[1.26.05] Fix count of upper-case letters in a word to consider multibyte characters.

Mbrola translations, option to not match across a word boundary.
Mbrola translations, allow check of lengthen indicator [:].
espeakedit, add operation to apply a voice's formants modifications to a keyframe.


git-svn-id: https://espeak.svn.sourceforge.net/svnroot/espeak/trunk@47 d46cf337-b52f-0410-862d-fd96e6ae7743

dictsource/dict_phonemes

 
 *    :    b    c    C    d    dZ   f    
 g    h    j    J    k    l    m    n    
-N    n^   p    r    R    s    S    s2   
-t    tS   ts   v    z    Z    z2   
+N    n^   p    r    R    R2   s    S    
+s2   t    tS   ts   v    z    Z    z2   
 
 
 Dictionary it_dict
 m    m#   n    N    n#   N#   n^   n^#  
 p    Q    r    R    r#   R2   s    S    
 t    T    tl#  v    x    z    
+
+
+Dictionary la_dict
+
+@-   a    a:   aI   aU   E    e:   EU   
+I    i:   O    o:   OI   U    u:   y    
+y:   
+
+*    :    <h>  b    d    f    g    h    
+j    k    l    m    n    N    p    r    
+R    s    t    w    z    
+
+
+Dictionary sr_dict
+
+&    @    @-   @2   a    A    a:   aI   
+aU   E    e    e:   E~   i    I    i:   
+l-   o    o:   oU   r-   u    U    u:   
+y    
+
+*    b    d    dZ   dZ;  f    g    h    
+j    k    l    l^   m    n    N    n^   
+p    r    R    s    S    t    tS   ts   
+tS;  v    x    z    Z    

dictsource/hu_list

 mm      milime:tER
 
 a.m	An^n^i||m'int  $dot
+dr	doktoR  $dot
 gpu	$abbrev
+id	idY:SEb:  $dot
 ill	illEdvE  $dot
+ifj	ifjAb:  $dot
 mta	$abbrev
 pl	pe:lda:ul  $dot
 stb	SAtYb:i  $dot
+ú.n	u:JnEvEzEt:
 vö	v'EZd||_'Yss2E
 
+ii	kEttY:  $abbrev
+iii	ha:Rom  $abbrev
+iv	ne:J  $abbrev
+vi	hAt  $abbrev
+vii	he:t  $abbrev
+viii	n^olts  $abbrev
 
 
 // Articles
 ra	$u
 re	$u
 ik	$u
+as	$u
 es	$u
 os	$u
 ós	$u
 
 
 // Names
-ricárd		RiCArd
+Engels		ENgEls
+Kéthly		ke:thli
+ricárd		RiCARd
 richard		RiCARd
-
+Szamuelly	sAmuElli
+Szamuely	sAmuElli

dictsource/hu_rules

         cc       ts:
         cs       tS
         ccs      tS:
+        cz       ts
 
   anar) ch (i    C
 hierar) ch (i    C
 .group k
         k        k
         kk       k:
+     _) köz      kYz
 
 .group l
         l        l
 .group m
         m        m
 
-    ko) mm (unis  m
+    ko) mm (uni  m
 
 
 .group n
         qu       kv
 
 .group r
-        r        R
-     A) r (A     *
-     C) r (A     R
-        r (C     R
-        rr       R*
+        r        R2
+     A) r (A     R2
+     C) r (A     R2
+        r (C     R2
+        rr       R2*
 
