[1.39.26]

Added languages: ca (Catalan) lv (Latvian)
Numbers: allow for languages which have numbers for 100,000 and (or not) 10,000,000 (eg. hi, sw)
Phoneme [ts] new files in phsource/ustop
*_rules: fix crash for lettergroups numbers greater than 10, and better error reporting.
Languages: eo, jbo: adjust syllable lengths.
Language: en, fix "1st, 2nd, 3rd, 5th".  But a full ordinal number feature is still needed.


git-svn-id: https://espeak.svn.sourceforge.net/svnroot/espeak/trunk@202 d46cf337-b52f-0410-862d-fd96e6ae7743

dictsource/dict_phonemes

 Dictionary eo_dict
 
 @-   a    aI   aU   e    eI   eU   i    
-o    OI   u    uI   
+o    oI   u    uI   
 
 *    b    d    dZ   f    g    h    j    
 k    l    m    n    p    r    R    s    
 
 Dictionary jbo_dict
 
-@    a    aI   aU   e    eI   i    l-   
-o    OI   r-   u    
+@    @-   a    aI   aU   e    eI   i    
+l-   o    oI   r-   u    
 
-b    d    dZ   f    g    h    j    k    
-l    m    n    N    p    R    s    S    
-t    tS   v    w    x    z    Z    
+*    -    b    d    dZ   f    g    h    
+j    k    l    m    n    N    p    R    
+s    S    t    tS   v    w    x    z    
+Z    
 
 
 Dictionary ko_dict
 Dictionary lv_dict
 
 a    a:   ai   au   e    e:   ei   i    
-i:   ie   iu   o    o:   oi   u    u:   
-ui   uo   
+i:   ie   iu   o    o:   u    u:   ui   
+uo   
 
 :    b    c    d    dZ   f    g    h    
 j    J    k    l    l^   m    n    n^   
 :    b    c    ch   d    d.   dZ   f    
 g    h    j    J    k    kh   l    l.   
 m    n    N    n.   n^   p    ph   R    
-R2   s    s.   S;   t    t.   t.h  th   
-tS   v    w    z    
+R2   s    S    s.   S;   t    t.   t.h  
+th   tS   v    w    z    
+
+
+Dictionary om_dict
+
+@#   a    a:   E    e    I    I#   i:   
+O    o:   U    u:   
+
+**   :    a#   b    B    c`   d    d`   
+dZ   f    g    h    j    k    k`   l    
+m    n    n^   p    p`   R    s    S    
+t    t`   tS   v    w    z    
+
+
+Dictionary ca_dict
+
+@    a    a#   aI   e    E    i    o    
+O    u    U    
+
+*    **   :    b    B    d    D    dZ;  
+f    g    j    J    J^   k    l    l^   
+m    n    N    n^   p    Q    r    R    
+R2   s    S;   t    T    tS   ts   tS;  
+v    v#   w    z    Z    Z;   

dictsource/en_list

 ?5 _%	pVRs'Ent
 _&	amp@sand
 _'	kwoUt
-_(	lEftbrakI2t
-_)	raItbrakI2t
+_[	lEftbrakI2t
+_]	raItbrakI2t
+_(	lEftpa2rEn
+_)	raItpa2rEn
 _*	ast@rIsk
 _+	plVs
 _,	k0m@
 _-	h,aIf@n
+?3 _-	daS
 _.	d0t
 _:	koUl@n
 _;	sEmIk'oUl@n
-_<	ElaNg@L
+_<	lEsDan
 _=	i:kw@Lz
-_>	A@aNg@L
+_>	greIt@Dan
 _?	kwEstS@n
 _@	at|saIn
-_[	lEftskwe@
-_]	raItskwe@
 _^	s3:kVmfl,Eks
 ?5 _^	sIRkVmfl,Eks
 __	'Vnd3sk,o@
 ₨	ru:pi:
 ₩	w0n
 §	sEkS@n
-¶	par@graf   // in en_rules
+¶	par@graf
+¶¶	par@grafs
 ¤	kVr@nsIsaIn
 ©	k0pIraIt
 ®	rEdZIst3d
 _dpt	pOInt
 _roman	roUm@n_
 
+(1 st)	f3:st
+(2 nd)	sEk@nd
+(3 rd)	T3:d
+(5 th)	fIT
+
 
 // ABBREVIATIONS
 //**************
 i.e	aIi:_!   $pause
 irc	$abbrev
 lbs	paUndz
+LBS	$abbrev
 ltd	lImI2tI2d
 mc	m@k  $alt2   // combine with the following word
 oem	$abbrev
 xy	$abbrev
 
 
-nd                $only
-rd                $only
-th                $only
-st                $only
+nd	$only
+rd	$only
+th	$only
+st	$only
 
 mr	mIst3     $dot
 mrs	mIsI2z    $dot
 //ms               mIz         $dot
-dr      d0kt3     $dot $capital
-lt                $dot
-prof              $dot
-rev               $dot
-st                $dot
+dr	$dot
+lt	$dot
+prof	$dot
+rev	$dot
+st	$dot
 
 // not abbreviations when all-caps
 all	O:l	$allcaps

dictsource/en_rules

          %) % (%       
         %%) % (_         _::
 
-// extra symbols (mostly in en_list)
-            ¶¶       par@grafs

dictsource/eo_list

 _0M4		duili'ono
 _dpt		komo_
 
+// accent names
+_acu	dekst@-*akoRno
+_brv	hoko
+_cir	tsiRkumflekso
+_dia	dupunktasupeRsigno
+_grv	maldekst@-*akoRno
+_tld	tildo
+
+
+_??	liteRo
 _.	punkto
+_:	dupunkto
+_;	punktokomo
+_,	komo
+_?	demandosigno
+_-	dividost@-*eko
+~	tildo
+°	gradoI
+%	p@-*otsento
+@	po
+&	kaI
+
 
 v	vo	// not Roman numbers
 x	ikso
 
 // misc adverbs etc
 eĉ              $brk
-(pli ol)	pliol
+pli	pl,i
+plej	pl,eI
 
 // pronouns
 mi              $u

dictsource/eo_rules

 .group o
 	_)	o(_	o
 		o	o
-		oj	OI
+		oj	oI
+	-)	oj (_	%oI
+	-)	on (_	%on
 		
 .group p
 	_)	p(_	po
 	\.)     .
 
 
-		. (D	_:k'omo_
 		*	asteRisko
-			pundo
-		 (D	_pundOI_:
+		£	pundo
+		£ (D	_pundoI_:
 	        !	,ekk@-*io,
-		:	dupunkto
 		-	_
 		- (_a	
 		$	dolaRo
-		$ (D	_dolaROI_:
-		%	p@-*otsento
-		@	po
-		&	%kaI
+		$ (D	_dolaRoI_:
 			
 

dictsource/es_list

 _»	kom'iJ^as||dE**'EtSas
 
 ^	TiRkumfl'exo
-€	eU*o
+€	eU**o
 %	pOrTj'Ento
 &	ampERs'ant
 @	a*R'oBa

dictsource/hi_list

 _98       ath:'a:n@we:
 _99       nIn^j'a:n@we:
 
-_0C       s'O:
-_0M1      h@z'a:r
-_0M2      d,Vsla:kh@_    // until we fix numbers > 100,000
-_1M2      d,Vsla:kh@_    // until we fix numbers > 100,000
-_0L       l'a:kh
+_0C	s'O:
+_0M1	h@z'a:r
+_0M2	l'a:kh
+_0M3	cr'ore
+_0M4	b'ilion  // ?? English
 _dpt      _d@s@ml'o:_        // ?? what is Hi for "." ?
 
