1 /* Character set conversion.
2 Copyright (C) 2007, 2009-2023 Free Software Foundation, Inc.
3
4 This file is free software: you can redistribute it and/or modify
5 it under the terms of the GNU Lesser General Public License as
6 published by the Free Software Foundation; either version 2.1 of the
7 License, or (at your option) any later version.
8
9 This file is distributed in the hope that it will be useful,
10 but WITHOUT ANY WARRANTY; without even the implied warranty of
11 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 GNU Lesser General Public License for more details.
13
14 You should have received a copy of the GNU Lesser General Public License
15 along with this program. If not, see <https://www.gnu.org/licenses/>. */
16
17 #include <config.h>
18
19 /* Specification. */
20 #include <iconv.h>
21
22 #include <errno.h>
23 #include <string.h>
24 #include "c-ctype.h"
25 #include "c-strcase.h"
26
27 #define SIZEOF(a) (sizeof(a) / sizeof(a[0]))
28
29 /* Namespace cleanliness. */
30 #define mapping_lookup rpl_iconv_open_mapping_lookup
31
32 /* The macro ICONV_FLAVOR is defined to one of these or undefined. */
33
34 #define ICONV_FLAVOR_AIX "iconv_open-aix.h"
35 #define ICONV_FLAVOR_HPUX "iconv_open-hpux.h"
36 #define ICONV_FLAVOR_IRIX "iconv_open-irix.h"
37 #define ICONV_FLAVOR_OSF "iconv_open-osf.h"
38 #define ICONV_FLAVOR_SOLARIS "iconv_open-solaris.h"
39 #define ICONV_FLAVOR_ZOS "iconv_open-zos.h"
40
41 #ifdef ICONV_FLAVOR
42 # include ICONV_FLAVOR
43 #endif
44
45 iconv_t
46 rpl_iconv_open (const char *tocode, const char *fromcode)
47 #undef iconv_open
48 {
49 char fromcode_upper[32];
50 char tocode_upper[32];
51 char *fromcode_upper_end;
52 char *tocode_upper_end;
53
54 #if REPLACE_ICONV_UTF
55 /* Special handling of conversion between UTF-8 and UTF-{16,32}{BE,LE}.
56 Do this here, before calling the real iconv_open(), because OSF/1 5.1
57 iconv() to these encoding inserts a BOM, which is wrong.
58 We do not need to handle conversion between arbitrary encodings and
59 UTF-{16,32}{BE,LE}, because the 'striconveh' module implements two-step
60 conversion through UTF-8.
61 The _ICONV_* constants are chosen to be disjoint from any iconv_t
62 returned by the system's iconv_open() functions. Recall that iconv_t
63 is a scalar type. */
64 if (c_toupper (fromcode[0]) == 'U'
65 && c_toupper (fromcode[1]) == 'T'
66 && c_toupper (fromcode[2]) == 'F'
67 && fromcode[3] == '-')
68 {
69 if (c_toupper (tocode[0]) == 'U'
70 && c_toupper (tocode[1]) == 'T'
71 && c_toupper (tocode[2]) == 'F'
72 && tocode[3] == '-')
73 {
74 if (strcmp (fromcode + 4, "8") == 0)
75 {
76 if (c_strcasecmp (tocode + 4, "16BE") == 0)
77 return _ICONV_UTF8_UTF16BE;
78 if (c_strcasecmp (tocode + 4, "16LE") == 0)
79 return _ICONV_UTF8_UTF16LE;
80 if (c_strcasecmp (tocode + 4, "32BE") == 0)
81 return _ICONV_UTF8_UTF32BE;
82 if (c_strcasecmp (tocode + 4, "32LE") == 0)
83 return _ICONV_UTF8_UTF32LE;
84 }
85 else if (strcmp (tocode + 4, "8") == 0)
86 {
87 if (c_strcasecmp (fromcode + 4, "16BE") == 0)
88 return _ICONV_UTF16BE_UTF8;
89 if (c_strcasecmp (fromcode + 4, "16LE") == 0)
90 return _ICONV_UTF16LE_UTF8;
91 if (c_strcasecmp (fromcode + 4, "32BE") == 0)
92 return _ICONV_UTF32BE_UTF8;
93 if (c_strcasecmp (fromcode + 4, "32LE") == 0)
94 return _ICONV_UTF32LE_UTF8;
95 }
96 }
97 }
98 #endif
99
100 /* Do *not* add special support for 8-bit encodings like ASCII or ISO-8859-1
101 here. This would lead to programs that work in some locales (such as the
102 "C" or "en_US" locales) but do not work in East Asian locales. It is
103 better if programmers make their programs depend on GNU libiconv (except
104 on glibc systems), e.g. by using the AM_ICONV macro and documenting the
105 dependency in an INSTALL or DEPENDENCIES file. */
106
107 /* Try with the original names first.
108 This covers the case when fromcode or tocode is a lowercase encoding name
109 that is understood by the system's iconv_open but not listed in our
110 mappings table. */
111 {
112 iconv_t cd = iconv_open (tocode, fromcode);
113 if (cd != (iconv_t)(-1))
114 return cd;
115 }
116
117 /* Convert the encodings to upper case, because
118 1. in the arguments of iconv_open() on AIX, HP-UX, and OSF/1 the case
119 matters,
120 2. it makes searching in the table faster. */
121 {
122 const char *p = fromcode;
123 char *q = fromcode_upper;
124 while ((*q = c_toupper (*p)) != '\0')
125 {
126 p++;
127 q++;
128 if (q == &fromcode_upper[SIZEOF (fromcode_upper)])
129 {
130 errno = EINVAL;
131 return (iconv_t)(-1);
132 }
133 }
134 fromcode_upper_end = q;
135 }
136
137 {
138 const char *p = tocode;
139 char *q = tocode_upper;
140 while ((*q = c_toupper (*p)) != '\0')
141 {
142 p++;
143 q++;
144 if (q == &tocode_upper[SIZEOF (tocode_upper)])
145 {
146 errno = EINVAL;
147 return (iconv_t)(-1);
148 }
149 }
150 tocode_upper_end = q;
151 }
152
153 #ifdef ICONV_FLAVOR
154 /* Apply the mappings. */
155 {
156 const struct mapping *m =
157 mapping_lookup (fromcode_upper, fromcode_upper_end - fromcode_upper);
158
159 fromcode = (m != NULL ? m->vendor_name : fromcode_upper);
160 }
161 {
162 const struct mapping *m =
163 mapping_lookup (tocode_upper, tocode_upper_end - tocode_upper);
164
165 tocode = (m != NULL ? m->vendor_name : tocode_upper);
166 }
167 #else
168 fromcode = fromcode_upper;
169 tocode = tocode_upper;
170 #endif
171
172 return iconv_open (tocode, fromcode);
173 }