
 l/l_4           [ll] sq
-l/la            [l#] base
-                [”¦] base
+l/la            [hös] base
+                [l#] base
                 [l] fr
                 [l/2] fr
                 [K] nso
                 [K] tn
 l/l_a           [l/3] base
                 [l/] fr
-l/le            [l#] base
-                [”¦] base
+l/le            [hös] base
+                [l#] base
                 [l] fr
                 [l/2] fr
                 [K] nso
                 [&:] af
 l/l_front       [L] sq
 l/l_front_      [l/4] sq
-l/li            [l#] base
-                [”¦] base
+l/li            [hös] base
+                [l#] base
                 [l] fr
                 [l/2] fr
                 [l] zh
 ll/_ll          [L] bg
 l/l_long        [l] base
                 [l] fr
-l/lo            [l#] base
-                [”¦] base
+l/lo            [hös] base
+                [l#] base
                 [l/2] fr
                 [K] nso
                 [K] tn
 l^/l_rfx        [l.] base
                 [l] ru
                 [l^] ru
-l/lu            [l#] base
-                [”¦] base
+l/lu            [hös] base
+                [l#] base
                 [l] fr
                 [l/2] fr
                 [K] nso

src/dictionary.cpp

 /***************************************************************************
- *   Copyright (C) 2005 to 2010 by Jonathan Duddington                     *
+ *   Copyright (C) 2005 to 2011 by Jonathan Duddington                     *
  *   email: [email protected]                                    *
  *                                                                         *
  *   This program is free software; you can redistribute it and/or modify  *
 };
 
 
-static void WritePhMnemonic(char *phon_out, int *ix, PHONEME_TAB *ph, PHONEME_LIST *plist)
+static void WritePhMnemonic(char **buf, PHONEME_TAB *ph, PHONEME_LIST *plist)
 {//=======================================================================================
 	int c;
 	int mnem;
 	int len;
 	int first;
+	int ix = 0;
+	char *phon_out;
 	unsigned int ipa_control=0;  // first byte of ipa string may control the phoneme name interpretation. 0x20 = ignore this phoneme
 	PHONEME_DATA phdata;
 
+	phon_out = *buf;
+
 	if(option_phonemes == 3)
 	{
 		// has an ipa name been defined for this phoneme ?
 		{
 			if((ipa_control = phdata.ipa_string[0]) > 0x20)
 			{
-				strcpy(&phon_out[*ix], phdata.ipa_string);
-				*ix += len;
+				strcpy(&phon_out[ix], phdata.ipa_string);
+				ix += len;
 			}
 			if(ipa_control >= 0x20)
+			{
+				*buf += ix;
 				return;  // 0x20 = ignore phoneme
+			}
 		}
 	}
 
 			if((c >= 0x20) && (c < 128))
 				c = ipa1[c-0x20];
 
-			*ix += utf8_out(c, &phon_out[*ix]);
+			ix += utf8_out(c, &phon_out[ix]);
 		}
 		else
 		{
-			phon_out[(*ix)++]= c;
+			phon_out[ix++]= c;
 		}
 		first = 0;
 	}
+
+	*buf += ix;
 }  // end of WritePhMnemonic
 
 
 */
 
 	int  ix;
+	int  len;
+	int  max_len;
 	int  phon_out_ix=0;
 	int  stress;
 	unsigned int c;
 	char *p;
+	char *buf;
+	char phon_buf[30];
 	PHONEME_LIST *plist;
 	
 	static const char *stress_chars = "==,,''";
 
 	if(phon_out != NULL)
 	{
-		for(ix=1; ix<(n_phoneme_list-2) && (phon_out_ix < (n_phon_out - 6)); ix++)
+		for(ix=1; ix<(n_phoneme_list-2); ix++)
 		{
+			buf = phon_buf;
+
 			plist = &phoneme_list[ix];
 			if(plist->newword)
-				phon_out[phon_out_ix++] = ' ';
+				*buf++ = ' ';
 
 			if(plist->synthflags & SFLAG_SYLLABLE)
 			{
 
 					if(c != 0)
 					{
-						phon_out_ix += utf8_out(c, &phon_out[phon_out_ix]);
+						buf += utf8_out(c, buf);
 					}
 				}
 			}
 				// the tone_ph field contains a phoneme table number
 				p = phoneme_tab_list[plist->tone_ph].name;
 
-				sprintf(&phon_out[phon_out_ix], "(%s)", p);
-				phon_out_ix += (strlen(p) + 2);
+				sprintf(buf, "(%s)", p);
+				buf += (strlen(p) + 2);
 			}
 			else
 			{
-				WritePhMnemonic(phon_out, &phon_out_ix, plist->ph, plist);
+				WritePhMnemonic(&buf, plist->ph, plist);
 	
 				if(plist->synthflags & SFLAG_LENGTHEN)
 				{
-					WritePhMnemonic(phon_out, &phon_out_ix, phoneme_tab[phonLENGTHEN], NULL);
+					WritePhMnemonic(&buf, phoneme_tab[phonLENGTHEN], NULL);
 				}
 				if((plist->synthflags & SFLAG_SYLLABLE) && (plist->type != phVOWEL))
 				{
 					// syllablic consonant
-					WritePhMnemonic(phon_out, &phon_out_ix, phoneme_tab[phonSYLLABIC], NULL);
+					WritePhMnemonic(&buf, phoneme_tab[phonSYLLABIC], NULL);
 				}
 				if(plist->tone_ph > 0)
 				{
-					WritePhMnemonic(phon_out, &phon_out_ix, phoneme_tab[plist->tone_ph], NULL);
+					WritePhMnemonic(&buf, phoneme_tab[plist->tone_ph], NULL);
 				}
 			}
-		}
 	