 // Single consonants

dictsource/jbo_list

 _i	ibu
 j	Z@
 k	k@
-l	l@
-m	m@
-n	n@
+_l	l@
+_m	m@
+_n	n@
 _o	obu
 p	p@
-r	R@
+_r	R@
 s	s@
 t	t@
 _u	ubu
 i	i_:	$u	// sentence break (recognised by eSpeak program). Try a short pause after ".i"
 
 // end-of-clause [_;_] before these
-noi	_;_nOI	$u
-poi	_;_pOI	$u
+noi	_;_noI	$u
+poi	_;_poI	$u
 no'u	_;_nohu	$u
 po'u	_;_pohu	$u
-goi	_;_gOI	$u
+goi	_;_goI	$u
 
 ija	_;_iZa		// should this series be unstressed?
 ijanai	_;_iZanaI
 
 // terminators (include a pause after)
 ku	ku_::	$u
-boi	bOI_::	$u
+boi	boI_::	$u
 vau	vaU_::	$u
 kei	keI_::	$u
-toi	tOI_::	$u
+toi	toI_::	$u

dictsource/jbo_rules

 	a	a
 	ai	aI
 	au	aU
-     X) a (_	%a	// single-syllable cmavo are unstressed
-     X) ai (_	%aI
-     X) au (_	%aU
+     X) a (_	,a	// single-syllable cmavo are unstressed
+     X) ai (_	,aI
+     X) au (_	,aU
 
 .group b
 	b	b
 .group e
 	e	e
 	ei	eI
-     X) e (_	%e
-     X) ei (_	%eI
+     X) e (_	,e
+     X) ei (_	,eI
 
 .group f
 	f	f
 .group i
 	i	i
 	i (A	j
-     X) i (_	%i
-     X) i (A_	%j
+     X) i (_	,i
+     X) i (A_	,j
 
 .group j
 	j	Z
 	k	k
 
 .group l
-	l	l-
+	l	l-  // syllabic
      A) l       l
 	l (A    l
 
 	m	m
 
 .group n
-	n	n
+	n	n-   // syllabic
+     A) n       n
+        n (A    n
 	n (g	N
 	n (k	N
 
 .group o
 	o	o
-	oi	OI
-     X) o (_	%o
-     X) oi (_	%OI
+	oi	oI
+     X) o (_	,o
+     X) oi (_	,oI
 
 .group p
 	p	p
 	q	kw
 
 .group r
-	r	r-
+	r	r-  // syllabic
      A) r	R
+     C) r (A    @-*
 	r (A	R
 
 
 .group u
 	u	u
 	u (A	w
-     X) u (_	%u
-     X) u (A_	%w
+     X) u (_	,u
+     X) u (A_	,w
 
 .group v
 	v	v
 	'	h
      _) ' (_    '@h@
 
-	.	_!
+	.	_!	// dot
      _)	. (@P1	_!      // remove . prefix
 	. (_S1  _!	// remove . suffix
 
 
-	ˈ	'       // U+2c8  stress marker
+	ˈ	'       // U+2c8  stress marker (from syllable capitalisation)

dictsource/sw_list

 _7C     m,ias'aba
 _8C     m,ian'ane
 _9C     m,iat'isa
-_1L    l'akim'oJa
-_2L    l'aki_|mb'ili
-_3L    l'akit'atu
-_4L    l'aki'n-ne
-_5L    l'akit'ano
-_6L    l'akis'ita
-_7L    l'akis'aba
-_8L    l'akin'ane
-_9L    l'akit'isa
 _0M1    'elfu
 _1M1    ,elfum'oJa
 _2M1    ,elfu_|mb'ili
 _7M1    ,elfus'aba
 _8M1    ,elfun'ane
 _9M1    ,elfut'isa
-_0M2    _|mili'oni
+_1M2	l'akim'oJa
+_2M2	l'aki_|mb'ili
+_3M2	l'akit'atu
+_4M2	l'aki'n-ne
+_5M2	l'akit'ano
+_6M2	l'akis'ita
+_7M2	l'akis'aba
+_8M2	l'akin'ane
+_9M2	l'akit'isa
+_0M3	_|mili'oni  // 1,000,000
+_0M4	_|bili'oni
 _dpt    _nukta
 _0and   _|na
 

espeak-data/voices/ca

+name catalan-test
+language ca
+gender male
+

espeak-data/voices/lv

+name latvian
+language lv
+gender male
+
+replace 03  o  o:
+

phsource/compile_report

-58 phoneme tables
+59 phoneme tables
           new total
     base 101  101
    base2  24  120
       hi  52  139
       ta  17  142
       hu  23  117
+      lv  28  123
       nl  28  124
       pl  18  114
       sk  29  132
       la  21  119
       es   9  121
    es_la   1  121
+      ca   9  122
       pt  28  137
    pt_pt  20  137
       ro  36  144
       kn  15  141
       bn  59  147
       ne  23  151
-      lv  28  123
       hy  17  115
-      om  18  118
+      om  21  121
 
