[1.41.15]

Fixes to: announce punctuation ( --punct option)


git-svn-id: https://espeak.svn.sourceforge.net/svnroot/espeak/trunk@227 d46cf337-b52f-0410-862d-fd96e6ae7743

dictsource/dict_phonemes

 Dictionary pa_dict
 
 @    a    a~   e    E    e~   E~   i    
-I    i:   i~   I~   o    O    o:   O~   
-o~   r-   U    u    u~   U~   V    V~   
+I    i~   I~   o    O    O~   o~   r-   
+U    u    u~   U~   V    V~   
 
 :    ;    b    bh   c    ch   d    d.   
 dh   dh.  f    g    gh   H    j    J    
 Jh   k    kh   l    l.   m    n    N    
 n.   n^   p    ph   Q    R    s    S    
 t    t.   th   th.  v    x    z    
+
+
+Dictionary prs_dict
+
+@    A    i    o    
+
+b    d    h    l    t    X    

dictsource/en_list

 
 // numeric
 
-_0    zero $text  // TEST
-//_0	z'i@roU
+//_0    zero $text  // TEST
+_0	z'i@roU
 _1	w'0n
 ?6 _1	w'Vn
 _2	t'u:
 ogre             oUg3
 okay             oUk'eI
 oled             oUlEd     // abbrev
-omeg             oUm'eIg@
+omega            oUm'eIg@
 omelet           0ml@t
 omelette         0ml@t
 omnibus           $1
 omniscience      0mn'IsI@ns
 opaque           oUp'eIk
+oped             0p'Ed
 opine            oUp'aIn
 opossum          @p0s@m
 opponent         @poUn@nt

dictsource/fr_list

 
 // 2006-11-18  Gilles Casse  <[email protected]>
 //
-// Updated 2009-09-28 Michel Such  <[email protected]>
+// Updated 2009-10-09 Michel Such  <[email protected]>
 //
 //	* Numbers, a few abbreviations and exceptions.
 //
 afrikaans	afrikans
 bahamas	baamas
 bangladesh	bA~gladES
+bayeux	bajY
 bélarus	belarys
 belize	beliz
 bissau	bisa'o
 (burkina faso)	byrkina||faso
 burundi	burundi
+calvados	calvadoss $text
 croatie	kroasi
 esperanto	E2sperA~to
 groenland	groEnlA~d

dictsource/fr_rules

 
 // 2006-11-18  Gilles Casse  <[email protected]>
 //
-// Updated: 2009-09-25 Michel Such  <[email protected]>
+// Updated: 2009-10-09 Michel Such  <[email protected]>
 //
 
 // Letter classes:
 	ol)	en (t_      A~		// somnolent, violent
 	Vol)	en (t_        		// somnolent, violent
 	Xol)	ent (_      t2  		// volent
+	eul)	ent (_      t2  		// veulent
 	ul)	en (t_      A~		// turbulent
 	Vul)	en (t_
 	imCoC)	en (t_      A~		// impotent
 
 	_)	n (_°       nymero
 	_)	neuro (@@P5       nYro
+	_)	nvda (P4	Envedea
 
 
 .group o

dictsource/hu_list

 _0Z2	sa:zAd     // hundredths
 _0Z3	EzR2Ed
 _0Z4	ti:zEzR2Ed
-
+_0Z5	sa:zEzR2Ed
+_0Z6	miliomod
 
 // ordinal numbers
 
 október		$alt
 november	$alt
 december	$alt
+
+
 //The following 12 row contains shorter hungarian month names
 jan	$alt
 feb $alt

dictsource/hu_rules

 
 .group á
         á        a:
-
+     @) ában (_S4   a:bAn
+     @) ának (_S4   a:nAk
 
 
 .group b
         b        b
 bb b:       
+     @) ban (_S3    bAn
+     @) ben (_S3    bEn
 
         biz (A   b'iz,       // bizalmas  etc.
 	biedermeier  bi:dER2ma:jER2
 dd d:
 dts tS:
 dt t:
+apró) d (sz d
 a) dsz ts:
+beszé) d (szin d
 cselé) d d
 enge) dsz ts:
 engedelmeske) dsz ts:
+fogó) dz (kod ts
 hazu) dsz ts:
+ha) d (járat d
+hor) d (szék d
 mara) dsz  ts:
 tu) dsz ts:
 kére) dz ts
+mentegető) dz z
 //      dd       d:
         dz       dz
      A) dz (A    d:z
 
 .group e
         e        E
+
   D_-_) es (_    %ES
 _) egyezség _!'EJ:ESSe:g
 .group é
         é        e:
         éi (é    e::i:
+     @) ében (_S4   e:bEn
+     @) ének (_S4   e:nEk
 