 .group s
         s        S
 
    hel) s (inki  z
     ki) s (ebb   SS
+_munká) s        S
     _o) slo      slo:
     _o) sló      slo:
 

dictsource/it_list

 _>   kj,uza'angolo
 _?   p'untointeRogat'ivo
 _@   ki'otS:ola
-_[   ap'eRtakw'ad@-*a
+_[   ap'eRtakw2'ad@-*a
 _\   bakslaS
-_]   kj,uzakw'ad@-*a
+_]   kj,uzakw2'ad@-*a
 _^   kap:utS:o
 __   'undeRskoRe
 _`   bakp@-*aIm
 _1	'uno
 _2	d'ue
 _3	tR'e
-_4	kw'at:Ro
-_5	tS'inkwe
+_4	kw2'at:Ro
+_5	tS'inkw2e
 _6	s'ej
 _7	s'Et:e
 _8	'Ot:o
 _11     'unditSi
 _12     d'oditSi
 _13     t@-*'editSi
-_14     kwat:'ORditSi
-_15     kw'imditSi
+_14     kw2at:'ORditSi
+_15     kw2'imditSi
 _16     s'editSi
 _17     ditSas:'Et:e
 _18     ditS'Ot:o
 _19     ditSann'ove
 _2X     v'ente
 _3X     t@-*'enta
-_4X     kwaR'anta
-_5X     tSinkw'anta
+_4X     kw2aR'anta
+_5X     tSinkw2'anta
 _6X     sess'anta
 _7X     set:'anta
 _8X     ot'anta
 // If a letter has a "word" pronunciation which is different from its
 // "letter" name, then include the letter name here, with the letter
 // prefixed by a _ character.
+_a	a:
+a	a:  $atend
 _e	e:
 e       e:  $atend
 _i	i:

dictsource/it_rules

 
 .group q
 	_)	q(_	ku:
-		q	kw
-		qu	kw
+		q	kw2
+		qu	kw2
 
 .group r
 	_)	r(_	ER*e

phsource/compile_report

-35 phoneme tables
+37 phoneme tables
           new total
     base  96   96
    base2  24  114
       hu  24  114
       nl  26  117
       pl  15  107
-      sk  26  124
-      cs   5  124
-      hr  25  133
+      sk  29  125
+      cs   5  125
+      hr  19  129
+      sr   2  129
       ru  36  124
       it  12  114
+      la  21  114
       es   6  114
       pt  28  131
    pt_pt  20  131
   1  l^/j2u         base
   1  l^/_l^         base
   1  l^/l^          base
-  7  l/l            base en af sk ru pt ro
- 10  l/_l           base en af de fr sk ru pt ro
+  8  l/l            base en af sk ru la pt ro
+ 11  l/_l           base en af de fr sk ru la pt ro
   1  l/l_           base
   2  l/l@           base
   1  l/L1_aL        base
   2  l/le           base
   1  l/L_eL_af      af
   2  l/li           base
-  3  l/l_long       base fr ro
+  4  l/l_long       base fr la ro
   2  l/lo           base
   2  l/lu           base
-  8  l/tl           base en af fr sk ru pt ro
+  9  l/tl           base en af fr sk ru la pt ro
   6  l/xl           base
   2  m/_m           base ro
   2  m/m_           base ro
   2  ufric/ch       base de
   3  ufric/f        base de ro
   2  ufric/f_       base ro
-  4  ufric/h_       base fi hi
-  5  ufric/h@       base fi hi
-  4  ufric/ha       base fi hi
-  4  ufric/he       base fi hi
-  4  ufric/hi       base fi hi
-  3  ufric/ho       base hi
-  3  ufric/hu       base hi
+  5  ufric/h_       base fi hi la
+  6  ufric/h@       base fi hi la
+  5  ufric/ha       base fi hi la
+  5  ufric/he       base fi hi la
+  5  ufric/hi       base fi hi la
+  4  ufric/ho       base hi la
+  4  ufric/hu       base hi la
   2  ufric/hu_fi    fi
   2  ufric/ll       base is
   3  ufric/s        base es pt
   1  ufric/x_hr     hr
   1  ustop/c        base
   5  ustop/k        base en fr hi sw
-  9  ustop/k_       base base2 en fi fr hi hu el sw
- 12  ustop/ki       base base2 en af fi fr hi hu is sw
- 10  ustop/kl       base base2 en fi fr hi hu el sw
- 11  ustop/kr       base base2 en fi fr hi hu el sw
-  9  ustop/k_unasp  base base2 fi hi hu el
+ 10  ustop/k_       base base2 en fi fr hi hu sk el sw
+ 13  ustop/ki       base base2 en af fi fr hi hu sk is sw
+ 11  ustop/kl       base base2 en fi fr hi hu sk el sw
+ 12  ustop/kr       base base2 en fi fr hi hu sk el sw
+ 10  ustop/k_unasp  base base2 fi hi hu sk el
   2  ustop/p        base hi
   2  ustop/p_       base hi
   1  ustop/percus10 base
-  8  ustop/pl       base base2 fi hi hu hr ro
-  8  ustop/pr       base base2 fi hi hu ro
-  7  ustop/p_unasp  base base2 fi hi hu hr ro
-  6  ustop/p_unasp_ base2 fi hi hu hr ro
+  9  ustop/pl       base base2 fi hi hu sk la ro
+  9  ustop/pr       base base2 fi hi hu la ro
+  8  ustop/p_unasp  base base2 fi hi hu sk la ro
+  7  ustop/p_unasp_ base2 fi hi hu sk la ro
   3  ustop/t        base en hi
   6  ustop/t_       base en hi sk el
  27  ustop/t_dnt    base base2 en fi fr hi hu pl hr ru ro vi ++
   2  vdiph/aai_2    af cy
   1  vdiph/aai_3    fi
   1  vdiph/aau      fi
-  3  vdiph/aau_2    de nl zh_yue
+  4  vdiph/aau_2    de nl la zh_yue
   1  vdiph/aau_3    hi
   1  vdiph/aau_4    vi
   1  vdiph/ae       fr_ca
   1  vdiph/ae_2     en_n
-  6  vdiph/ai       base2 eo hr it pt vi
+  7  vdiph/ai       base2 eo sk hr it pt vi
   3  vdiph/ai_2     en_us cy
   1  vdiph/ai_3     no
   1  vdiph/ai_4     af
-  1  vdiph/ai_5     de
+  2  vdiph/ai_5     de la
   1  vdiph/ai_6     en_rp
   5  vdiph/ai_7     en_sc is
   1  vdiph/ai_en    en
   3  vdiph/eei_3    en_us en_rp sk
   1  vdiph/eei_4    nl
   3  vdiph/eeu      pt vi zh_yue
-  1  vdiph/eeu_2    pt_pt
+  2  vdiph/eeu_2    la pt_pt
   2  vdiph/eeu_3    en_n en_wm
   1  vdiph/eey      fi
-  6  vdiph/ei       base2 nl pt is vi zh_yue
-  1  vdiph/ei_2     hr
+  7  vdiph/ei       base2 nl la pt is vi zh_yue
   8  vdiph/eu       base2 en cy eo nl sk pt vi
   1  vdiph/eu_2     fi