-		if(phon_out_ix >= n_phon_out)
-			phon_out_ix = n_phon_out - 1;
+			len = buf - phon_buf;
+			max_len = (n_phon_out - phon_out_ix - 5);   // allow for " ..." and zero byte terminator
+			if(len > max_len)
+			{
+				strcpy(&phon_buf[max_len], " ...");
+				len = max_len + 4;
+			}
+			phon_buf[len] = 0;
+			strcpy(&phon_out[phon_out_ix], phon_buf);
+			phon_out_ix += len;
+
+			if(len > max_len)
+			{
+				break;
+			}
+		}
 		phon_out[phon_out_ix] = 0;
 	}
 }  // end of GetTranslatedPhonemeString

src/phonemelist.cpp

 /***************************************************************************
- *   Copyright (C) 2005 to 2007 by Jonathan Duddington                     *
+ *   Copyright (C) 2005 to 2011 by Jonathan Duddington                     *
  *   email: [email protected]                                    *
  *                                                                         *
  *   This program is free software; you can redistribute it and/or modify  *
 		// set consonant clusters to all voiced or all unvoiced
 		// Regressive
 		int type;
-		int word_end_devoice = 0;
+		int stop_propagation = 0;
 		voicing = 0;
 
 		for(j=n_ph_list2-1; j>=0; j--)
 
 			if(regression & 0x2)
 			{
-				// LANG=Russian, [v] amd [v;] don't cause regression, or [R^]
+				// [v] amd [v;] don't cause regression, or [R^]
 				if((ph->mnemonic == 'v') || (ph->mnemonic == ((';'<<8)+'v')) || ((ph->mnemonic & 0xff)== 'R'))
 				{
-					if(word_end_devoice == 1)
-						voicing = 0;
-					else
-						type = phLIQUID;
+					stop_propagation = 1;
 				}
 			}
 
 					voicing = 0;
 				}
 			}
+			if(stop_propagation)
+			{
+				voicing = 0;
+				stop_propagation = 0;
+			}
 
-			word_end_devoice = 0;
 			if(plist2[j].sourceix)
 			{
 				if(regression & 0x04)
 					if(voicing == 0)
 					{
 						voicing = 1;
-						word_end_devoice = 1;
 					}
 				}
 			}

src/readclause.cpp

 		}
 		*charix_top = ix;
 
-		if(((ix > (n_buf-20)) && !IsAlpha(c1) && !iswdigit(c1))  ||  (ix >= (n_buf-2)))
+		if(((ix > (n_buf-75)) && !IsAlpha(c1) && !iswdigit(c1))  ||  (ix >= (n_buf-4)))
 		{
 			// clause too long, getting near end of buffer, so break here
 			// try to break at a word boundary (unless we actually reach the end of buffer).
+			// (n_buf-4) is to allow for 3 bytes of multibyte character plus terminator.
 			buf[ix] = ' ';
 			buf[ix+1] = 0;
 			UngetC(c2);

src/synthdata.cpp

 #include "translate.h"
 #include "wave.h"
 
-const char *version_string = "1.45.04  25.Apr.11";
+const char *version_string = "1.45.21  27.Apr.11";
 const int version_phdata  = 0x014500;
 
 int option_device_number = -1;

src/translate.cpp

 unsigned int embedded_list[N_EMBEDDED_LIST];
 
 // the source text of a single clause (UTF8 bytes)
-#define N_TR_SOURCE    700
+#define N_TR_SOURCE    800
 static char source[N_TR_SOURCE+40];     // extra space for embedded command & voice change info at end
 
 int n_replace_phonemes;
 				strcpy(ph_buf,word_phonemes);
 
 				flags2[0] = TranslateWord(translator, p2+1, 0, wtab+1);
-				if(flags2[0] & FLAG_WAS_UNPRONOUNCABLE)
+				if((flags2[0] & FLAG_WAS_UNPRONOUNCABLE) || (word_phonemes[0] == phonSWITCH))
 					ok = 0;
 
 				if(sylimit & 0x100)

src/translate.h

 #define LOPT_PREFIXES        3
 
  // non-zero, change voiced/unoiced to match last consonant in a cluster
- // bit 1=LANG=ru,  don't propagate over [v]
+ // bit 1=LANG=cz,bg  don't propagate over [v]
  // bit 2=don't propagate acress word boundaries
  // bit 3=LANG=pl,  propagate over liquids and nasals
  // bit 4=devoice word-final consonants
 	int transpose_min;
 	char dictionary_name[40];
 
-	char phon_out[300];
+	char phon_out[400];
 	char phonemes_repeat[20];
 	int  phonemes_repeat_count;
 	int  phoneme_tab_ix;