 Data file      Used by
 b/b             [b] base
                 [<h>] la
                 [H] bn
                 [h] bn
+                [a#] om
 h/he            [h] base
                 [h] fi
                 [H] hi
                 [l] sk
                 [l] ru
                 [l] la
+                [l] ca
                 [l] pt
                 [l;] ro
 l/_l            [l] base
                 [l] sk
                 [l] ru
                 [l] la
+                [l] ca
                 [l] pt
                 [l;] ro
                 [l] zh
                 [l] sk
                 [l] ru
                 [l] la
+                [l] ca
                 [l] pt
                 [l;] ro
                 [l] sq
 r3/r_trill      [R2] base
                 [R3] base
                 [r] af
+                [R2] lv
                 [r] nl
                 [R] sk
                 [r*] sr
                 [x] pt
-                [R2] lv
 r3/r_trill2     [R] base
                 [r] cy
                 [R] hr
                 [r] af
                 [r] nl
 r3/r_trill.wav  [R2] base
+                [R2] lv
                 [R] sk
                 [r*] sr
                 [x] pt
-                [R2] lv
 r3/r_u          [(u)] base
 r3/r_ulv        [r"] hy
 r3/r_uvl        [r"] hy
                 [z2] fr
                 [z3] fr
                 [s2] hu
+                [s2] lv
                 [s] es
                 [s#] pt
                 [z;] ro
                 [s2] is
-                [s2] lv
 ufric/s!        [s] base
                 [s2] fi
                 [s] fr
                 [s2] hu
+                [s2] lv
                 [s] es
                 [s#] pt
                 [s2] is
-                [s2] lv
 ufric/s_continue  [s2] fi
                 [s2] hu
-                [s2] is
                 [s2] lv
+                [s2] is
 ufric/sh        [S] base
                 [S] fr
                 [s#] pt_pt
 ufric/xx3       [X] base
                 [Q] hy
 ustop/c         [c] base
+ustop/c_ejc     [c`] om
 ustop/k         [k] base
                 [k] en
                 [k] fr
                 [k] fr
                 [kh] hi
                 [k] hu
+                [k] lv
                 [k] sk
                 [k] el
                 [kh] zh
                 [k] sw
                 [k] sq
                 [kh] bn
-                [k] lv
+ustop/k_ejc     [k`] om
 ustop/ki        [c] base
                 [k] base
                 [k] base2
                 [k] fi
                 [k] fr
                 [k] hu
+                [k] lv
                 [k] sk
                 [k^] mk
                 [c] is
                 [c] tr
                 [J] ku
                 [c] ku
-                [k] lv
+ustop/ki_ejc    [k`] om
 ustop/kl        [k] base
                 [k] base2
                 [k] en
                 [k] fr
                 [kh] hi
                 [k] hu
+                [k] lv
                 [k] sk
                 [k] el
                 [k] zhy
                 [k] sw
                 [kh] bn
-                [k] lv
 ustop/kr        [k] base
                 [k] base2
                 [k] en
                 [k] fr
                 [kh] hi
                 [k] hu
+                [k] lv
                 [k] sk
                 [k] el
                 [k] zhy
                 [k] sw
                 [kh] bn
-                [k] lv
 ustop/k_unasp   [k] base
                 [k] base2
                 [k] fi
                 [k] hi
                 [k] hu
+                [k] lv
                 [k] sk
                 [k] el
                 [k] zhy
                 [k`] ko
                 [k] sq
                 [k] bn
-                [k] lv
-                [k`] om
 ustop/k_unasp_  [k] zh
 ustop/null      [?] base
+                [dh] om
 ustop/p         [p] base
                 [p] fr
                 [p2] fr
                 [ph] hy
 ustop/p_asp     [ph] hi
                 [ph] hy
+ustop/p_ejc     [p`] om
 ustop/percus10  [(X1] base
 ustop/pl        [p] base
                 [p] base2
                 [p] hi
                 [ph] hi
                 [p] hu
+                [p] lv
                 [p] sk
                 [p] la
                 [p;] ro
                 [p] sq
                 [p] bn
                 [ph] bn
-                [p] lv
                 [ph] hy
 ustop/pr        [p] base
                 [p] base2
                 [p] fr
                 [p] hi
                 [p] hu
+                [p] lv
                 [p] sk
                 [p] la
                 [p;] ro
                 [p`] ko
                 [p] sq
                 [p] bn
-                [p] lv
-                [p`] om
 ustop/p_unasp_  [p] base2
                 [p] fi
                 [p] hi
                 [p] hu
+                [p] lv
                 [p] sk
                 [p] la
                 [p;] ro
                 [p] sq
                 [p] bn
-                [p] lv
 ustop/q         [q] base
 ustop/q_u       [q] base
 ustop/t         [t] base
                 [t3] fr
                 [t.] hi
                 [t.h] hi
+                [t] lv
                 [t] sk
                 [t] el
                 [th] zh
                 [t] sq
                 [t.] bn
                 [t.h] bn
-                [t] lv
 ustop/t_dnt     [t] base
                 [t[] base
                 [t] base2
                 [t] sw
                 [t`] ko
                 [t] bn
-                [t`] om
 ustop/t_dnt2    [t[] vi
+ustop/t_ejc     [t`] om
 ustop/t_hi      [t] hi
                 [t] bn
                 [th] bn
                 [ts;] ro
                 [z] zhy
                 [tsh] zh
+                [ts] hy
+ustop/ts_       [ts] base2
+                [ts] de
+                [ts] eo
+                [ts] hu
                 [ts] lv
+                [ts] pl
+                [ts] ru
+                [ts;] ro
                 [ts] hy
+ustop/ts2       [ts] lv
 ustop/tsh       [tS] base
                 [c] zhy
 ustop/tsh_      [tS] base
                 [t2] fr
                 [t3] fr
                 [t.] hi
+                [t] lv
                 [t] sk
                 [t] el
                 [t] sq
                 [t.] bn
-                [t] lv
 ustop/tsh_pzd   [tS;] zh
 ustop/tsh_pzd_unasp  [tS;] zh
 ustop/tsh_sr    [tS] sr
                 [c] hu
                 [c`] ko
                 [c] bn
-                [c`] om
 ustop/ts_pzd3   [tS;] hr
                 [dZ;] hr
 ustop/t_sr      [d] sr
 vdiph/8u        [oU] en_us
                 [ou] zh
 vdiph/aae       [aI] en_us
-vdiph/aai       [aai] zhy
+vdiph/aai       [ai] lv
+                [aai] zhy
                 [aI] ne
-                [ai] lv
 vdiph/aai_2     [AI] af
                 [AY] cy
 vdiph/aai_3     [ai] fi
 vdiph/aau       [au] fi
+                [au] lv
 vdiph/aau_2     [aU] en_wi
                 [aU] de
                 [VU] nl
                 [eI] hy
 vdiph/eei_2     [eI] eo
                 [ei] fi
-                [eI] id
                 [ei] lv
+                [eI] id
 vdiph/eei_3     [eI] sk
                 [eI] ku
 vdiph/eeu       [EU] pt
                 [OI] en_n
                 [OI] cy
                 [OY] cy
-                [OI] eo
+                [oI] eo
                 [oi] fi
+                [oi] lv
                 [OI] no
                 [oi] zhy
-                [oi] lv
 vdiph/ooi_2     [OI] af
 vdiph/ooi_3     [OI] en_rp
                 [aI] en_wm
 vdiph/ooi_4     [OI] en_us
 vdiph/oou       [oU] cs
                 [OU] grc
-                [au] lv
 vdiph/ou        [ou] fi
                 [ou] zhy
                 [oU] ne
 vdiph/ui        [uI] base2
                 [uI] eo
                 [ui] fi
+                [ui] lv
                 [uI] vi
                 [ui] zhy
                 [uI] ne
-                [ui] lv
 vdiph/u-i       [yI] vi
 vdiph/ui_2      [uI] af
 vdiph/ui_3      [uI] cy
                 [D] es
 voc/dh_         [D] base
                 [D] es
+voc/dh_om       [d`] om
 voc/j           [J^] base
 voc/Q           [Q] base
                 [Q^] base
 vowel/&         [a] en_rp
                 [&] fi
                 [&:] hi
+                [&] lv
+                [&:] lv
                 [&] sv
                 [&] bn
                 [&:] bn
-                [&] lv
-                [&:] lv
 vowel/0         [0] en
                 [O] hi
                 [O] pt
                 [&/] pt
                 [a#] rw
                 [a/] rw
-                [a] om
 vowel/a_2       [a] base2
                 [a] en_wi
                 [A:] en_wi
                 [A@] en_wi
                 [aa] en_wi
                 [a] eo
+                [a] ca
                 [A] pt
                 [a] pt_pt
                 [a] ro
                 [a/] sw
                 [a/] sq
                 [a] kn
+                [a] om
 vowel/a_3       [a] en_sc
                 [a/] en_sc
                 [A:] en_sc
                 [a] is
                 [A] zh
                 [A] da
-                [a:] om
 vowel/a#_3      [a2] en
                 [a2] en_n
                 [a2] en_us
                 [@] de
                 [a] hi
                 [a#] ru
+                [a#] ca
                 [&] pt_pt
                 [&/] pt_pt
                 [@] vi
                 [a] bn
 vowel/a_4       [a/] el
                 [a] vi
+                [a:] om
 vowel/a_5       [a:] ta
                 [a/] it
                 [a] la
                 [a] sq
                 [a:] ml
                 [a:] kn
+                [A:] om
 vowel/a_6       [a] fr
                 [a2] fr
 vowel/aa        [a] fi
                 [A] fr_ca
+                [a] lv
+                [a:] lv
                 [A] no
                 [A:] no
                 [aa] zhy
-                [a] lv
-                [a:] lv
                 [a] hy
 vowel/aa#       [O] en_us
                 [0] en_wi
                 [e] hr
                 [e/] it
                 [e:] la
+                [e] ca
                 [e] pt
                 [e] pt_pt
                 [e] grc
                 [E3] sv
                 [E] ku
 vowel/e_mid     [E] en_rp
-                [e] jbo
                 [E] fr
                 [E2] fr
                 [E] fr_ca
                 [E] hr
                 [E] it
                 [E] es
+                [E] ca
                 [E] pt_pt
                 [E] no
                 [E] is
                 [E] de
                 [E2] de
                 [E:] de
+                [e] jbo
                 [e] fi
+                [e] lv
+                [e:] lv
                 [E] nl
                 [e] sk
                 [e] es
                 [e] tr
                 [E] ko
                 [e] kn
-                [e] lv
-                [e:] lv
                 [E] om
 vowel/@_fnt     [@] en_wi
                 [@2] en_wi
                 [i] fr
                 [i:] fr
                 [i:] sk
-                [i:] lv
 vowel/i_4       [i] fi
                 [i] hu
                 [I] it
                 [i/] pt_pt
                 [y] ro
 vowel/i_6       [i] ta
+                [i] lv
+                [i:] lv
                 [i] hr
                 [i] tr
                 [i] ku
                 [i] id
                 [i] ml
                 [i] kn
-                [i] lv
 vowel/i#_6      [i.] zh
 vowel/i_7       [i] pl
 vowel/i#_7      [i[] zh
                 [o] it
                 [o/] it
                 [O] la
+                [o] ca
                 [o] pt_pt
                 [o] grc
                 [o:] sv
                 [O] de
                 [o] eo
                 [O] es
+                [O] ca
                 [o] el
                 [O] sv
                 [O] no
                 [O@] en_wm
                 [O:] af
                 [o] fi
+                [o] lv
+                [o:] lv
                 [o] sk
                 [o:] sk
                 [o] hr
                 [o] vi
                 [O/] sq
-                [o] lv
-                [o:] lv
 vowel/oo_2      [0] en_sc
                 [O] cy
                 [o] cs
                 [U:] ta
 vowel/u#_4      [U] en_sc
                 [u:] en_sc
-vowel/u_5       [u] sw
+vowel/u_5       [u] lv
+                [u:] lv
+                [u] sw
 vowel/u_6       [U] pt_pt
                 [u] pt_pt
                 [u] ku
 vowel/uu#       [U] ku
 vowel/uu_2      [U] base2
                 [U] de
+                [U] ca
                 [U] tr
-                [u] lv
-                [u:] lv
 vowel/uu_3      [u] af
                 [y] zh
 vowel/uu_4      [U] fi
                 [y] is
                 [y] hy
 vweak/@         [@#] om
-vweak/a         [a#] om
 vwl_af/@        [@] af
 vwl_af/I        [I] af
 vwl_af/r@       [@] af
 vwl_zh/uo       [uo] zh
 vwl_zh/y&       [y&] zh
 vwl_zh/yee      [yE] zh
+w/iw_           [w/] base
 w/w             [w] base
 w/_w            [w] base
                 [w] zh
 w/xw            [(@)] base
                 [(a)] base
                 [(e)] base
+                [(i)] base
                 [(o)] base
                 [(u)] base
 x/b             [b] base
                 [d] is
                 [d] id
                 [d.] bn
+                [d`] om
 x/d_            [d] base
                 [d[] base
                 [d] base2

