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 #include "mbfilter.h"
31 #include "mbfilter_cp1252.h"
32 #include "unicode_table_cp1252.h"
33
34 static int mbfl_filt_ident_cp1252(int c, mbfl_identify_filter *filter);
35
36 static const char *mbfl_encoding_cp1252_aliases[] = {"cp1252", NULL};
37
38 const mbfl_encoding mbfl_encoding_cp1252 = {
39 mbfl_no_encoding_cp1252,
40 "Windows-1252",
41 "Windows-1252",
42 (const char *(*)[])&mbfl_encoding_cp1252_aliases,
43 NULL,
44 MBFL_ENCTYPE_SBCS,
45 &vtbl_cp1252_wchar,
46 &vtbl_wchar_cp1252
47 };
48
49 const struct mbfl_identify_vtbl vtbl_identify_cp1252 = {
50 mbfl_no_encoding_cp1252,
51 mbfl_filt_ident_common_ctor,
52 mbfl_filt_ident_cp1252
53 };
54
55 const struct mbfl_convert_vtbl vtbl_cp1252_wchar = {
56 mbfl_no_encoding_cp1252,
57 mbfl_no_encoding_wchar,
58 mbfl_filt_conv_common_ctor,
59 NULL,
60 mbfl_filt_conv_cp1252_wchar,
61 mbfl_filt_conv_common_flush,
62 NULL,
63 };
64
65 const struct mbfl_convert_vtbl vtbl_wchar_cp1252 = {
66 mbfl_no_encoding_wchar,
67 mbfl_no_encoding_cp1252,
68 mbfl_filt_conv_common_ctor,
69 NULL,
70 mbfl_filt_conv_wchar_cp1252,
71 mbfl_filt_conv_common_flush,
72 NULL,
73 };
74
75 #define CK(statement) do { if ((statement) < 0) return (-1); } while (0)
76
77 /*
78 * wchar => cp1252
79 */
mbfl_filt_conv_wchar_cp1252(int c,mbfl_convert_filter * filter)80 int mbfl_filt_conv_wchar_cp1252(int c, mbfl_convert_filter *filter)
81 {
82 int s=-1, n;
83
84 if (c >= 0x100) {
85 /* look it up from the cp1252 table */
86 s = -1;
87 n = 31;
88 while (n >= 0) {
89 if (c == cp1252_ucs_table[n] && c != 0xfffe) {
90 s = 0x80 + n;
91 break;
92 }
93 n--;
94 }
95 if (s <= 0 && (c & ~MBFL_WCSPLANE_MASK) == MBFL_WCSPLANE_8859_1)
96 {
97 s = c & MBFL_WCSPLANE_MASK;
98 }
99 }
100 else if (c >= 0 && c < 0x100) {
101 s = c;
102 }
103 if (s >= 0) {
104 CK((*filter->output_function)(s, filter->data));
105 } else {
106 CK(mbfl_filt_conv_illegal_output(c, filter));
107 }
108 return c;
109 }
110
111 /*
112 * cp1252 => wchar
113 */
mbfl_filt_conv_cp1252_wchar(int c,mbfl_convert_filter * filter)114 int mbfl_filt_conv_cp1252_wchar(int c, mbfl_convert_filter *filter)
115 {
116 int s;
117
118 if (c >= 0x80 && c < 0xa0) {
119 s = cp1252_ucs_table[c - 0x80];
120 } else {
121 s = c;
122 }
123
124 CK((*filter->output_function)(s, filter->data));
125
126 return c;
127 }
128
129 /* We only distinguish the MS extensions to ISO-8859-1.
130 * Actually, this is pretty much a NO-OP, since the identification
131 * system doesn't allow us to discriminate between a positive match,
132 * a possible match and a definite non-match.
133 * The problem here is that cp1252 looks like SJIS for certain chars.
134 * */
mbfl_filt_ident_cp1252(int c,mbfl_identify_filter * filter)135 static int mbfl_filt_ident_cp1252(int c, mbfl_identify_filter *filter)
136 {
137 if (c >= 0x80 && c < 0xa0)
138 filter->flag = 0;
139 else
140 filter->flag = 1; /* not it */
141 return c;
142 }
143