   2  vdiph/&i       fi hi
   3  vdiph/@i_3     en_wm ro zh_yue
   1  vdiph/&i_3     vi
   1  vdiph/@i_4     vi
-  2  vdiph/ii       hr ro
+  1  vdiph/ii       ro
   1  vdiph/i#i      ro
-  5  vdiph/oi       base2 en_wm de it vi
+  6  vdiph/oi       base2 en_wm de it la vi
   1  vdiph/oi_2     af
  10  vdiph/ooi      en en_n en_us en_rp cy eo fi no zh_yue
   1  vdiph/ooi_2    af
-  2  vdiph/ooi_3    en_wm hr
+  1  vdiph/ooi_3    en_wm
   2  vdiph/ou       fi zh_yue
   2  vdiph/ou_2     sk ro
   2  vdiph/ou_3     is
   1  vdiph/u-i      vi
   1  vdiph/ui_2     af
   2  vdiph/ui_3     cy
-  1  vdiph/ui_4     hr
   1  vdiph/Vi       vi
   1  vdiph/Vu       af
   1  vdiph/Vu_2     en_wm
   1  vowel/3_2      en_n
   1  vowel/3_3      en_rp
   3  vowel/3_en     en en_wm af
-  2  vowel/@_4      nl hr
+  1  vowel/@_4      nl
   2  vowel/8_2      en_us sv
   1  vowel/8_3      zh_yue
- 11  vowel/a        en_n cy de hu nl pl sk hr
+ 13  vowel/a        en_n cy de hu nl pl sk hr sr la
   5  vowel/a#       base2 en_sc it pt
-  7  vowel/a_2      base2 eo it pt pt_pt ro vi
+  8  vowel/a_2      base2 eo it la pt pt_pt ro vi
   4  vowel/a#_2     hr sv is sw
   6  vowel/a_3      en_sc cs is
  12  vowel/a#_3     en en_n en_us en_wm de hi ru pt_pt vi zh_yue
   1  vowel/aa_7     nl
   4  vowel/a_en     en fr
   1  vowel/@_bck    hi
- 14  vowel/e        base2 en en_n af cy eo fr hu hr it pt pt_pt ++
+ 15  vowel/e        base2 en en_n af cy eo fr hu hr it la pt ++
   2  vowel/e#       en_sc
   6  vowel/e_2      en_sc de hi sv no
   1  vowel/e_3      hu
   1  vowel/e_5      en_sc
   3  vowel/ee       fr pl sv
   2  vowel/e_e      en_sc is
-  7  vowel/ee_1     base2 en en_n pl sv no zh_yue
+  8  vowel/ee_1     base2 en en_n pl la sv no zh_yue
   3  vowel/ee_2     en cy nl
   1  vowel/ee#_2    sv
   3  vowel/ee_3     af pt vi
   1  vowel/@_hgh    no
  19  vowel/i        base2 en_n en_rp en_wm cy eo fr hu pl it pt pt_pt ++
   1  vowel/i#       cy
-  3  vowel/i_2      de nl sv
+  4  vowel/i_2      de nl la sv
   2  vowel/i_3      af sk
   3  vowel/i_4      fi hu is
   5  vowel/i_5      en_sc
   2  vowel/i_fnt    hi
  10  vowel/ii       en_wm cy de fr_ca nl is zh_yue
   3  vowel/ii#      cy pl ru
+  1  vowel/ii_2     la
   7  vowel/ii_3     en_rp fi hi sk hr sv no
   2  vowel/ii_4     en_rp
   6  vowel/ii_en    en en_n
   5  vowel/@_low    hi ro no
- 10  vowel/o        base2 en en_wm de fr hi it pt_pt sv
+ 11  vowel/o        base2 en en_wm de fr hi it la pt_pt sv
   4  vowel/o_2      cy hi hu no
   2  vowel/o-_2     en_n en_wm
   2  vowel/o_3      en_sc
   1  vowel/oe_2     no
   1  vowel/oe_4     sv
   2  vowel/o_mid    fr hu
-  8  vowel/oo       en_sc de eo es el sv no zh_yue
+  9  vowel/oo       en_sc de eo la es el sv no zh_yue
  11  vowel/oo_1     en_n en_rp en_wm af fi sk hr vi
   3  vowel/oo_2     en_sc cy cs
   1  vowel/oo_3     af
-  2  vowel/oo_4     hi pl
+  3  vowel/oo_4     hi pl it
   1  vowel/oo_5     is
   6  vowel/oo_en    en en_n en_rp
   2  vowelr/aa_r    en_sc
   4  vowelr/r-voc   hi sk hr
   2  vowelr/V3_r    en en_sc
   1  vowelr/V_r     en
-  7  vowel/u        en_n cy de eo fr cs
+  8  vowel/u        en_n cy de eo fr cs la
   5  vowel/u#       en en_sc
   4  vowel/u_2      fi sk hr ro
   1  vowel/u#_2     sv
   1  vowel/u_7      vi
  16  vowel/u_bck    base2 cy fi fr_ca hi hu nl pl sk hr it pt ++
   2  vowel/uu       en en_wm
-  2  vowel/uu_2     base2 de
+  3  vowel/uu_2     base2 de la
   1  vowel/uu_3     af
   2  vowel/uu_4     fi sv
   7  vowel/uu_bck   fr_ca hi hu pt no zh_yue
   1  vowel/yy       fr_ca
   1  vowel/yy_2     no
   1  vowel/yy_3     sv
-  3  vowel/yy_4     de hu is
+  5  vowel/yy_4     de hu la is
   1  vwl_af/@       af
   1  vwl_af/I       af
   2  vwl_af/r@      af
   1  w/wa           base
   1  w/we           base
   3  w/wi           base vi zh_yue
- 10  w/wi2          en en_n en_us en_sc en_rp en_wm de nl sv no
+ 11  w/wi2          en en_n en_us en_sc en_rp en_wm de nl la sv no
   1  w/wo           base
   1  w/wu           base
   5  w/xw           base