phsource/ph_base2

   vowelout f1=0 f2=1700 -300 250  f3=-100 80  rms=20
   lengthmod 2
   wave ustop/ts
+  before _ ustop/ts_
 endphoneme
 
 

phsource/ph_catalan

+
+//====================================================
+//  Catalan  - based on Spanish and Base2
+//====================================================
+
+phoneme a
+  vowel starttype (a) endtype (a)
+  length 180
+  formants vowel/a_2
+  reduceto a#  2
+endphoneme
+
+phoneme a#
+  vowel starttype (@) endtype (@)
+  length 170
+  formants vowel/a#_3
+  unstressed
+endphoneme
+
+phoneme e
+  vowel starttype (e) endtype (e)
+  length 170
+  formants vowel/e
+  reduceto @  2
+endphoneme
+
+phoneme E
+  vowel starttype (e) endtype (e)
+  length 180
+  formants vowel/e_mid
+  reduceto @  2
+endphoneme
+
+phoneme O
+  vowel starttype (o) endtype (o)
+  length 170
+  formants vowel/oo
+  reduceto U  2
+endphoneme
+
+phoneme o
+  vowel starttype (o) endtype (o)
+  length 170
+  formants vowel/o
+  reduceto U  2
+endphoneme
+
+phoneme U
+  vowel starttype (u) endtype (u)
+  length 160
+  formants vowel/uu_2
+  unstressed
+endphoneme
+
+
+
+phoneme l
+  liquid
+  length 100
+  lengthmod 7
+  beforenotvowel l/2       // use 'dark' [l] after a vowel
+  formants l/l
+  after _ l/_l  t l/tl
+endphoneme
+
+
+

phsource/ph_esperanto

 
 phoneme aU
   vowel starttype (a) endtype (u)
-  length 230
+  length 250
   formants vdiph/au_4
 endphoneme
 
 
 phoneme eU
   vowel starttype (e) endtype (u)
-  length 230
+  length 250
   formants vdiph/eu
 endphoneme
 
 
 phoneme aI
   vowel starttype (a) endtype (i)
-  length 240
+  length 250
   formants vdiph/ai
 endphoneme
 
 
 phoneme eI
   vowel starttype (e) endtype (i)
-  length 230
+  length 250
   formants vdiph/eei_2
 endphoneme
 