 .group ë
         ë        Y
      C) gyj      J
      A) gyj (A   J:
 ha) gyj J:
-    _e) gy (es   J:
+e) gy (jelentésű J
+_e) gy (es   J:
     _e) gy (et_  J:
     _e) gy (etlen  J:
     _e) gy (ik   J:
     _e) gy (üt   J:
+e) gyj (elentésű Jj
+gyön) gy (sor J
 _minde) gy (ik   J:
 
     na) gysz (a  Js
 
 .group i
         i        i
+     @) i (_S1   i
+
 _) itthon  ithon
 
 .group í
         í        i:
 íts i:tS:
+fesz) ítsd i:tSd
+bizony) ítsd i:tSd
 
 .group j
         j        j
+     @) jában  (_S5  ja:bAn
+     @) jának  (_S5  ja:nAk
+
 _) javíts  jAvi:tS:
 _) javítsd  jAvi:tSd
 
+
 .group k
         k        k
         kk       k:
 föld) s (ánc S
 s (színű S
    hel) s (inki  z
+
     ki) s (ebb   SS
    _má) s (szó   S
 _munká) s        S
         tty      c:
 apá) ts (ág tS:
 becsüle) t t
+csalódo) tts (ág tS:
 cson) t t
 cson) tj c
 ezüs) t t
 éle) t t
 
 felej) ts tS
+
+gyapo) t t
+
+gráni) t t
+
 já) tsz (ani ts:
 tsz ts:
 szorí) ts (a tS:
 á) t (sző t
 á) t (szú t
 á) t (szű t
+á) t (jut t
+á) t (jö t
 bizo) tts (ág tS:
 állí) ts tS:
 állapo) t (sor t
 köve) ts (ég  tS:
 kür) t t
 kür) tj c
+küldö) ttj c:
 korlátozo) tts tS:
 neve) ts (ég tS:
 on) ts (u tS
+o) tt t:
 szen)t (szék t
 szé) t t
 szöve) ts (ég  tS:
+söté) ts (ég tS:
+sajá) ts (ág tS:
+
 pillan) ts tS
 romlo) tts tS:
 tar) ts tS
 e) z  (t s
 ho) z (ta s
 ho) z (tá s
+helye) z (ked s
 inté) z (ked s
 költö) z (köd s
 

dictsource/ta_list

-//_xx  விழுக்காடு   $text  // TESTING doesn't reduce vowels
 
 // This file is UTF8 encoded 
 // Spelling to phoneme words and exceptions for Tamil
 
 _0M5	a:jirVma:jirVmko:d.i	// 10,000,000,000,000
 _1M5	a:jirVma:jirVmko:d.i
-	
 
 _dpt	_pul.l.i
+
+
+// exceptions
+பாபு	$alt	// initial ப as [b]
+பாலு	$alt
+பவானி	$alt
+பகவதி	$alt
+பானம்	$alt
+பாறம்	$alt
+பாரதி	$alt
+பாரதம்	$alt
+பரதம்	$alt
+பாப	$alt
+பாகம்	$alt
+பண்தம்	$alt
+பவனி	$alt
+
+தானம்	$alt	// initial த as [d]
+தேகம்	$alt
+திரவியம்	$alt
+தூபம்	$alt
+தீபம் 	$alt
+தீபாவளி		$alt
+தீரம்	$alt
+தைரியம்	$alt
+திடம்	$alt
+திரவம்	$alt
+திடீர்	$alt
+திவசம்	$alt
+தியானம்	$alt
+தானியம்	$alt
+தண்டனை	$alt
+துவம்சம்	$alt
+திருஷ்டி	$alt
+தயாநிதி	$alt
+தயாளு	$alt
+துக்கம்	$alt
+திகில்	$alt
+துந்துவி	$alt
+தயவு	$alt
+தடால்	$alt
+திக்	$alt
+தடக்	$alt
+தீக்ஷை	$alt
+
+
+
+கக	$alt	// initial க as [g]
+கனம்	$alt
+கணம்	$alt
+கணபதி	$alt
+கணீர்	$alt
+கிலி	$alt
+கில்லி	$alt
+கோதுமை	$alt
+கோபுரம் 	$alt
+காந்தம்	$alt
+காந்தி	$alt
+குண்டு 	$alt
+கோலி	$alt
+குகை 	$alt
+கிரி	$alt
+கோமதி	$alt
+
+
+
+
+பப	pVpV

dictsource/ta_rules

      _) எ	;e     // add a short [j] sound at start of word ?
 
 	ஏ	e:
-     _) ஏ	;e::     // add a short [j] sound at start of word ?
+     _) ஏ	;e:     // add a short [j] sound at start of word ?
 
