libstdc++-v3/testsuite/ext/codecvt/char-1.cc

   1 // { dg-require-iconv "UCS-2BE" }
   2 // { dg-require-iconv "ISO-8859-15" }
   3
   4 // 2000-08-22 Benjamin Kosnik <bkoz@cygnus.com>
   5
   6 // Copyright (C) 2000-2019 Free Software Foundation, Inc.
   7 //
   8 // This file is part of the GNU ISO C++ Library.  This library is free
   9 // software; you can redistribute it and/or modify it under the
  10 // terms of the GNU General Public License as published by the
  11 // Free Software Foundation; either version 3, or (at your option)
  12 // any later version.
  13
  14 // This library is distributed in the hope that it will be useful,
  15 // but WITHOUT ANY WARRANTY; without even the implied warranty of
  16 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  17 // GNU General Public License for more details.
  18
  19 // You should have received a copy of the GNU General Public License along
  20 // with this library; see the file COPYING3.  If not see
  21 // <http://www.gnu.org/licenses/>.
  22
  23 // 22.2.1.5 - Template class codecvt [lib.locale.codecvt]
  24 #include <locale>
  25 #include <cstring>
  26 #include <testsuite_hooks.h>
  27 #include <ext/codecvt_specializations.h>
  28
  29 /*
  30 > how do I check that these conversions are correct?
  31 Very easy.  Since all the characters are from ASCII you simply
  32 zero-extend the values.
  33
  34 drepper$ echo 'black pearl jasmine tea' | od -t x1
  35 0000000 62 6c 61 63 6b 20 70 65 61 72 6c 20 6a 61 73 6d
  36 0000020 69 6e 65 20 74 65 61 0a
  37
  38 So the UCS-2 string is
  39
  40 0x0062, 0x006c, 0x0061, ...
  41
  42 You get the idea.  With iconv() you have to take care of the
  43 byte-order, though.  UCS-2 can mean little- or big endian.  Looking at
  44 your result
  45
  46 > $9 = 25856
  47
  48 it shows that the other byte-order is used (25856 == 0x6500).
  49 */
  50
  51 // Partial specialization using encoding_state.
  52 // codecvt<unicode_t, char, encoding_state>
  53 // UNICODE - UCS2 (big endian)
  54 void test01()
  55 {
  56   using namespace std;
  57   typedef codecvt_base::result                  result;
  58   typedef unsigned short                        int_type;
  59   typedef char                                  ext_type;
  60   typedef __gnu_cxx::encoding_state                     state_type;
  61   typedef codecvt<int_type, ext_type, state_type>       unicode_codecvt;
  62   typedef char_traits<int_type>                 int_traits;
  63   typedef char_traits<ext_type>                 ext_traits;
  64
  65   const ext_type*       e_lit = "black pearl jasmine tea";
  66   int                   size = strlen(e_lit);
  67
  68   char  i_lit_base[50] __attribute__((aligned(__alignof__(int_type)))) =
  69   {
  70     char(0x00), char(0x62), char(0x00), char(0x6c), char(0x00), char(0x61),
  71     char(0x00), char(0x63), char(0x00), char(0x6b), char(0x00), char(0x20),
  72     char(0x00), char(0x70), char(0x00), char(0x65), char(0x00), char(0x61),
  73     char(0x00), char(0x72), char(0x00), char(0x6c), char(0x00), char(0x20),
  74     char(0x00), char(0x6a), char(0x00), char(0x61), char(0x00), char(0x73),
  75     char(0x00), char(0x6d), char(0x00), char(0x69), char(0x00), char(0x6e),
  76     char(0x00), char(0x65), char(0x00), char(0x20), char(0x00), char(0x74),
  77     char(0x00), char(0x65), char(0x00), char(0x61), char(0x00), char(0xa0)
  78   };
  79   const int_type*       i_lit = reinterpret_cast<int_type*>(i_lit_base);
  80
  81   const ext_type*       efrom_next;
  82   const int_type*       ifrom_next;
  83   ext_type*             e_arr = new ext_type[size + 1];
  84   ext_type*             eto_next;
  85   int_type*             i_arr = new int_type[size + 1];
  86   int_type*             ito_next;
  87
  88   // construct a locale object with the specialized facet.
  89   locale                loc(locale::classic(), new unicode_codecvt);
  90   // sanity check the constructed locale has the specialized facet.
  91   VERIFY( has_facet<unicode_codecvt>(loc) );
  92   const unicode_codecvt& cvt = use_facet<unicode_codecvt>(loc);
  93
  94   // in
  95   //  unicode_codecvt::state_type state01("UCS-2BE", "ISO-8859-15", 0xfeff, 0);
  96   unicode_codecvt::state_type state01("UCS-2BE", "ISO-8859-15", 0, 0);
  97
  98   // internal encoding is bigger because of bom
  99   result r1 = cvt.in(state01, e_lit, e_lit + size, efrom_next,
 100                      i_arr, i_arr + size + 1, ito_next);
 101   VERIFY( r1 == codecvt_base::ok );
 102   VERIFY( !int_traits::compare(i_arr, i_lit, size) );
 103   VERIFY( efrom_next == e_lit + size );
 104   VERIFY( ito_next == i_arr + size );
 105
 106   // out
 107   unicode_codecvt::state_type state02("UCS-2BE", "ISO-8859-15", 0, 0);
 108   result r2 = cvt.out(state02, i_lit, i_lit + size, ifrom_next,
 109                        e_arr, e_arr + size, eto_next);
 110   VERIFY( r2 == codecvt_base::ok );
 111   VERIFY( !ext_traits::compare(e_arr, e_lit, size) );
 112   VERIFY( ifrom_next == i_lit + size );
 113   VERIFY( eto_next == e_arr + size );
 114
 115   // unshift
 116   ext_traits::copy(e_arr, e_lit, size);
 117   unicode_codecvt::state_type state03("UCS-2BE", "ISO-8859-15", 0, 0);
 118   result r3 = cvt.unshift(state03, e_arr, e_arr + size, eto_next);
 119   VERIFY( r3 == codecvt_base::noconv );
 120   VERIFY( !ext_traits::compare(e_arr, e_lit, size) );
 121   VERIFY( eto_next == e_arr );
 122
 123   int i = cvt.encoding();
 124   VERIFY( i == 2 ); // Target-dependent.
 125
 126   VERIFY( !cvt.always_noconv() );
 127
 128   unicode_codecvt::state_type state04("UCS-2BE", "ISO-8859-15", 0, 0);
 129   int j = cvt.length(state03, e_lit, e_lit + size, 5);
 130   VERIFY( j == 5 );
 131
 132   int k = cvt.max_length();
 133   VERIFY( k == 1 );
 134
 135   delete [] e_arr;
 136   delete [] i_arr;
 137 }
 138
 139 int main ()
 140 {
 141   test01();
 142   return 0;
 143 }