release/src/router/libiconv/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 # Make sure the charset.alias file is found.
   6 CHARSETALIASDIR=../lib
   7 export CHARSETALIASDIR
   8
   9 options_ascii='--unicode-subst=<U+%04X> --byte-subst=<0x%02x> --widechar-subst=<%08x>'
  10 options_utf8='--unicode-subst=«U+%04X» --byte-subst=«0x%02x» --widechar-subst=«%08x»'
  11
  12 # Test of --byte-subst with an ASCII substitution.
  13
  14 cat > tmp-in <<\EOF
  15 Böse Bübchen
  16 EOF
  17 $iconv $options_ascii -f ASCII -t ASCII < tmp-in > tmp-out
  18 cat > tmp-ok <<\EOF
  19 B<0xc3><0xb6>se B<0xc3><0xbc>bchen
  20 EOF
  21 cmp tmp-out tmp-ok
  22
  23 # Test of --byte-subst with a non-ASCII substitution.
  24
  25 if test "`(locale charmap) 2>/dev/null`" = UTF-8; then
  26   cat > tmp-in <<\EOF
  27 Böse Bübchen
  28 EOF
  29   $iconv $options_utf8 -f ASCII -t UTF-8 2>/dev/null < tmp-in > tmp-out
  30   cat > tmp-ok <<\EOF
  31 B«0xc3»«0xb6»se B«0xc3»«0xbc»bchen
  32 EOF
  33   cmp tmp-out tmp-ok
  34 fi
  35
  36 if test "`(locale charmap) 2>/dev/null`" = UTF-8; then
  37   cat > tmp-in <<\EOF
  38 Böse Bübchen
  39 EOF
  40   $iconv $options_utf8 -f ASCII -t ISO-8859-1 2>/dev/null < tmp-in > tmp-out
  41   $iconv -f ISO-8859-1 -t UTF-8 < tmp-out > tmp-out2
  42   cat > tmp-ok <<\EOF
  43 B«0xc3»«0xb6»se B«0xc3»«0xbc»bchen
  44 EOF
  45   cmp tmp-out2 tmp-ok
  46 fi
  47
  48 # Test of --byte-subst with a very long substitution.
  49
  50 cat > tmp-in <<\EOF
  51 Böse Bübchen
  52 EOF
  53 $iconv --byte-subst='<0x%010000x>' -f ASCII -t ASCII < tmp-in > tmp-out
  54 # This printf command crashes on Solaris 10.
  55 # See <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6550204>.
  56 # Likewise on OSF/1 5.1.
  57 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
  58   cmp tmp-out tmp-ok
  59 fi
  60
  61 # Test of --unicode-subst with an ASCII substitution.
  62
  63 cat > tmp-in <<\EOF
  64 Böse Bübchen
  65 EOF
  66 $iconv $options_ascii -f UTF-8 -t ASCII < tmp-in > tmp-out
  67 cat > tmp-ok <<\EOF
  68 B<U+00F6>se B<U+00FC>bchen
  69 EOF
  70 cmp tmp-out tmp-ok
  71
  72 cat > tmp-in <<\EOF
  73 Russian (Русский)
  74 EOF
  75 $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
  76 cat > tmp-ok <<\EOF
  77 Russian (<U+0420><U+0443><U+0441><U+0441><U+043A><U+0438><U+0439>)
  78 EOF
  79 cmp tmp-out tmp-ok
  80
  81 # Test of --unicode-subst with a non-ASCII substitution.
  82
  83 if test "`(locale charmap) 2>/dev/null`" = UTF-8; then
  84   cat > tmp-in <<\EOF
  85 Russian (Русский)
  86 EOF
  87   $iconv $options_utf8 -f UTF-8 -t ISO-8859-1 2>/dev/null < tmp-in > tmp-out
  88   $iconv -f ISO-8859-1 -t UTF-8 < tmp-out > tmp-out2
  89   cat > tmp-ok <<\EOF
  90 Russian («U+0420»«U+0443»«U+0441»«U+0441»«U+043A»«U+0438»«U+0439»)
  91 EOF
  92   cmp tmp-out2 tmp-ok
  93 fi
  94
  95 # Test of --unicode-subst with a very long substitution.
  96
  97 cat > tmp-in <<\EOF
  98 Böse Bübchen
  99 EOF
 100 $iconv --unicode-subst='<U+%010000X>' -f UTF-8 -t ASCII < tmp-in > tmp-out
 101 # This printf command crashes on Solaris 10.
 102 # See <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6550204>.
 103 if printf 'B<U+%010000X>se B<U+%010000X>bchen\n' 0x00F6 0x00FC > tmp-ok 2>/dev/null; then
 104   cmp tmp-out tmp-ok
 105 fi
 106
 107 cat > tmp-in <<\EOF
 108 Böse Bübchen
 109 EOF
 110 $iconv --byte-subst='<0x%010000x>' -f ASCII -t ASCII < tmp-in > tmp-out
 111 # This printf command crashes on Solaris 10.
 112 # See <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6550204>.
 113 # Likewise on OSF/1 5.1.
 114 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
 115   cmp tmp-out tmp-ok
 116 fi
 117
 118 # Test of --widechar-subst:
 119 # wcrtomb() doesn't exist on FreeBSD 4.0 and is broken on MacOS X 10.3.
 120 # So far this has been tested only on a glibc system with !__STDC_ISO_10646__.
 121
 122 if false && test "`(locale charmap) 2>/dev/null`" = UTF-8; then
 123   cat > tmp-in <<\EOF
 124 Russian (Русский)
 125 EOF
 126   $iconv -f char -t wchar_t < tmp-in > tmp-inw
 127   LC_ALL=C                $iconv $options_ascii -f wchar_t -t ASCII < tmp-inw > tmp-out1
 128   LC_ALL=de_DE.ISO-8859-1 $iconv $options_ascii -f wchar_t -t ASCII < tmp-inw > tmp-out2
 129   cat > tmp-ok <<\EOF
 130 Russian (<00000420><00000443><00000441><00000441><0000043a><00000438><00000439>)
 131 EOF
 132   cmp tmp-out1 tmp-ok
 133   cmp tmp-out2 tmp-ok
 134   if test "`(LC_ALL=de_DE.ISO-8859-1 locale charmap) 2>/dev/null`" = ISO-8859-1; then
 135     options_latin1=`echo " $options_utf8" | $iconv -f UTF-8 -t ISO-8859-1`
 136     LC_ALL=de_DE.ISO-8859-1 $iconv $options_latin1 -f wchar_t -t UTF-8 < tmp-inw > tmp-out1
 137     cat > tmp-ok <<\EOF
 138 Russian («00000420»«00000443»«00000441»«00000441»«0000043a»«00000438»«00000439»)
 139 EOF
 140     cmp tmp-out1 tmp-ok
 141   fi
 142 fi
 143
 144 rm -f tmp-in* tmp-out* tmp-ok
 145 exit 0