 	ஐ	aI
 
      ்) க (B	g
      _) க	kV
      _) க (B	k
+     _) க (T	gV	// if word has $alt attribute in ta_list
+     _) க (BT	g	// $alt
 	க்க	kkV
 	க்க (B	kk
+     _) க (ௌ 		g
+	க (ீத		g
+	க (ஜ		gV
+
+     _) க (ம்பீர	gV
+     _) க (ந்தர்வ	gV
+     _) க (ோவிந்த	g
 
 	ங	NV
 	ங (B	N
 
+
 	ச	sV    // ?? [z]
 	ச (B	s
      _) ச	sa
      ஞ்) ச	dZV
      ஞ்) ச (B	dZ
 
+
 	ஜ	dZV
 	ஜ (B	dZ
 
+
 	ஞ	n^V
 	ஞ (B	n^
 
+
 	ட	d.V
 	ட (B	d.
      _) ட	t.V
 	ட்ட	t.t.V
 	ட்ட (B	t.t.
 
+
 	ண	n.V
 	ண (B	n.
 
+
 	த	dV
 	த (B	d
      _) த	tV
      _) த (B	t
+     _) த (T	dV	// if word has $alt attribute in ta_list
+     _) த (BT	d	// if word has $alt
 	த்த	ttV
 	த்த (B	tt
 
+			// exceptions, த as [d]
+     _) த (ன	dV
+     _) த (ர்ம	dV
+     _) த (ுர்	d
+     _) த (ூர	d
+     _) த (ெய்வ	d
+     _)	த (ுஷ்	d
+     _) த (க்ஷிணா	dV
+
 	ந	nV
 	ந (B	n  // dental n
 
+
 	ன	nV
 	ன (B	n  // alveolar n
 
+
 	ப	bV
 	ப (B	b
      _) ப	pV
      _) ப (B	p
+     _) ப (T	bV	// if word has $alt attribute in ta_list
+     _) ப (BT	b	// if word has $alt
 	ப்ப	ppV
 	ப்ப (B	pp
      ட்)	ப	pV
      ஃ) ப	fV
      ஃ) ப (B	f
 
+
 	ம	mV
 	ம (B	m
 
+
 	ய	jV
 	ய (B	j
 
+
 	ர	rV
 	ர (B	r
 
+
 	ற	RV
 	ற (B	R
 	ற் (ற	t.	// RR -> t.R
 
+
 	ல	lV
 	ல (B	l
 
+
 	ள	l.V
 	ள (B	l.
 
+
 	ழ	z.V
 	ழ (B	z.
 
+
 	வ	vV
 	வ (B	v
 
+
 	ஶ	SV
 	ஶ (B	S
 
+
 	ஷ	s.V
 	ஷ (B	s.
 
+
 	ஸ	sV
 	ஸ (B	s
 
+
 	ஹ	hV
 	ஹ (B	h
 
+
 	க்ஷ 	ks.V
 	க்ஷ (B	ks.
 

phsource/compile_report

-61 phoneme tables
+62 phoneme tables
           new total
     base 104  104
    base2  26  125
       eu   6  126
       mn  15  115
       pa  13  153
+     prs   8  111
 