 
-phoneme OI
+phoneme oI
   vowel starttype (o) endtype (i)
-  length 240
+  length 250
   formants vdiph/ooi
 endphoneme
 
 
 phoneme uI
   vowel starttype (u) endtype (i)
-  length 230
+  length 250
   formants vdiph/ui
 endphoneme
 
   vowelout f1=0 f2=1700 -300 250  f3=-100 80  rms=20
   lengthmod 2
   wave ustop/ts
+  before _ ustop/ts_
 endphoneme
 
 

phsource/ph_german

   vowelout f1=0 f2=1700 -300 250  f3=-100 80  rms=20
   lengthmod 2
   wave ustop/ts
+  before _ ustop/ts_
 endphoneme
 
 

phsource/ph_hungarian

   vowelout f1=0 f2=1700 -300 250  f3=-100 80  rms=20
   lengthmod 2
   wave ustop/ts
+  before _ ustop/ts_
 endphoneme
 
 

phsource/ph_latvian

+
+phoneme @
+  vowel starttype (@) endtype (@)
+  length 120
+  formants vowel/8_7
+endphoneme
+ 
+phoneme i
+  vowel starttype (i) endtype (i)
+  length 110
+  formants vowel/i_6
+  linkout ;
+endphoneme
+
+phoneme i:
+  vowel starttype (i) endtype (i)
+  length 300
+  formants vowel/i_6
+  linkout ;
+endphoneme
+
+
+phoneme e
+  vowel starttype (e) endtype (e)
+  length 120
+  formants vowel/e_mid2
+endphoneme
+
+phoneme e:
+  vowel starttype (e) endtype (e)
+  length 330
+  formants vowel/e_mid2
+endphoneme
+
+
+phoneme &
+  vowel starttype (a) endtype (a)
+  length 120
+  formants vowel/&
+endphoneme
+
+phoneme &:
+  vowel starttype (a) endtype (a)
+  length 330
+  formants vowel/&
+endphoneme
+
+
+phoneme a
+  vowel starttype (a) endtype (a)
+  length 120
+  formants vowel/aa
+endphoneme
+
+phoneme a:
+  vowel starttype (a) endtype (a)
+  length 350
+  formants vowel/aa
+endphoneme
+
+
+phoneme o
+  vowel starttype (o) endtype (o)
+  length 120
+  formants vowel/oo_1
+endphoneme
+
+phoneme o:
+  vowel starttype (o) endtype (o)
+  length 330
+  formants vowel/oo_1
+endphoneme
+
+
+phoneme u
+  vowel starttype (u) endtype (u)
+  length 110
+  formants vowel/u_5
+endphoneme
+
+phoneme u:
+  vowel starttype (u) endtype (u)
+  length 330
+  formants vowel/u_5
+endphoneme
+
+
+phoneme ai
+  vowel starttype (a) endtype (i)
+  length 310
+  long
+  formants vdiph/aai
+endphoneme
+
+phoneme ei
+  vowel starttype (e) endtype (i)
+  length 300
+  long
+  formants vdiph/eei_2
+endphoneme
+
+phoneme au
+  vowel starttype (a) endtype (u)
+  length 300
+  long
+  formants vdiph/aau
+endphoneme
+
+phoneme uo
+  vowel starttype (u) endtype (a)
+  length 300
+  long
+//  formants vdiph2/uo
+  formants vdiph2/uaa
+endphoneme
+
+phoneme oi
+  vowel starttype (o) endtype (i)
+  length 290
+  long
+  formants vdiph/ooi
+endphoneme
+
+phoneme ui
+  vowel starttype (u) endtype (i)
+  length 290
+  long
+  formants vdiph/ui
+endphoneme
+
+phoneme ie
+  vowel starttype (i) endtype (e)
+  length 290
+  long
+  formants vdiph2/ie_2
+endphoneme
+
+phoneme iu
+  vowel starttype (i) endtype (u)
+  length 290
+  long
+  formants vdiph2/iu_2
+endphoneme
+
+
+
+
+phoneme R2   // this is [R] from Slovak/Czech
+  liquid
+  vowelin  f1=0  f2=1700 -300 300  f3=-300 80
+  vowelout f1=2  f2=1700 -300 300  f3=-300 80 brk
+  formants r3/r_trill+r3/r_trill.wav%50
+  trill
+  lengthmod 6
+endphoneme
+
+
+phoneme ts
+  vls alv afr sibilant lengthenstop
+  vowelin f1=0  f2=1700 -300 300  f3=-100 80
+  vowelout f1=0 f2=1700 -300 250  f3=-100 80  rms=20
+  lengthmod 2
+  wave ustop/ts2%90
+  before _ ustop/ts_
+endphoneme
+
+
+phoneme t   // try disable this and try using English [t]
+  vls alv stop lengthenstop
+  vowelin f1=0  f2=1700 -300 300  f3=-100 80
+  vowelout f1=0 f2=1700 -300 250  f3=-100 80  rms=20
+  lengthmod 2
+  wave  ustop/t_short%115
+  before _ ustop/t_
+  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
+  before l^ ustop/ki
+  switchvoicing g
+endphoneme
+
+
+phoneme s2   // second part of long [s:]
+  vls alv frc sibilant
+  vowelin  f1=0  f2=1700 -300 300  f3=-100 80
+  vowelout f1=0  f2=1700 -300 250  f3=-100 80  rms=20
+  lengthmod 3
+  wave ufric/s_continue
+  before _ ufric/s_
+  before p ufric/s!
+  before t ufric/s!
+  before k ufric/s!
+  switchvoicing z
+endphoneme
+

phsource/ph_lojban

 phoneme e
   vowel starttype (e) endtype (e)
   length 170
-  formants vowel/e_mid
+  formants vowel/e_mid2
 endphoneme
 
 

phsource/ph_polish

   vowelout f1=0 f2=1700 -300 250  f3=-100 80  rms=20
   lengthmod 2
   wave ustop/ts
+  before _ ustop/ts_
 endphoneme
 
 

phsource/ph_romanian

   vowelout f1=0 f2=1700 -300 250  f3=-100 80  rms=20 colr=1
   lengthmod 2
   wave ustop/ts
+  before _ ustop/ts_
 endphoneme
 
 

phsource/ph_russian

   vowelout f1=0 f2=1700 -300 250  f3=-100 80  rms=20
   lengthmod 2
   wave ustop/ts
+  before _ ustop/ts_
 endphoneme
 
 

phsource/ph_tamil

 
 phoneme u:
   vowel  starttype (u)  endtype (u)
-  length 250
+  length 260
   formants vowel/u
 endphoneme
 

phsource/phonemes

   formants NULL
   starttype _ endtype _
   length 1
-  lengthmod 1
+  lengthmod 5
 endphoneme
 
 phoneme _:   //   Standard pause
   before N DFT-40+nn/inn
   before ; DFT-40+j2/xj2
   before l DFT-40+l/xl
+  before w DFT-30+w/xw
   before j DFT-30+j/xj
   before r- DFT-60+r/xr
   before r/ DFT+r3/r_i
   vowelout  len=50
   length 100
   formants w/w_
+  after (i) w/iw_
   lengthmod 7
   beforevowel w
 endphoneme
 phonemetable hu base
 include ph_hungarian
 
