Return to check-subst CVS log

Up to [ELWIX - Embedded LightWeight unIX -] / embedaddon / libiconv / tests

libiconv

    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: 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
   56:   cmp tmp-out tmp-ok
   57: fi
   58: 
   59: # Test of --unicode-subst with an ASCII substitution.
   60: 
   61: cat > tmp-in <<\EOF
   62: Böse Bübchen
   63: EOF
   64: $iconv $options_ascii -f UTF-8 -t ASCII < tmp-in > tmp-out
   65: cat > tmp-ok <<\EOF
   66: B<U+00F6>se B<U+00FC>bchen
   67: EOF
   68: cmp tmp-out tmp-ok
   69: 
   70: cat > tmp-in <<\EOF
   71: Russian (Русский)
   72: EOF
   73: $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
   74: cat > tmp-ok <<\EOF
   75: Russian (<U+0420><U+0443><U+0441><U+0441><U+043A><U+0438><U+0439>)
   76: EOF
   77: cmp tmp-out tmp-ok
   78: 
   79: # Test of --unicode-subst with a non-ASCII substitution.
   80: 
   81: if test "`(locale charmap) 2>/dev/null`" = UTF-8; then
   82:   cat > tmp-in <<\EOF
   83: Russian (Русский)
   84: EOF
   85:   $iconv $options_utf8 -f UTF-8 -t ISO-8859-1 2>/dev/null < tmp-in > tmp-out
   86:   $iconv -f ISO-8859-1 -t UTF-8 < tmp-out > tmp-out2
   87:   cat > tmp-ok <<\EOF
   88: Russian («U+0420»«U+0443»«U+0441»«U+0441»«U+043A»«U+0438»«U+0439»)
   89: EOF
   90:   cmp tmp-out2 tmp-ok
   91: fi
   92: 
   93: # Test of --unicode-subst with a very long substitution.
   94: 
   95: cat > tmp-in <<\EOF
   96: Böse Bübchen
   97: EOF
   98: $iconv --unicode-subst='<U+%010000X>' -f UTF-8 -t ASCII < tmp-in > tmp-out
   99: # This printf command crashes on Solaris 10.
  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: 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
  110:   cmp tmp-out tmp-ok
  111: fi
  112: 
  113: # Test of --widechar-subst:
  114: # wcrtomb() doesn't exist on FreeBSD 4.0 and is broken on MacOS X 10.3.
  115: # So far this has been tested only on a glibc system with !__STDC_ISO_10646__.
  116: 
  117: if false && test "`(locale charmap) 2>/dev/null`" = UTF-8; then
  118:   cat > tmp-in <<\EOF
  119: Russian (Русский)
  120: EOF
  121:   $iconv -f char -t wchar_t < tmp-in > tmp-inw
  122:   LC_ALL=C                $iconv $options_ascii -f wchar_t -t ASCII < tmp-inw > tmp-out1
  123:   LC_ALL=de_DE.ISO-8859-1 $iconv $options_ascii -f wchar_t -t ASCII < tmp-inw > tmp-out2
  124:   cat > tmp-ok <<\EOF
  125: Russian (<00000420><00000443><00000441><00000441><0000043a><00000438><00000439>)
  126: EOF
  127:   cmp tmp-out1 tmp-ok
  128:   cmp tmp-out2 tmp-ok
  129:   if test "`(LC_ALL=de_DE.ISO-8859-1 locale charmap) 2>/dev/null`" = ISO-8859-1; then
  130:     options_latin1=`echo " $options_utf8" | $iconv -f UTF-8 -t ISO-8859-1`
  131:     LC_ALL=de_DE.ISO-8859-1 $iconv $options_latin1 -f wchar_t -t UTF-8 < tmp-inw > tmp-out1
  132:     cat > tmp-ok <<\EOF
  133: Russian («00000420»«00000443»«00000441»«00000441»«0000043a»«00000438»«00000439»)
  134: EOF
  135:     cmp tmp-out1 tmp-ok
  136:   fi
  137: fi
  138: 
  139: rm -f tmp-in* tmp-out* tmp-ok
  140: exit 0