 Data file      Used by
 b/b             [b] base
                 [k] base2
                 [k] en
                 [k] fi
-                [k] fr
                 [kh] hi
                 [k] hu
                 [k] lv
 vdiph/au        [aU] en
                 [aU] af
                 [a:U] vi
+                [aU] prs
 vdiph/au#       [aU] en_sc
 vdiph/au_2      [au] zhy
 vdiph/au_3      [aU] en_rp
                 [EI] vi
                 [eI] hy
                 [eI] ne
+                [eI] prs
 vdiph/eei_2     [eI] eo
                 [ei] fi
                 [ei] lv
                 [e~] hi
                 [e~] pt
                 [e~] bn
+vnasal/ii_n     [I~] hi
 vnasal/i_n      [i~] pt
                 [i~] bn
                 [i:~] bn
-vnasal/i_n2     [I~] hi
-                [i~] hi
+vnasal/i_n2     [i~] hi
 vnasal/m-       [m-] sw
 vnasal/n-       [n-] sw
 vnasal/nn-      [N-] sw
                 [A] zh
 vowel/aa_3      [A:] af
                 [A] sq
+                [A] prs
 vowel/aa_4      [A:] sv
                 [O] vi
 vowel/aa_5      [A:] en_n
                 [E2] id
                 [e] hy
                 [E] bn
+                [E] prs
 vowel/e_mid2    [E] af
                 [E] de
                 [E2] de
                 [i:] kn
                 [i] ne
                 [i] mn
+                [i] prs
 vowel/i#        [y:] cy
 vowel/i_2       [i2] de
                 [i:] de
                 [oU] en_n
                 [oU] en_wi
                 [o:] de
+                [o2] fr
                 [o] hi
                 [o] ta
                 [o] it
                 [o] bn
                 [o] mr
                 [o:] mn
+                [o:] pa
 vowel/o_2       [o:] cy
-                [o2] fr
                 [o:] hi
                 [o:] hu
                 [o:] la
                 [o] vi
                 [o] da
                 [o:] bn
-                [o:] pa
 vowel/o-_2      [V] en_n
                 [V] en_wm
 vowel/o_3       [oU] en_sc
-                [o] fr
                 [o:] ta
 vowel/o-_3      [U] en_rp
 vowel/o-_4      [o] ro
                 [O] da
 vowel/o_6       [O:] nl
 vowel/o_7       [o] ku
+vowel/o_8       [o] fr
 vowel/oe        [W] en
                 [W] af
                 [W] de
                 [o] zhy
                 [O] tr
                 [O] id
+                [o] prs
 vowel/oo_1      [O:] en_n
                 [O:] en_rp
                 [O@] en_rp
                 [u] kn
                 [u:] kn
                 [u] bn
+                [u] prs
 vowel/u#        [u:] en_sc
                 [Y] tr
 vowel/u_2       [u1] fi
 vwl_en_us/or    [o@] en_us
                 [O:] en_sc
 vwl_en_us/ur    [U@] en_us
-vwl_fr/r        [R] fr
-                [r] fr
+vwl_fr/r        [r] fr
                 [r2] fr
 vwl_fr/_r       [R] fr
                 [r] fr
                 [E~] fr
                 [W~] fr
 vwl_fr/@R       [x] pt_pt
+vwl_fr/_r2      [R] fr
 vwl_fr/@R2      [R] fr_ca
 vwl_fr/@R3      [Q2] hy
 vwl_fr/@R4      [Q2] hy

phsource/ph_french

 //====================================================
 //  French
 //====================================================
-// Updated 2009-09-30 Michel Such  <[email protected]>
+// Updated 2009-10-07 Michel Such  <[email protected]>
 
 phoneme (l)  virtual
 // Used for l and l/
 
 phoneme i- // shorter
   vowel starttype (i) endtype (i)
-  length 130
+  length 110
   formants vowel/i_2
   after (l) l/li-15
   after (r) vwl_fr/ri
 phoneme o
   vowel starttype (o) endtype (o)
   length 170
-  formants vowel/o_3
+  formants vowel/o_8
   after (l) l/lo-10
   after (r) vwl_fr/ro
   before l/ DFT+l/l_o
 phoneme o2
   vowel starttype (o) endtype (o)
   length 170
-  formants vowel/o_2
+  formants vowel/o
   after (l) l/lo-10
   after (r) vwl_fr/ro
   before l/ vowel/o_mid2+l/l_o
 
 phoneme wA
   vowel starttype (o) endtype (a)
-  length 230
+  length 220
   formants vwl_fr/w_a
   after (l) l/lo-10
-  after (r) vwl_fr/ro-15
+  after (r) vwl_fr/ro
   before l/ DFT-10+l/l_a
   before j DFT
   before (r) DFT+vwl_fr/xr
   lengthmod 2
   wave ustop/k%40     // reduce strength of noise burst
   before _ ustop/k_
-  before (r) ustop/kr%45
+  before (r) ustop/k%50
   before r/2 ustop/k%60
   before (l) ustop/kl%45
   before (i) ustop/ki%45
 
 phoneme r // between vowels in a word
   liquid uvl starttype (r) endtype (r)
-  vowelin  f1=0  f2=1600 -300 300  f3=-200 80
+  vowelin  f1=0  f2=1600 -300 300  f3=-200 80 len=20
   vowelout f1=2  f2=1600 -300 300  f3=-300 80
   length 100
   lengthmod 7
 
 phoneme R // First letter of a word
   liquid uvl starttype (r) endtype (r)
-  vowelin  f1=0  f2=1600 -300 300  f3=-200 80
+  vowelin  f1=0  f2=1600 -300 300  f3=-200 80 len=20
   vowelout f1=2  f2=1600 -300 300  f3=-300 80
   length 100
   lengthmod 0
   beforenotvowel r/2
-  formants vwl_fr/r
+  formants vwl_fr/_r2
   after _ vwl_fr/_r
   after (a) vwl_fr/_r
   after (e) vwl_fr/_r
 
 phoneme r/2   // variant of [r] when not preceding a vowel
   liquid uvl starttype r/2 endtype r/2
-  vowelout f1=2  f2=1600 -300 300  f3=-300 80
+  vowelout f1=2  f2=1600 -300 300  f3=-300 100
   length 100
   lengthmod 2
   formants vwl_fr/r_+r3/rx%25

phsource/ph_hindi

 phoneme I~
   vowel starttype (i) endtype (i)
   length 170
-  formants vnasal/i_n2
+  formants vnasal/ii_n
 endphoneme
 
 phoneme e~

phsource/phonemes

 
 phonemetable pa hi
 include ph_punjabi
+
+phonemetable prs base
+include ph_dari

src/compiledict.cpp

 	unsigned char bad_phoneme[4];
 static char nullstring[] = {0};
 