+phonemetable lv base
+include ph_latvian
+
 phonemetable nl base
 include ph_dutch
 
 phonemetable es_la es
 include ph_spanish_la
 
+phonemetable ca es
+include ph_catalan
+
 phonemetable pt base2
 include ph_pt_brazil
 
 phonemetable ne hi
 include ph_nepali
 
-phonemetable lv base
-include ph_latvian
-
 phonemetable hy base
 include ph_armenian_west
 
 phonemetable om base
 include ph_oromo
+

src/compiledata.cpp

 	{"long",       0x2000000+phLONG},
 	{"brkafter",   0x2000000+phBRKAFTER},
 	{"nonsyllabic",0x2000000+phNONSYLLABIC},
+	{"lengthenstop",0x2000000+phLENGTHENSTOP},
 
 	// voiced / unvoiced
 	{"vcd",	0x2000000+phVOICED},
 	float total;
 	float pkheight;
 	int marker1_set=0;
-	int frame_vowelbreak=NULL;
+	int frame_vowelbreak=0;
 	SpectFrame *fr;
 	wxString path_sep = _T("/");
 

src/compiledict.cpp

 
 int hash_counts[N_HASH_DICT];
 char *hash_chains[N_HASH_DICT];
+char letterGroupsDefined[N_LETTER_GROUPS];
 
 MNEM_TAB mnem_flags[] = {
 	// these in the first group put a value in bits0-3 of dictionary_flags
 	text_not_phonemes = 0;
 	phonetic = word = nullstring;
 
+if(memcmp(linebuf,"_-",2)==0)
+{
+step=1;  // TEST
+}
 	p = linebuf;
 //	while(isspace2(*p)) p++;
 
 			break;
 	
 		case 1:
-			if(c == '-')
+			if((c == '-') && (word[0] != '_'))
 			{
 				flag_codes[n_flag_codes++] = BITNUM_FLAG_HYPHENATED;
 				c = ' ';
 					c = *p++ - '0';
 					value = *p++ - '0';
 					c = c * 10 + value;
-					if((value < 0) || (value > 9) || (c <= 0) || (c >= N_LETTER_GROUPS))
+					if((value < 0) || (value > 9))
 					{
 						c = 0;
-						fprintf(f_log,"%5d: Expected 2 digits after 'L'",linenum);
+						fprintf(f_log,"%5d: Expected 2 digits after 'L'\n",linenum);
+						error_count++;
+					}
+					else
+					if((c <= 0) || (c >= N_LETTER_GROUPS) || (letterGroupsDefined[(int)c] == 0))
+					{
+						fprintf(f_log,"%5d: Letter group L%.2d not defined\n",linenum,c);
 						error_count++;
 					}
 					c += 'A';
 	p = input;
 	if(!isdigit(p[0]) || !isdigit(p[1]))
 	{
+		fprintf(f_log,"%5d: Expected 2 digits after '.L'\n",linenum);
+		error_count++;
 		return(1);
 	}
 
-	group = atoi(&p[1]);
+	group = atoi(&p[0]);
 	if(group >= N_LETTER_GROUPS)
+	{
+		fprintf(f_log,"%5d: lettergroup out of range (01-%.2d)\n",linenum,N_LETTER_GROUPS);
+		error_count++;
 		return(1);
+	}
 
 	while(!isspace2(*p)) p++;
 
 	fputc(RULE_GROUP_START,f_out);
 	fputc(RULE_LETTERGP2,f_out);
 	fputc(group + 'A', f_out);
+	letterGroupsDefined[group] = 1;
 
 	for(;;)
 	{
 
 			if(memcmp(buf,".L",2)==0)
 			{
-				if(compile_lettergroup(&buf[2], f_out) != 0)
-				{
-					fprintf(f_log,"%5d: Bad lettergroup\n",linenum);
-					error_count++;
-				}
+				compile_lettergroup(&buf[2], f_out);
 				continue;
 			}
 
 	char path[sizeof(path_home)+40];       // path_dsource+20
 
 	error_count = 0;
+	memset(letterGroupsDefined,0,sizeof(letterGroupsDefined));
+
 	debug_flag = flags & 1;
 
 	if(dsource == NULL)

src/dictionary.cpp

 	char *w;
 
 	p = letterGroups[group];
+	if(p == NULL)
+		return(0);
 
 	while(*p != RULE_GROUP_END)
 	{
 	int  unstressed_word = 0;
 	char *max_output;
 	int final_ph;
+	int final_ph2;
 	int mnem;
+	int mnem2;
 	int post_tonic;
 	int opt_length;
 	int done;
 	}
 	if(ix == 0) return;
 	final_ph = phonetic[ix-1];
+	final_ph2 =  phonetic[ix-2];
 
 	max_output = output + (N_WORD_PHONEMES-3);   /* check for overrun */
 
 						else
 						{
 							mnem = phoneme_tab[final_ph]->mnemonic;
-							if((mnem != 'n') && (mnem != 's'))
+							mnem2 = phoneme_tab[final_ph2]->mnemonic;
+
+							if((mnem == 's') && (mnem2 == 'n'))
+							{
+								// -ns  stress remains on penultimate syllable
+							}
+							else
+							if(((mnem != 'n') && (mnem != 's')) || (phoneme_tab[final_ph2]->type != phVOWEL))
 							{
 								stressed_syllable = vowel_count - 1;
 							}

src/numbers.cpp

 	int subtract;
 	int repeat = 0;
 	unsigned int flags;
+	char ph_roman[30];
 	char number_chars[N_WORD_BYTES];
 
 	static const char *roman_numbers = "ixcmvld";
 	if(acc > langopts.max_roman)
 		return(0);
 
-	Lookup("_roman",ph_out);   // precede by "roman" if _rom is defined in *_list
-	p = &ph_out[strlen(ph_out)];
+	Lookup("_roman",ph_roman);   // precede by "roman" if _rom is defined in *_list
+	p = &ph_out[0];
+
+	if((langopts.numbers & NUM_ROMAN_AFTER) == 0)
+	{
+		strcpy(ph_out,ph_roman);
+		p = &ph_out[strlen(ph_out)];
+	}
 
 	sprintf(number_chars," %d ",acc);
 	TranslateNumber(&number_chars[1],p,&flags,0);
+
+	if(langopts.numbers & NUM_ROMAN_AFTER)
+		strcat(ph_out,ph_roman);
 	return(1);
 }  // end of TranslateRoman
 