phsource/ph_croatian

   length 140
 endphoneme
 
-phoneme @3    //  Schwa (raised)
-  vowel   starttype (@)  endtype (@)
-  formants vowel/@_4
-  length 140
-endphoneme
-
 
 phoneme a
   vowel starttype (a) endtype (a)
   reduceto U  4
 endphoneme
 
+
 phoneme U
   vowel starttype (u) endtype (u)
   length 140
-//  formants vowel/uu_bck
   formants vowel/u_2
 endphoneme
 
 endphoneme
 
 
-phoneme eI
-  vowel starttype (e) endtype (i)
-  length 250
-  formants vdiph/ei_2
-  linkout ;
-endphoneme
-
-
-phoneme iI
-  vowel starttype (i) endtype (i)
-  length 250
-  formants vdiph/ii
-  linkout ;
-endphoneme
-
-
-phoneme oI
-  vowel starttype (o) endtype (i)
-  length 250
-  formants vdiph/ooi_3
-  linkout ;
-endphoneme
-
-
-phoneme uI
-  vowel starttype (u) endtype (i)
-  length 250
-  formants vdiph/ui_4
-  linkout ;
-endphoneme
-
-
-phoneme p
-  vls blb stop
-  vowelin  f1=0  f2=1000 -50 -100  f3=-200 80
-  vowelout f1=0  f2=1000 -500 -350  f3=-300 80 rms=30
-  lengthmod 2
-  wave ustop/p_unasp%120
-  before _ ustop/p_unasp_%80
-  before l ustop/pl
-  switchvoicing b
-endphoneme
 
 
 phoneme t        // dental variant of /t/

phsource/ph_hungarian

 
 phoneme Y:
   vowel starttype (@) endtype (@)
-  length 240
+  length 270
   formants vowel/y#
 endphoneme
 

phsource/ph_italian

 phoneme O
   vowel starttype (o) endtype (o)
   length 170
-//  formants vowel/oo_4
+  formants vowel/oo_4
 //  formants vowel/0
   reduceto  o  4        // [O] only in stressed syllables
 endphoneme

phsource/ph_slovak

   formants vdiph/ou_2
 endphoneme
 
+
+phoneme aI
+  vowel starttype (a) endtype (i)
+  length 270
+  formants vdiph/ai
+endphoneme
+
 phoneme eI
   vowel starttype (e) endtype (i)
   length 260
   switchvoicing d
 endphoneme
 
+phoneme p
+  vls blb stop
+  vowelin  f1=0  f2=1000 -50 -100  f3=-200 80
+  vowelout f1=0  f2=1000 -500 -350  f3=-300 80 rms=30
+  lengthmod 2
+  wave ustop/p_unasp%120
+  before _ ustop/p_unasp_%80
+  before l ustop/pl
+  switchvoicing b
+endphoneme
+
+
+phoneme k
+  vls vel stop
+  vowelin f1=0  f2=2300 200 400  f3=-100 80
+  vowelout f1=0 f2=2300 300 400  f3=-100 80  rms=20
+  lengthmod 2
+  wave ustop/k_unasp%60   // weaker
+  before _ ustop/k_
+  before r ustop/kr
+  before l ustop/kl
+  before (i) ustop/ki
+  switchvoicing g
+endphoneme
+
 