-	WORD_TAB winfo;
-
 	comment = NULL;
 	text_not_phonemes = 0;
 	phonetic = word = nullstring;
 		if(word[0] == '_')
 		{
 			// This is a special word, used by eSpeak.  Translate this into phonemes now
-			memset(&winfo,0,sizeof(winfo));
 			strcat(phonetic, " ");     // need a space to indicate word-boundary
 
 	// PROBLEM  vowel reductions are not applied to the translated phonemes
 	// condition rules are not applied
-			TranslateWord(translator,phonetic,0,&winfo);
+			TranslateWord(translator,phonetic,0,NULL);
 			text_not_phonemes = 0;
 			strncpy0(encoded_ph, word_phonemes, N_WORD_BYTES-4);
 

src/dictionary.cpp

 	sprintf(fname,"%s%c%s_dict",path_home,PATHSEP,name);
 	size = GetFileLength(fname);
 
+	if(tr->data_dictlist != NULL)
+	{
+		Free(tr->data_dictlist);
+		tr->data_dictlist = NULL;
+	}
+
 	f = fopen(fname,"rb");
 	if((f == NULL) || (size <= 0))
 	{
 		return(1);
 	}
 
-	if(tr->data_dictlist != NULL)
-		Free(tr->data_dictlist);
-
 	tr->data_dictlist = Alloc(size);
 	fread(tr->data_dictlist,size,1,f);
 	fclose(f);

src/extras.cpp

 	char phonemes[80];
 	char phonemes2[80];
 	char buf_out[120];
-	WORD_TAB winfo;
 	static const char *vowels1 = "aeiou";
 	static const char *vowels2 = "aeou";
 
 			}
 		}
 		// translate
-		memset(&winfo,0,sizeof(winfo));
-		TranslateWord(translator,&word1[1],0,&winfo);
+		TranslateWord(translator,&word1[1],0, NULL);
 		DecodePhonemes(word_phonemes,phonemes);
 
 		stress_posn1 = 0;
 				vowel_ix++;
 		}
 
-		memset(&winfo,0,sizeof(winfo));
-		TranslateWord(translator,&word2[1],0,&winfo);
+		TranslateWord(translator,&word2[1],0, NULL);
 		DecodePhonemes(word_phonemes,phonemes2);
 
 		vowel_ix = 1;
 	char pronounce2[80];
 	char phonemes[80];
 	char phonemes2[80];
-	WORD_TAB winfo;
 
 	static const char *vowels = "aeiouyAEIOUY29@";
 
 		}
 
 		// translate
-		memset(&winfo,0,sizeof(winfo));
-		TranslateWord(translator,&word2[1],0,&winfo);
+		TranslateWord(translator,&word2[1],0, NULL);
 
 		DecodePhonemes2(word_phonemes,phonemes);  // also need to change some phoneme names
 
 	int wlen;
 	int len;
 	int check_root;
-	WORD_TAB winfo;
 
 	char word[80];
 	char word2[80];
 		}
 
 		// translate
-		memset(&winfo,0,sizeof(winfo));
-		TranslateWord(translator, &word2[1],0,&winfo);
+		TranslateWord(translator, &word2[1],0, NULL);
 		DecodePhonemes(word_phonemes,phonemes);
 
 		// find the stress position in the translation

src/numbers.cpp

 	char ph_buf2[50];
 	char suffix[20];
 	char *wordptr;
-	unsigned int dictflags[2];
+	unsigned int dictflags;
 
 	static const char str_pause[2] = {phonPAUSE_NOLINK,0};
 
 		// but not if the next word starts with an upper-case letter
 		ordinal = 2;
 		word[ix] = ' ';
-
 		if(tr->translator_name == L('h','u'))
 		{
 			// lang=hu don't treat dot as ordinal indicator if the next word is a month name ($alt)
-			dictflags[0] = 0;
 			wordptr = &word[ix+2];
-			LookupDictList(tr, &wordptr, ph_buf, dictflags, 0, NULL);
-			if(dictflags[0] & FLAG_ALT_TRANS) // TEST
+			dictflags = TranslateWord(tr, &word[ix+2], 0, NULL);
+			if(dictflags & FLAG_ALT_TRANS)
 				ordinal = 0;
 		}
 	}
 				}
 				break;
 