 		{
 			if((thousandplex > 0) && (value < 1000))
 			{
-				if(langopts.numbers2 & 0x100)
-				{
-					if((thousandplex == 1) && (value >= 100))
-					{
-						// special word for 100,000's
-						char ph_buf3[20];
-						sprintf(string,"_%dL",value / 100);
-						if(Lookup(string,ph_buf2) == 0)
-						{
-							LookupNum2(value/100,0,ph_buf2);
-							Lookup("_0L",ph_buf3);
-							strcat(ph_buf2,ph_buf3);
-						}
-						value %= 100;
-						if(value == 0)
-							suppress_null = 1;
-					}
-				}
 				if((suppress_null == 0) && (LookupThousands(value,thousandplex,ph_append)))
 				{
 					// found an exact match for N thousand

src/phoneme.h

 #define phBEFORENOTVOWEL2 0x1000
 #define phSWITCHVOICING  0x0800
 
-#define phNONSYLLABIC 0x100000   // don't count this vowel as a syllable when finding the stress position 
-#define phLONG        0x200000
+#define phNONSYLLABIC  0x100000   // don't count this vowel as a syllable when finding the stress position 
+#define phLONG         0x200000
+#define phLENGTHENSTOP 0x400000  // make the pre-pause slightly longer
 
 // fixed phoneme code numbers, these can be used from the program code
 #define phonCONTROL     1

src/phonemelist.cpp

 			{
 				if((x = (langopts.word_gap & 0x7)) != 0)
 				{
-					insert_ph = pause_phonemes[x];
+					if((x > 1) || ((insert_ph != phonPAUSE_SHORT) && (insert_ph != phonPAUSE_NOLINK)))
+					{
+						// don't reduce the pause
+						insert_ph = pause_phonemes[x];
+					}
 				}
 				if(option_wordgap > 0)
 				{

src/setlengths.cpp

 			if((langopts.word_gap & 0x10) && (p->newword))
 				p->prepause = 60;
 
+			if(p->ph->phflags & phLENGTHENSTOP)
+				p->prepause += 30;
+
 			if(p->synthflags & SFLAG_LENGTHEN)
 				p->prepause += langopts.long_stop;
 			break;
 			}
 
 			// calc length modifier
-			if(next->ph->code == phonPAUSE_VSHORT)
+			if((next->ph->code == phonPAUSE_VSHORT) && (next2->type == phPAUSE))
 			{
-				// ignore very short pause
+				// if PAUSE_VSHORT is followed by a pause, then use that
 				next = next2;
 				next2 = next3;
 				next3 = &phoneme_list[ix+4];

src/synthdata.cpp

 #include "translate.h"
 #include "wave.h"
 
-const char *version_string = "1.39.22  10.Nov.08";
+const char *version_string = "1.39.26  15.Nov.08";
 const int version_phdata  = 0x013900;
 
 int option_device_number = -1;

src/tr_languages.cpp

 
 	case L('e','o'):
 		{
-			static const short stress_lengths_eo[8] = {145, 145,  200, 170,    0,   0,  320, 340};
+			static const short stress_lengths_eo[8] = {145, 145,  230, 170,    0,   0,  360, 370};
 			static const unsigned char stress_amps_eo[] = {16,14, 20,20, 20,24, 24,22 };
 			static const wchar_t eo_char_apostrophe[2] = {'l',0};
 		
 			tr->char_plus_apostrophe = eo_char_apostrophe;
 
 			tr->langopts.word_gap = 1;
+			tr->langopts.vowel_pause = 2;
 			tr->langopts.stress_rule = 2;
 			tr->langopts.stress_flags =  0x6 | 0x10; 
 			tr->langopts.unstressed_wd1 = 3;
 		break;
 
 	case L('e','s'):   // Spanish
+	case L('c','a'):   // Catalan
 		{
-			static const short stress_lengths_es[8] = {170, 200,  180, 180,  0, 0,  220, 250};
+			static const short stress_lengths_es[8] = {180, 210,  190, 190,  0, 0,  230, 260};
+//			static const short stress_lengths_es[8] = {170, 200,  180, 180,  0, 0,  220, 250};
 			static const unsigned char stress_amps_es[8] = {16,12, 18,18, 20,20, 20,20 };    // 'diminished' is used to mark a quieter, final unstressed syllable
+			static const wchar_t ca_punct_within_word[] = {'\'',0xb7,0};   // ca: allow middle-dot within word
 
 			tr = new Translator();
 			SetupTranslator(tr,stress_lengths_es,stress_amps_es);
 			tr->langopts.unstressed_wd2 = 2;
 			tr->langopts.param[LOPT_SONORANT_MIN] = 120;  // limit the shortening of sonorants before short vowels
 
-			tr->langopts.numbers = 0x529 + NUM_ROMAN;
+			tr->langopts.numbers = 0x529 + NUM_ROMAN + NUM_ROMAN_AFTER;
+
+			if(name2 == L('c','a'))
+			{
+				tr->punct_within_word = ca_punct_within_word;
+				tr->langopts.stress_flags = 0x200 | 0x6 | 0x30;  // stress last syllable unless word ends with a vowel
+			}
 		}
 		break;
 
 			tr->langopts.stress_rule = 6;      // stress on last heaviest syllable, excluding final syllable
 			tr->langopts.stress_flags =  0x10004;   // use 'diminished' for unstressed final syllable
 			tr->langopts.numbers = 0x011;
-			tr->langopts.numbers2 = 0x100;
+			tr->langopts.numbers2 = NUM2_100000;
 			tr->letter_bits_offset = OFFSET_DEVANAGARI;
 			SetIndicLetters(tr);
 		}
 
 	case L_jbo:   // Lojban
 		{
-			static const short stress_lengths_jbo[8] = {185,170, 200,200, 0,0, 290,300};
+			static const short stress_lengths_jbo[8] = {145,145, 170,160, 0,0, 330,350};
 			static const wchar_t jbo_punct_within_word[] = {'.',',','\'',0x2c8,0};  // allow period and comma within a word, also stress marker (from LOPT_SYLLABLE_CAPS)
 			tr = new Translator();
 			SetupTranslator(tr,stress_lengths_jbo,NULL);
 			tr->langopts.stress_rule = 2;
 			tr->langopts.vowel_pause = 0x20c;  // pause before a word which starts with a vowel, or after a word which ends in a consonant
+//			tr->langopts.word_gap = 1;
 			tr->punct_within_word = jbo_punct_within_word;
 			tr->langopts.param[LOPT_SYLLABLE_CAPS] = 1;  // capitals indicate stressed syllables
 			SetLetterVowel(tr,'y');
 
 	case L('l','v'):  // latvian
 		{
+			static const unsigned char stress_amps_lv[8] = {17,14, 20,20, 20,22, 22,22 };
+			static const short stress_lengths_lv[8] = {180,140, 210,210, 0,0, 210,210};
 			tr = new Translator();
-			SetupTranslator(tr,stress_lengths_sk,stress_amps_sk);
+			SetupTranslator(tr,stress_lengths_lv,stress_amps_lv);
 
 			tr->langopts.stress_rule = 0;
 			tr->langopts.spelling_stress = 1;
 		}
 		break;
 