phsource/phonemes

 phonemetable hr sk
 include ph_croatian
 
+phonemetable sr hr
+include ph_serbian
+
 phonemetable ru base
 include ph_russian
 
 phonemetable it base2
 include ph_italian
 
+phonemetable la base
+include ph_latin
+
 phonemetable es base2
 include ph_spanish
 

src/menus.cpp

 	menu_bar->Append(voice_menu, _T("&Voice"));
 	menu_bar->Append(option_menu, _T("&Options"));
 	menu_bar->Append(tools_menu, _T("&Tools"));
-	menu_bar->Append(data_menu, _T("&Data"));
+	menu_bar->Append(data_menu, _T("&Compile"));
 	menu_bar->Append(help_menu, _T("&Help"));
 
 	return(menu_bar);

src/readclause.cpp

 	int c;
 	int c1;
 	int c2;
+	int cbuf[4];
+	int ix;
 	int n_bytes;
 	unsigned char m;
+	static int ungot2 = 0;
 	static const unsigned char mask[4] = {0xff,0x1f,0x0f,0x07};
 	static const unsigned char mask2[4] = {0,0x80,0x20,0x30};
 
 		return(c1);
 	}
 
-	c1 = GetC_get();
+	if(ungot2 != 0)
+	{
+		c1 = ungot2;
+		ungot2 = 0;
+	}
+	else
+	{
+		c1 = GetC_get();
+	}
+
 	if(option_multibyte == espeakCHARS_WCHAR)
 	{
 		count_characters++;
 		if(((c1 & 0xf8) == 0xf0) && ((c1 & 0x0f) <= 4))
 			n_bytes = 3;
 
-		if(n_bytes > 0)
+		if((ix = n_bytes) > 0)
 		{
-			c = c1 & mask[n_bytes];
-			m = mask2[n_bytes];
-			while(n_bytes > 0)
+			c = c1 & mask[ix];
+			m = mask2[ix];
+			while(ix > 0)
 			{
-				c2 = GetC_get();
-				if(c2 == 0)
+				if((c2 = cbuf[ix] = GetC_get()) == 0)
 				{
 					if(option_multibyte==espeakCHARS_AUTO)
 						option_multibyte=espeakCHARS_8BIT;   // change "auto" option to "no"
 
 				if((c2 & 0xc0) != 0x80)
 				{
+					// This is not UTF8.  Change to 8-bit characterset.
+					if(n_bytes > 1)
+						ungot2 = cbuf[2];
 					GetC_unget(c2);
 					break;
 				}
 				m = 0x80;
 				c = (c << 6) + (c2 & 0x3f);
-				n_bytes--;
+				ix--;
 			}
-			if(n_bytes == 0)
+			if(ix == 0)
 			{
 				count_characters++;
 				return(c);

src/spect.cpp

 }
 
 
+void SpectFrame::ApplyVoiceMods()
+{//==============================
+	// apply the modifications to the formants which are defined in the current voice
+	int pk;
+
+	for(pk=0; pk<N_PEAKS; pk++)
+	{
+		peaks[pk].pkfreq = (peaks[pk].pkfreq * voice->freq2[pk])/256;
+		peaks[pk].pkheight = (peaks[pk].pkheight * voice->height2[pk])/256;
+		peaks[pk].pkwidth = (peaks[pk].pkwidth * voice->width2[pk])/256;
+		peaks[pk].pkright = (peaks[pk].pkright * voice->width2[pk])/256;
+	}
+}
+
+
 
 double SpectFrame::GetRms(int seq_amplitude)
 {//=========================================

src/spect.h

 	void ZeroPeaks();
 	void CopyPeaks(SpectFrame *sf);
 	void ToggleMarker(int n);
+	void ApplyVoiceMods();
 	
 	void MakeWave(int peaks, PitchEnvelope &pitch, int amplitude, int duration);
 	void MakeHtab(int numh, int *htab, int pitch);

src/spectdisplay.cpp

 	int code;
 	wxKeyEvent keyevent;
 	SpectFrame *sf;
-	static int key[] = {0x1044,0x3044,0x104e,0x1047,0x105a,0x1051,0x3051,WXK_F1,WXK_F2,0x1049};
+	static int key[] = {0x1044,0x3044,0x104e,0x1047,0x105a,0x1051,0x3051,WXK_F1,WXK_F2,0x1049,WXK_F10};
 
 	id = event.GetId();
 
 		display = 2;
 		break;
 
