[1.26.08] Fixed --stdout option in "espeak" program was writing to file "stdout".

Fix to UTF-8/8-bit character set recognition.
lang=de, more additions to translation rules, changed phoneme name [OI] to [OY].
lang=it, turned of "syntactic doubling" (temporarily?)


git-svn-id: https://espeak.svn.sourceforge.net/svnroot/espeak/trunk@51 d46cf337-b52f-0410-862d-fd96e6ae7743

phsource/compile_report

@@ -16,7 +16,7 @@
       fr  32  118
    fr_ca  11  118
       hi  49  131
-      hu  24  114
+      hu  23  113
       nl  26  117
       pl  15  107
       sk  29  125
@@ -224,7 +224,7 @@
   2  ufric/hu_fi    fi
   2  ufric/ll       base is
   3  ufric/s        base es pt
- 13  ufric/s_       base fi fr hu es pt ro
+ 11  ufric/s_       base fi fr hu es pt ro
  15  ufric/s!       base fi hu es pt
   2  ufric/s_continue fi hu
   2  ufric/sh       base pt_pt
@@ -399,8 +399,8 @@
   6  vocw/v         base ro
   4  vocw/zh        base ro
   2  vocw/zh_rfx    base
-  4  voc/z          base fr hu ro
-  4  voc/z_         base fr hu ro
+  3  voc/z          base fr ro
+  3  voc/z_         base fr ro
   4  voc/zh         base sk ro
   3  voc/zh_        base ro
   1  voc/zh_2       sk

phsource/ph_german

@@ -129,7 +129,7 @@ phoneme aI
 endphoneme
 
 
-phoneme OI
+phoneme OY
   vowel starttype (o) endtype (i)
   length 220
   formants vdiph/oi

phsource/ph_hungarian

@@ -189,12 +189,3 @@ phoneme s2     // second part of geminated [s:]
 endphoneme
 
 
-phoneme z2      // [z] but without switchvoicing
-  vcd alv frc sibilant
-  vowelout f1=0  f2=1700 -300 300  f3=-100 60  len=50 
-  formants voc/z+ufric/s_
-  before _ voc/z_+ufric/s_%80
-  lengthmod 6
-endphoneme
-
-