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: Den V. Tsopa <tdv@edisoft.ru>
22 *
23 */
24 /*
25 * The source code included in this files was separated from mbfilter_ru.c
26 * by moriyoshi koizumi <moriyoshi@php.net> on 4 dec 2002.
27 *
28 */
29
30 #include "mbfilter.h"
31 #include "mbfilter_cp1251.h"
32 #include "unicode_table_cp1251.h"
33
34 static int mbfl_filt_ident_cp1251(int c, mbfl_identify_filter *filter);
35
36 static const char *mbfl_encoding_cp1251_aliases[] = {"CP1251", "CP-1251", "WINDOWS-1251", NULL};
37
38 const mbfl_encoding mbfl_encoding_cp1251 = {
39 mbfl_no_encoding_cp1251,
40 "Windows-1251",
41 "Windows-1251",
42 (const char *(*)[])&mbfl_encoding_cp1251_aliases,
43 NULL,
44 MBFL_ENCTYPE_SBCS,
45 &vtbl_cp1251_wchar,
46 &vtbl_wchar_cp1251
47 };
48
49 const struct mbfl_identify_vtbl vtbl_identify_cp1251 = {
50 mbfl_no_encoding_cp1251,
51 mbfl_filt_ident_common_ctor,
52 mbfl_filt_ident_cp1251
53 };
54
55 const struct mbfl_convert_vtbl vtbl_wchar_cp1251 = {
56 mbfl_no_encoding_wchar,
57 mbfl_no_encoding_cp1251,
58 mbfl_filt_conv_common_ctor,
59 NULL,
60 mbfl_filt_conv_wchar_cp1251,
61 mbfl_filt_conv_common_flush,
62 NULL,
63 };
64
65 const struct mbfl_convert_vtbl vtbl_cp1251_wchar = {
66 mbfl_no_encoding_cp1251,
67 mbfl_no_encoding_wchar,
68 mbfl_filt_conv_common_ctor,
69 NULL,
70 mbfl_filt_conv_cp1251_wchar,
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 * cp1251 => wchar
79 */
80 int
mbfl_filt_conv_cp1251_wchar(int c,mbfl_convert_filter * filter)81 mbfl_filt_conv_cp1251_wchar(int c, mbfl_convert_filter *filter)
82 {
83 int s;
84
85 if (c >= 0 && c < cp1251_ucs_table_min) {
86 s = c;
87 } else if (c >= cp1251_ucs_table_min && c < 0x100) {
88 s = cp1251_ucs_table[c - cp1251_ucs_table_min];
89 if (s <= 0) {
90 s = c;
91 s &= MBFL_WCSPLANE_MASK;
92 s |= MBFL_WCSPLANE_CP1251;
93 }
94 } else {
95 s = c;
96 s &= MBFL_WCSGROUP_MASK;
97 s |= MBFL_WCSGROUP_THROUGH;
98 }
99
100 CK((*filter->output_function)(s, filter->data));
101
102 return c;
103 }
104
105 /*
106 * wchar => cp1251
107 */
108 int
mbfl_filt_conv_wchar_cp1251(int c,mbfl_convert_filter * filter)109 mbfl_filt_conv_wchar_cp1251(int c, mbfl_convert_filter *filter)
110 {
111 int s, n;
112
113 if (c < 0x80) {
114 s = c;
115 } else {
116 s = -1;
117 n = cp1251_ucs_table_len-1;
118 while (n >= 0) {
119 if (c == cp1251_ucs_table[n]) {
120 s = cp1251_ucs_table_min + n;
121 break;
122 }
123 n--;
124 }
125 if (s <= 0 && (c & ~MBFL_WCSPLANE_MASK) == MBFL_WCSPLANE_CP1251) {
126 s = c & MBFL_WCSPLANE_MASK;
127 }
128 }
129
130 if (s >= 0) {
131 CK((*filter->output_function)(s, filter->data));
132 } else {
133 CK(mbfl_filt_conv_illegal_output(c, filter));
134 }
135
136 return c;
137 }
138
139 /* all of this is so ugly now! */
mbfl_filt_ident_cp1251(int c,mbfl_identify_filter * filter)140 static int mbfl_filt_ident_cp1251(int c, mbfl_identify_filter *filter)
141 {
142 if (c >= 0x80 && c <= 0xff)
143 filter->flag = 0;
144 else
145 filter->flag = 1; /* not it */
146 return c;
147 }
148