-			case NUM_DFRACTION_1:   // italian, say "hundredths" is leading zero
+			case NUM_DFRACTION_1:   // italian, say "hundredths" if leading zero
 			case NUM_DFRACTION_5:   // hungarian, always say "tenths" etc.
-				if(decimal_count <= 4)
+				LookupNum3(tr, atoi(&word[n_digits]), ph_buf, 0,0,0);
+				if((word[n_digits]=='0') || (decimal_mode == NUM_DFRACTION_5))
 				{
-					LookupNum3(tr, atoi(&word[n_digits]), ph_buf, 0,0,0);
-					if((word[n_digits]=='0') || (decimal_mode == 0xa000))
-					{
-						// decimal part has leading zeros, so add a "hundredths" or "thousandths" suffix
-						sprintf(string,"_0Z%d",decimal_count);
-						if(Lookup(tr, string, buf1) == 0)
-							break;   // revert to speaking single digits
+					// decimal part has leading zeros, so add a "hundredths" or "thousandths" suffix
+					sprintf(string,"_0Z%d",decimal_count);
+					if(Lookup(tr, string, buf1) == 0)
+						break;   // revert to speaking single digits
 
-						strcat(ph_buf,buf1);
-					}
-					strcat(ph_out,ph_buf);
-					n_digits += decimal_count;
+					strcat(ph_buf,buf1);
 				}
+				strcat(ph_out,ph_buf);
+				n_digits += decimal_count;
 				break;
 
 			case NUM_DFRACTION_3:

src/readclause.cpp

 
 
 
-static int AnnouncePunctuation(Translator *tr, int c1, int c2, char *buf, int bufix)
-{//=================================================================================
+static int AnnouncePunctuation(Translator *tr, int c1, int *c2_ptr, char *output, int *bufix, int end_clause)
+{//==========================================================================================================
 	// announce punctuation names
 	// c1:  the punctuation character
 	// c2:  the following character
 	const char *punctname;
 	int found = 0;
 	int soundicon;
-	char *p;
+	int c2;
+	int len;
+	int bufix1;
+	char buf[200];
+	char buf2[80];
+
+	c2 = *c2_ptr;
+	buf[0] = 0;
 
 	if((soundicon = LookupSoundicon(c1)) >= 0)
 	{
 		// add an embedded command to play the soundicon
-		sprintf(&buf[bufix],"\001%dI ",soundicon);
+		sprintf(buf,"\001%dI ",soundicon);
 		UngetC(c2);
 		found = 1;
 	}
 	if((punctname = LookupCharName(tr, c1)) != NULL)
 	{
 		found = 1;
-		if(bufix==0)
+		if((*bufix==0) || (end_clause==0))
 		{
 			punct_count=1;
 			while(c2 == c1)
 				punct_count++;
 				c2 = GetC();
 			}
-			UngetC(c2);
+			*c2_ptr = c2;
+			if(end_clause)
+			{
+				UngetC(c2);
+			}
 
-			p = &buf[bufix];
 			if(punct_count==1)
 			{
-				sprintf(p,"%s %s %s",tone_punct_on,punctname,tone_punct_off);
+				sprintf(buf,"%s %s %s",tone_punct_on,punctname,tone_punct_off);
 			}
 			else
 			if(punct_count < 4)
 			{
-				sprintf(p,"\001+10S%s",tone_punct_on);
+				sprintf(buf,"\001+10S%s",tone_punct_on);
 				while(punct_count-- > 0)
-					sprintf(buf,"%s %s",buf,punctname);
-				sprintf(p,"%s %s\001-10S",buf,tone_punct_off);
+				{
+					sprintf(buf2," %s",punctname);
+					strcat(buf, buf2);
+				}
+				sprintf(buf2," %s\001-10S",tone_punct_off);
+				strcat(buf, buf2);
 			}
 			else
 			{
-				sprintf(p,"%s %s %d %s %s",
+				sprintf(buf,"%s %s %d %s %s",
 						tone_punct_on,punctname,punct_count,punctname,tone_punct_off);
-				return(CLAUSE_COMMA);
 			}
 		}
 		else
 					ssml_ignore_l_angle = c1;  // this was &lt; which was converted to <, don't pick it up again as <
 			}
 			ungot_char2 = c1;
-			buf[bufix] = ' ';
-			buf[bufix+1] = 0;
+			buf[0] = ' ';
+			buf[1] = 0;
 		}
 	}
 
 	if(found == 0)
 		return(-1);
 
