tests/check-subst

   1 #!/bin/sh
   2 # Check of --unicode-subst, --byte-subst, --widechar-subst options.
   3 set -e
   4 iconv=../src/iconv_no_i18n
   5
   6 options_ascii='--unicode-subst=<U+%04X> --byte-subst=<0x%02x> --widechar-subst=<%08x>'
   7 options_utf8='--unicode-subst=«U+%04X» --byte-subst=«0x%02x» --widechar-subst=«%08x»'
   8
   9 # Test of --byte-subst with an ASCII substitution.
  10
  11 cat > tmp-in <<\EOF
  12 Böse Bübchen
  13 EOF
  14 $iconv $options_ascii -f ASCII -t ASCII < tmp-in > tmp-out
  15 cat > tmp-ok <<\EOF
  16 B<0xc3><0xb6>se B<0xc3><0xbc>bchen
  17 EOF
  18 cmp tmp-out tmp-ok
  19
  20 # Test of --byte-subst with a non-ASCII substitution.
  21
  22 if test "`(locale charmap) 2>/dev/null`" = UTF-8; then
  23   cat > tmp-in <<\EOF
  24 Böse Bübchen
  25 EOF
  26   $iconv $options_utf8 -f ASCII -t UTF-8 2>/dev/null < tmp-in > tmp-out
  27   cat > tmp-ok <<\EOF
  28 B«0xc3»«0xb6»se B«0xc3»«0xbc»bchen
  29 EOF
  30   cmp tmp-out tmp-ok
  31 fi
  32
  33 if test "`(locale charmap) 2>/dev/null`" = UTF-8; then
  34   cat > tmp-in <<\EOF
  35 Böse Bübchen
  36 EOF
  37   $iconv $options_utf8 -f ASCII -t ISO-8859-1 2>/dev/null < tmp-in > tmp-out
  38   $iconv -f ISO-8859-1 -t UTF-8 < tmp-out > tmp-out2
  39   cat > tmp-ok <<\EOF
  40 B«0xc3»«0xb6»se B«0xc3»«0xbc»bchen
  41 EOF
  42   cmp tmp-out2 tmp-ok
  43 fi
  44
  45 # Test of --byte-subst with a very long substitution.
  46
  47 cat > tmp-in <<\EOF
  48 Böse Bübchen
  49 EOF
  50 $iconv --byte-subst='<0x%010000x>' -f ASCII -t ASCII < tmp-in > tmp-out
  51 # This printf command crashes on Solaris 10.
  52 # See <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6550204>.
  53 # Likewise on OSF/1 5.1.
  54 if printf 'B<0x%010000x><0x%010000x>se B<0x%010000x><0x%010000x>bchen\n' 0xC3 0xB6 0xC3 0xBC > tmp-ok 2>/dev/null; then
  55   cmp tmp-out tmp-ok
  56 fi
  57
  58 # Test of --unicode-subst with an ASCII substitution.
  59
  60 cat > tmp-in <<\EOF
  61 Böse Bübchen
  62 EOF
  63 $iconv $options_ascii -f UTF-8 -t ASCII < tmp-in > tmp-out
  64 cat > tmp-ok <<\EOF
  65 B<U+00F6>se B<U+00FC>bchen
  66 EOF
  67 cmp tmp-out tmp-ok
  68
  69 cat > tmp-in <<\EOF
  70 Russian (Русский)
  71 EOF
  72 $iconv $options_ascii -f UTF-8 -t ISO-8859-1 2>/dev/null < tmp-in | $iconv -f ISO-8859-1 -t UTF-8 > tmp-out
  73 cat > tmp-ok <<\EOF
  74 Russian (<U+0420><U+0443><U+0441><U+0441><U+043A><U+0438><U+0439>)
  75 EOF
  76 cmp tmp-out tmp-ok
  77
  78 # Test of --unicode-subst with a non-ASCII substitution.
  79
  80 if test "`(locale charmap) 2>/dev/null`" = UTF-8; then
  81   cat > tmp-in <<\EOF
  82 Russian (Русский)
  83 EOF
  84   $iconv $options_utf8 -f UTF-8 -t ISO-8859-1 2>/dev/null < tmp-in > tmp-out
  85   $iconv -f ISO-8859-1 -t UTF-8 < tmp-out > tmp-out2
  86   cat > tmp-ok <<\EOF
  87 Russian («U+0420»«U+0443»«U+0441»«U+0441»«U+043A»«U+0438»«U+0439»)
  88 EOF
  89   cmp tmp-out2 tmp-ok
  90 fi
  91
  92 # Test of --unicode-subst with a very long substitution.
  93
  94 cat > tmp-in <<\EOF
  95 Böse Bübchen
  96 EOF
  97 $iconv --unicode-subst='<U+%010000X>' -f UTF-8 -t ASCII < tmp-in > tmp-out
  98 # This printf command crashes on Solaris 10.
  99 # See <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6550204>.
 100 if printf 'B<U+%010000X>se B<U+%010000X>bchen\n' 0x00F6 0x00FC > tmp-ok 2>/dev/null; then
 101   cmp tmp-out tmp-ok
 102 fi
 103
 104 cat > tmp-in <<\EOF
 105 Böse Bübchen
 106 EOF
 107 $iconv --byte-subst='<0x%010000x>' -f ASCII -t ASCII < tmp-in > tmp-out
 108 # This printf command crashes on Solaris 10.
 109 # See <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6550204>.
 110 # Likewise on OSF/1 5.1.
 111 if printf 'B<0x%010000x><0x%010000x>se B<0x%010000x><0x%010000x>bchen\n' 0xC3 0xB6 0xC3 0xBC > tmp-ok 2>/dev/null; then
 112   cmp tmp-out tmp-ok
 113 fi
 114
 115 # Test of --widechar-subst:
 116 # wcrtomb() doesn't exist on FreeBSD 4.0 and is broken on MacOS X 10.3.
 117 # So far this has been tested only on a glibc system with !__STDC_ISO_10646__.
 118
 119 if false && test "`(locale charmap) 2>/dev/null`" = UTF-8; then
 120   cat > tmp-in <<\EOF
 121 Russian (Русский)
 122 EOF
 123   $iconv -f char -t wchar_t < tmp-in > tmp-inw
 124   LC_ALL=C                $iconv $options_ascii -f wchar_t -t ASCII < tmp-inw > tmp-out1
 125   LC_ALL=de_DE.ISO-8859-1 $iconv $options_ascii -f wchar_t -t ASCII < tmp-inw > tmp-out2
 126   cat > tmp-ok <<\EOF
 127 Russian (<00000420><00000443><00000441><00000441><0000043a><00000438><00000439>)
 128 EOF
 129   cmp tmp-out1 tmp-ok
 130   cmp tmp-out2 tmp-ok
 131   if test "`(LC_ALL=de_DE.ISO-8859-1 locale charmap) 2>/dev/null`" = ISO-8859-1; then
 132     options_latin1=`echo " $options_utf8" | $iconv -f UTF-8 -t ISO-8859-1`
 133     LC_ALL=de_DE.ISO-8859-1 $iconv $options_latin1 -f wchar_t -t UTF-8 < tmp-inw > tmp-out1
 134     cat > tmp-ok <<\EOF
 135 Russian («00000420»«00000443»«00000441»«00000441»«0000043a»«00000438»«00000439»)
 136 EOF
 137     cmp tmp-out1 tmp-ok
 138   fi
 139 fi
 140
 141 rm -f tmp-in* tmp-out* tmp-ok
 142 exit 0