+	case WXK_F10:
+		sf->ApplyVoiceMods();
+		display = 2;
+		break;
+
 	case WXK_F2:
 		// make and play sequence
 		ReadDialogValues();
 	menu_spectdisplay->Append(207,_T("Hide Interpolation	CTRL+SHIFT+Q"));
 	menu_spectdisplay->Append(205,_T("Zero Peaks	CTRL+Z"));
 	menu_spectdisplay->Append(210,_T("Toggle Marker"),menu_markers);
+	menu_spectdisplay->Append(211,_T("Apply the formant adjustments from current voice"));
 	menu_spectdisplay->AppendSeparator();
 	menu_spectdisplay->Append(208,_T("Play Selected Frame	F1"));
 	menu_spectdisplay->Append(209,_T("Play All	F2"));

src/synth_mbrola.cpp

 	// bit 0  skip the next phoneme
 	// bit 1  match this and Previous phoneme
 	// bit 2  only at the start of a word
+	// bit 4  don't match two phonemes across a word boundary
 
 	pr = mbrola_tab;
 	while(pr->name != 0)
 			if(pr->next_phoneme == 0)
 				found = 1;
 			else
+			if((pr->next_phoneme == ':') && (plist->synthflags & SFLAG_LENGTHEN))
+			{
+				found = 1;
+			}
+			else
 			{
 				if(pr->control & 2)
 					other_ph = ph_prev;
+				else
+				if((pr->control & 8) && ((plist+1)->newword))
+					other_ph = phoneme_tab[phPAUSE];  // don't match the next phoneme over a word boundary
 				else
 					other_ph = ph_next;
 

src/synthdata.cpp

 #include "translate.h"
 #include "wave.h"
 
-const char *version_string = "1.26.04  13.Jun.07";
+const char *version_string = "1.26.05  14.Jun.07";
 const int version_phdata  = 0x012601;
 
 int option_device_number = -1;

src/tr_languages.cpp

 
 
 	case L('h','r'):   // Croatian
+	case L('s','r'):   // Serbian
 		{
 			static const wchar_t replace_chars_hr[] = 
 				{0x430,0x431,0x446,0x45b,0x447,0x434,0x452,0x45f,
 			tr = new Translator();
 			tr->charset_a0 = charsets[4];   // ISO-8859-4, includes a,e,i,o,u-macron
 			tr->langopts.stress_rule = 2;
+			tr->langopts.stress_flags = 0x21;
+			tr->langopts.unstressed_wd1 = 0;
+			tr->langopts.unstressed_wd2 = 2;
+			tr->langopts.param[LOPT_DIERESES] = 1;
 		}
 		break;
 

src/translate.cpp

 	int dict_flags;        // returned from dictionary lookup
 	int word_flags;        // set here
 	int embedded_count = 0;
+	int letter_count = 0;
 	char *word;
 	char *p;
 	int j, k;
 			{
 				if(!IsAlpha(prev_out) || (langopts.ideographs && (c >= 0x3000)))
 				{
+					letter_count = 0;
+
 					if((prev_out != ' ') && (prev_out != '\''))
 					{
 						// start of word, insert space if not one there already
 					}
 				}
 
+				letter_count++;
+
 				if(iswupper(c))
 				{
 					c = towlower(c);
 						source_index = prev_source_index;  // unget
 					}
 #endif
-					if((all_upper_case) && ((ix - words[word_count].start) > 1))
+					if((all_upper_case) && (letter_count > 2))
 					{
 						if((c == 's') && (next_in==' '))
 						{