+	bufix1 = *bufix;
+	len = strlen(buf);
+	strcpy(&output[*bufix],buf);
+	*bufix += len;
+
+	if(end_clause==0)
+		return(-1);
+
 	if(c1 == '-')
 		return(CLAUSE_NONE);   // no pause
-	if(bufix > 0)
+	if(bufix1 > 0)
 		return(CLAUSE_SHORTCOMMA);
 	if((strchr_w(punct_close,c1) != NULL) && !iswalnum(c2))
 		return(CLAUSE_SHORTFALL+4);
 	int any_alnum = 0;
 	int self_closing;
 	int punct_data;
-	int is_punctuation;
-	int save_c2;
+	int is_end_clause;
+	int announced_punctuation;
 	int stressed_word = 0;
 	const char *p;
 	wchar_t xml_buf[N_XML_BUF+1];
 			linelength = 0;
 		}
 
-		if(option_punctuation && (phoneme_mode==0) && (sayas_mode==0) && iswpunct(c1))
-		{
-			// option is set to explicitly speak punctuation characters
-			// if a list of allowed punctuation has been set up, check whether the character is in it
-			if((option_punctuation == 1) || (wcschr(option_punctlist,c1) != NULL))
-			{
-				if((terminator = AnnouncePunctuation(tr, c1, c2, buf, ix)) >= 0)
-					return(terminator);
-			}
-		}
+		announced_punctuation = 0;
 
-		if((phoneme_mode==0) && (sayas_mode==0) && ((punct = lookupwchar(punct_chars,c1)) != 0))
+		if((phoneme_mode==0) && (sayas_mode==0))
 		{
-			punct_data = punct_attributes[punct];
+			is_end_clause = 0;
 
-			if(punct_data & PUNCT_IN_WORD)
+			if((punct = lookupwchar(punct_chars,c1)) != 0)
 			{
-				// Armenian punctuation inside a word
-				stressed_word = 1;
-				*tone_type = punct_data >> 12 & 0xf;   // override the end-of-sentence type
-				continue;
+				punct_data = punct_attributes[punct];
+	
+				if(punct_data & PUNCT_IN_WORD)
+				{
+					// Armenian punctuation inside a word
+					stressed_word = 1;
+					*tone_type = punct_data >> 12 & 0xf;   // override the end-of-sentence type
+					continue;
+				}
+
+				if((iswspace(c2) || (punct_data & 0x8000) || IsBracket(c2) || (c2=='?') || (c2=='-') || Eof()))
+				{
+					// note: (c2='?') is for when a smart-quote has been replaced by '?'
+					is_end_clause = 1;
+				}
 			}
 
-			if((iswspace(c2) || (punct_data & 0x8000) || IsBracket(c2) || (c2=='?') || (c2=='-') || Eof()))
+			if(option_punctuation && iswpunct(c1))
 			{
-				// note: (c2='?') is for when a smart-quote has been replaced by '?'
-
-				is_punctuation = 1;
+				// option is set to explicitly speak punctuation characters
+				// if a list of allowed punctuation has been set up, check whether the character is in it
+				if((option_punctuation == 1) || (wcschr(option_punctlist,c1) != NULL))
+				{
+					if((terminator = AnnouncePunctuation(tr, c1, &c2, buf, &ix, is_end_clause)) >= 0)
+						return(terminator);
+					announced_punctuation = c1;
+				}
+			}
+	
 
+			if(is_end_clause)
+			{
+				int c_next;
+	
 				if((c1 == '.') && (cprev == '.'))
 				{
 					c1 = 0x2026;
 				}
 	
 				nl_count = 0;
-				save_c2 = c2;
+				c_next = c2;
 
-				if(iswspace(c2))
+				if(iswspace(c_next))
 				{
-					while(!Eof() && iswspace(c2))
+					while(!Eof() && iswspace(c_next))
 					{
-						if(c2 == '\n')
+						if(c_next == '\n')
 							nl_count++;
-						c2 = GetC();   // skip past space(s)
-					}
-					if(!Eof())
-					{
-						UngetC(c2);
+						c_next = GetC();   // skip past space(s)
 					}
 				}
 
 				if((nl_count==0) && (c1 == '.'))
 				{
-//					if(iswdigit(cprev) && (tr->langopts.numbers & NUM_ORDINAL_DOT) && islower(c2))
 					if(iswdigit(cprev) && (tr->langopts.numbers & NUM_ORDINAL_DOT))
 					{
 						// dot after a number indicates an ordinal number
-						is_punctuation = 0;
+						if(islower(c_next) || (c_next == '<'))
+							is_end_clause = 0;      // only if followed by lower-case, (or if there is a XML tag)
 					}
 					else
-					if(iswlower(c2))
+					if(iswlower(c_next))
 					{
 						// next word has no capital letter, this dot is probably from an abbreviation
 						c1 = ' ';
-						is_punctuation = 0;
+						is_end_clause = 0;
 					}
 					if(any_alnum==0)
 					{
 						c1 = ' ';   // no letters or digits yet, so probably not a sentence terminator
-						is_punctuation = 0;
+						is_end_clause = 0;
 					}
 				}
-				c2 = save_c2;
 