+	case L('o','m'):
+		{
+			static const unsigned char stress_amps_om[] = {16,16, 20,20, 20,24, 24,22 };
+			static const short stress_lengths_om[8] = {200,200, 200,200, 0,0, 200,200};
+			tr = new Translator();
+			SetupTranslator(tr,stress_lengths_om,stress_amps_om);
+
+			tr->langopts.stress_rule = 3;
+		}
+		break;
+
 	case L('p','l'):   // Polish
 		{
 			static const short stress_lengths_pl[8] = {160, 190,  175, 175,  0, 0,  200, 210};
 			tr->langopts.stress_flags =  0x6 | 0x10; 
 
 			tr->langopts.numbers = 0x4e1;
-			tr->langopts.numbers2 = 0x100;
+			tr->langopts.numbers2 = NUM2_100000a;
 		}
 		break;
 
 			if(name2 == L('k','n'))
 			{
 				tr->letter_bits_offset = OFFSET_KANNADA;
+				tr->langopts.numbers = 0x1;
+				tr->langopts.numbers2 = NUM2_100000;
 			}
 			tr->langopts.param[LOPT_WORD_MERGE] = 1;   // don't break vowels betwen words
 			SetIndicLetters(tr);   // call this after setting OFFSET_
 		}
 		break;
 
-
+#ifdef deleted
 	case L('t','h'):  // Thai
 		{
 			static const short stress_lengths_th[8] = {230,150, 230,230, 230,0, 230,250};
 			tr->langopts.word_gap = 0x21;   // length of a final vowel is less dependent on the next consonant, don't merge consonant with next word
 		}
 		break;
+#endif
 
 	case L('t','r'):   // Turkish
 		{

src/translate.cpp

 
 	for(ix=0; ix<word_count; ix++)
 	{
-		int j;
+		int nx;
 		int c_temp;
 		char *pn;
 		char *pw;
+		static unsigned int break_numbers1 = 0x49249248;
+		static unsigned int break_numbers2 = 0x492492a8;  // for languages which have numbers for 100,000 and 100,00,000
+		static unsigned int break_numbers3 = 0x49249268;  // for languages which have numbers for 100,000 and 1,000,000
+		unsigned int break_numbers;
 		char number_buf[80];
 
 		// start speaking at a specified word position in the text?
 
 		// digits should have been converted to Latin alphabet ('0' to '9')
 		word = pw = &sbuf[words[ix].start];
+
+		if(iswdigit(word[0]) && (langopts.numbers2 & NUM2_100000))
+		{
+			// Languages with 100000 numbers.  Remove thousands separators so that we can insert them again later
+			pn = number_buf;
+			while(pn < &number_buf[sizeof(number_buf)-3])
+			{
+				if(iswdigit(*pw))
+				{
+					*pn++ = *pw++;
+				}
+				else
+				if((*pw == langopts.thousands_sep) && (pw[1] == ' ') && iswdigit(pw[2]))
+				{
+					pw += 2;
+					ix++;  // skip "word"
+				}
+				else
+				{
+					nx = pw - word;
+					memset(word,' ',nx);
+					nx = pn - number_buf;
+					memcpy(word,number_buf,nx);
+					break;
+				}
+			}
+			pw = word;
+		}
+
 		for(n_digits=0; iswdigit(word[n_digits]); n_digits++);  // count consecutive digits
+
 		if((n_digits > 4) && (word[0] != '0'))
 		{
 			// word is entirely digits, insert commas and break into 3 digit "words"
 			number_buf[0] = ' ';
 			pn = &number_buf[1];
-			j = n_digits;
+			nx = n_digits;
+
+			if(langopts.numbers2 & NUM2_100000a)
+				break_numbers = break_numbers3;
+			else
+			if(langopts.numbers2 & NUM2_100000)
+				break_numbers = break_numbers2;
+			else
+				break_numbers = break_numbers1;
+
 			while(pn < &number_buf[sizeof(number_buf)-3])
 			{
 				if(!isdigit(c = *pw++) && (c != langopts.decimal_sep))
 					break;
 
 				*pn++ = c;
-				if((--j > 0) && (j % 3)==0)
+				if((--nx > 0) && (break_numbers & (1 << nx)))
 				{
 					if(langopts.thousands_sep != ' ')
+					{
 						*pn++ = langopts.thousands_sep;
+					}
 					*pn++ = ' ';
+					if(break_numbers & (1 << (nx-1)))
+					{
+						// the next group only has 1 digits (i.e. NUM2_10000), make it three
+						*pn++ = '0';
+						*pn++ = '0';
+					}
+					if(break_numbers & (1 << (nx-2)))
+					{
+						// the next group only has 2 digits (i.e. NUM2_10000), make it three
+						*pn++ = '0';
+					}
 				}
 			}
 			pn[0] = ' ';
 				for(pw = word; *pw != ' ';)
 				{
 					memset(number_buf,' ',9);
-					j = utf8_in(&c_temp, pw, 0);
-					memcpy(&number_buf[2],pw,j);
+					nx = utf8_in(&c_temp, pw, 0);
+					memcpy(&number_buf[2],pw,nx);
 					TranslateWord2(&number_buf[2], &words[ix], 0, 0 );
-					pw += j;
+					pw += nx;
 				}
 			}
 

src/translate.h

 #define N_RULE_GROUP2    120          // max num of two-letter rule chains
 #define N_HASH_DICT     1024
 #define N_CHARSETS        20
-#define N_LETTER_GROUPS   20
+#define N_LETTER_GROUPS   25
 
 
 /* dictionary flags, word 1 */
 	unsigned char *length_mods;
 	unsigned char *length_mods0;
 
-#define NUM_ROMAN    0x20000
-#define NUM_ROMAN_UC 0x40000
-#define NUM_NOPAUSE  0x80000
+#define NUM_ROMAN        0x20000
+#define NUM_ROMAN_UC     0x40000
+#define NUM_NOPAUSE      0x80000
+#define NUM_ROMAN_AFTER 0x200000
 
 	// bits0-1=which numbers routine to use.
 	// bit2=  thousands separator must be space
 	// bit18=Roman numbers only if upper case
 	// bit19=don't add pause after a number
 	// bit20='and' before hundreds
+	// bit21= say "roman" after the number, not before
 	int numbers;
 
+#define NUM2_100000     0x100   // numbers for 100,000 and 10,000,000
+#define NUM2_100000a    0x500   // numbers for 100,000 and 1,000,000
 	// bits 1-4  use variant form of numbers before thousands,millions,etc.
 	// bit6=(LANG=pl) two forms of plural, M or MA
 	// bit7=(LANG-ru) use MB for 1 thousand, million, etc
-	// bit8=(LANG=sw) special word for 100,000s
+	// bit8=(LANG=hi) special word for 100,000 and 10,000,000
 	// bit9=(LANG=rw) say "thousand" and "million" before its number, not after
+   // bit10=(LANG=sw) special word for 100,000 and 1,000,000
 	int numbers2;
 
 	int max_roman;