root/ext/mbstring/libmbfl/filters/mbfilter_cp1252.c

DEFINITIONS

1. mbfl_filt_conv_wchar_cp1252
2. mbfl_filt_conv_cp1252_wchar
3. mbfl_filt_ident_cp1252

   1 /*
   2  * "streamable kanji code filter and converter"
   3  * Copyright (c) 1998-2002 HappySize, Inc. All rights reserved.
   4  *
   5  * LICENSE NOTICES
   6  *
   7  * This file is part of "streamable kanji code filter and converter",
   8  * which is distributed under the terms of GNU Lesser General Public
   9  * License (version 2) as published by the Free Software Foundation.
  10  *
  11  * This software is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14  * GNU Lesser General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU Lesser General Public
  17  * License along with "streamable kanji code filter and converter";
  18  * if not, write to the Free Software Foundation, Inc., 59 Temple Place,
  19  * Suite 330, Boston, MA  02111-1307  USA
  20  *
  21  * The author of this part: Wez Furlong <wez@thebrainroom.com>
  22  *
  23  */
  24 /*
  25  * the source code included in this files was separated from mbfilter.c
  26  * by moriyoshi koizumi <moriyoshi@php.net> on 4 dec 2002.
  27  *
  28  */
  29 
  30 #ifdef HAVE_CONFIG_H
  31 #include "config.h"
  32 #endif
  33 
  34 #include "mbfilter.h"
  35 #include "mbfilter_cp1252.h"
  36 #include "unicode_table_cp1252.h"
  37 
  38 static int mbfl_filt_ident_cp1252(int c, mbfl_identify_filter *filter);
  39 
  40 static const char *mbfl_encoding_cp1252_aliases[] = {"cp1252", NULL};
  41 
  42 const mbfl_encoding mbfl_encoding_cp1252 = {
  43         mbfl_no_encoding_cp1252,
  44         "Windows-1252",
  45         "Windows-1252",
  46         (const char *(*)[])&mbfl_encoding_cp1252_aliases,
  47         NULL,
  48         MBFL_ENCTYPE_SBCS
  49 };
  50 
  51 const struct mbfl_identify_vtbl vtbl_identify_cp1252 = {
  52         mbfl_no_encoding_cp1252,
  53         mbfl_filt_ident_common_ctor,
  54         mbfl_filt_ident_common_dtor,
  55         mbfl_filt_ident_cp1252
  56 };
  57 
  58 const struct mbfl_convert_vtbl vtbl_cp1252_wchar = {
  59         mbfl_no_encoding_cp1252,
  60         mbfl_no_encoding_wchar,
  61         mbfl_filt_conv_common_ctor,
  62         mbfl_filt_conv_common_dtor,
  63         mbfl_filt_conv_cp1252_wchar,
  64         mbfl_filt_conv_common_flush
  65 };
  66 
  67 const struct mbfl_convert_vtbl vtbl_wchar_cp1252 = {
  68         mbfl_no_encoding_wchar,
  69         mbfl_no_encoding_cp1252,
  70         mbfl_filt_conv_common_ctor,
  71         mbfl_filt_conv_common_dtor,
  72         mbfl_filt_conv_wchar_cp1252,
  73         mbfl_filt_conv_common_flush
  74 };
  75 
  76 #define CK(statement)   do { if ((statement) < 0) return (-1); } while (0)
  77 
  78 /*
  79  * wchar => cp1252
  80  */
  81 int mbfl_filt_conv_wchar_cp1252(int c, mbfl_convert_filter *filter)
  82 {
  83         int s=-1, n;
  84 
  85         if (c >= 0x100) {
  86                 /* look it up from the cp1252 table */
  87                 s = -1;
  88                 n = 31;
  89                 while (n >= 0) {
  90                         if (c == cp1252_ucs_table[n] && c != 0xfffe) {
  91                                 s = 0x80 + n;
  92                                 break;
  93                         }
  94                         n--;
  95                 }
  96                 if (s <= 0 && (c & ~MBFL_WCSPLANE_MASK) == MBFL_WCSPLANE_8859_1)
  97                 {
  98                         s = c & MBFL_WCSPLANE_MASK;
  99                 }
 100         }
 101         else if (c >= 0 && c < 0x100) {
 102                 s = c;
 103         }
 104         if (s >= 0) {
 105                 CK((*filter->output_function)(s, filter->data));
 106         } else {
 107                 if (filter->illegal_mode != MBFL_OUTPUTFILTER_ILLEGAL_MODE_NONE) {
 108                         CK(mbfl_filt_conv_illegal_output(c, filter));
 109                 }
 110         }
 111         return c;
 112 }
 113 
 114 /*
 115  * cp1252 => wchar
 116  */
 117 int mbfl_filt_conv_cp1252_wchar(int c, mbfl_convert_filter *filter)
 118 {
 119         int s;
 120 
 121         if (c >= 0x80 && c < 0xa0) {
 122                 s = cp1252_ucs_table[c - 0x80];
 123         } else {
 124                 s = c;
 125         }
 126 
 127         CK((*filter->output_function)(s, filter->data));
 128 
 129         return c;
 130 }
 131 
 132 /* We only distinguish the MS extensions to ISO-8859-1.
 133  * Actually, this is pretty much a NO-OP, since the identification
 134  * system doesn't allow us to discriminate between a positive match,
 135  * a possible match and a definite non-match.
 136  * The problem here is that cp1252 looks like SJIS for certain chars.
 137  * */
 138 static int mbfl_filt_ident_cp1252(int c, mbfl_identify_filter *filter)
 139 {
 140         if (c >= 0x80 && c < 0xa0)
 141                 filter->flag = 0;
 142         else
 143                 filter->flag = 1; /* not it */
 144         return c;
 145 }
 146 
 147