-				if(is_punctuation)
+				if(is_end_clause)
 				{
+					UngetC(c_next);
 					buf[ix] = ' ';
 					buf[ix+1] = 0;
 		
 					}
 					return(punct_data);   // only recognise punctuation if followed by a blank or bracket/quote
 				}
+				else
+				{
+					if(!Eof())
+					{
+						if(iswspace(c2))
+							UngetC(c_next);
+					}
+				}
 			}
 		}
 
 		if(speech_parameters[espeakSILENCE]==1)
 			continue;
 
+		if(c1 == announced_punctuation)
+		{
+//			const unsigned short keep_punctuation[] = {
+//			'\'', '-', 0x92, 0xb4, 0x2019, 0x2032, 0 };
+
+			// This character has already been announced, so delete it so that it isn't spoken a second time.
+			// Unless it's a hyphen or apostrophe (which is used by TranslateClause() )
+			if(!IsBracket(c1))
+			{
+				c1 = ' ';
+			}
+		}
+
 		j = ix+1;
 		ix += utf8_out(c1,&buf[ix]);    //	buf[ix++] = c1;
 		if(!iswspace(c1) && !IsBracket(c1))

src/synthdata.cpp

 #include "translate.h"
 #include "wave.h"
 
-const char *version_string = "1.41.13  10.Oct.09";
+const char *version_string = "1.41.15  12.Oct.09";
 const int version_phdata  = 0x014100;
 
 int option_device_number = -1;

src/synthesize.cpp

 //	fprintf(stderr,"FMT%d %3s  %3d-%3d f1=%d  f2=%4d %4d %4d  f3=%4d %3d\n",
 //		which,WordToString(other_ph->mnemonic),len,rms,f1,f2,f2_min,f2_max,f3_adj,f3_amp);
 
-	if(other_ph->mnemonic == '?')
+	if((other_ph != NULL) && (other_ph->mnemonic == '?'))
 		flags |= 8;
 
 	if(which == 1)

src/tr_languages.cpp

 		break;
 
 
+	case L('f','a'):   // Farsi
+		{
+			tr->langopts.param[LOPT_UNPRONOUNCABLE] = 1;   // disable check for unpronouncable words
+		}
+		break;
+
+
 	case L('f','i'):   // Finnish
 		{
 			static const unsigned char stress_amps_fi[8] = {18,16, 22,22, 20,22, 22,22 };

src/translate.cpp

 	int spell_word;
 	int stress_bits;
 	int emphasize_allcaps = 0;
-	int wflags = wtab->flags;
-	int wmark = wtab->wmark;
+	int wflags;
+	int wmark;
+	WORD_TAB wtab_null[8];
 
 	// translate these to get pronunciations of plural 's' suffix (different forms depending on
 	// the preceding letter
 	static char word_iz[4] = {0,'i','z',0};
 	static char word_ss[4] = {0,'s','s',0};
 
+	if(wtab == NULL)
+	{
+		memset(wtab_null, 0, sizeof(wtab_null));
+		wtab = wtab_null;
+	}
+	wflags = wtab->flags;
+	wmark = wtab->wmark;
+
 	dictionary_flags[0] = 0;
 	dictionary_flags[1] = 0;
 	dictionary_flags2[0] = 0;
 	end_phonemes[0] = 0;
 	ph_limit = &phonemes[N_WORD_PHONEMES];
 
+	if(tr->data_dictlist == NULL)
+	{
+		// dictionary is not loaded
+		word_phonemes[0] = 0;
+		return(0);
+	}
+
 	// count the length of the word
 	if(*word1 == ' ') word1++;   // possibly a dot was replaced by space:  $dot
 	wordx = word1;
 					AppendPhonemes(tr,phonemes, N_WORD_PHONEMES, end_phonemes);
 					end_phonemes[0] = 0;
 				}
+				memcpy(wordx,word_copy,strlen(word_copy));
 			}
 			wordx[-1] = c_temp;
 		}

src/translate.h

 #define FLAG_DONT_SWITCH_TRANSLATOR  0x1000
 #define FLAG_SUFFIX_REMOVED  0x2000
 #define FLAG_HYPHEN_AFTER    0x4000
+#define FLAG_ORDINAL       0x8000   // passed to TranslateNumber() to indicate an ordinal number
 
 #define FLAG_NO_TRACE      0x10000   // passed to TranslateRules() to suppress dictionary lookup printout
 #define FLAG_NO_PREFIX     0x20000
-#define FLAG_ORDINAL       0x40000   // passed to TranslateNumber() to indicate an ordinal number
 
 // prefix/suffix flags (bits 8 to 14, bits 16 to 22) don't use 0x8000, 0x800000
 #define SUFX_E        0x0100   // e may have been added