i18n: make GETTEXT_POISON a runtime option
[git/git.git] / gettext.c
1 /*
2 * Copyright (c) 2010 Ævar Arnfjörð Bjarmason
3 */
4
5 #include "cache.h"
6 #include "exec-cmd.h"
7 #include "gettext.h"
8 #include "strbuf.h"
9 #include "utf8.h"
10 #include "config.h"
11
12 #ifndef NO_GETTEXT
13 # include <locale.h>
14 # include <libintl.h>
15 # ifdef HAVE_LIBCHARSET_H
16 # include <libcharset.h>
17 # else
18 # include <langinfo.h>
19 # define locale_charset() nl_langinfo(CODESET)
20 # endif
21 #endif
22
23 static const char *charset;
24
25 /*
26 * Guess the user's preferred languages from the value in LANGUAGE environment
27 * variable and LC_MESSAGES locale category if NO_GETTEXT is not defined.
28 *
29 * The result can be a colon-separated list like "ko:ja:en".
30 */
31 const char *get_preferred_languages(void)
32 {
33 const char *retval;
34
35 retval = getenv("LANGUAGE");
36 if (retval && *retval)
37 return retval;
38
39 #ifndef NO_GETTEXT
40 retval = setlocale(LC_MESSAGES, NULL);
41 if (retval && *retval &&
42 strcmp(retval, "C") &&
43 strcmp(retval, "POSIX"))
44 return retval;
45 #endif
46
47 return NULL;
48 }
49
50 int use_gettext_poison(void)
51 {
52 static int poison_requested = -1;
53 if (poison_requested == -1) {
54 const char *v = getenv("GIT_TEST_GETTEXT_POISON");
55 poison_requested = v && strlen(v) ? 1 : 0;
56 }
57 return poison_requested;
58 }
59
60 #ifndef NO_GETTEXT
61 static int test_vsnprintf(const char *fmt, ...)
62 {
63 char buf[26];
64 int ret;
65 va_list ap;
66 va_start(ap, fmt);
67 ret = vsnprintf(buf, sizeof(buf), fmt, ap);
68 va_end(ap);
69 return ret;
70 }
71
72 static void init_gettext_charset(const char *domain)
73 {
74 /*
75 This trick arranges for messages to be emitted in the user's
76 requested encoding, but avoids setting LC_CTYPE from the
77 environment for the whole program.
78
79 This primarily done to avoid a bug in vsnprintf in the GNU C
80 Library [1]. which triggered a "your vsnprintf is broken" error
81 on Git's own repository when inspecting v0.99.6~1 under a UTF-8
82 locale.
83
84 That commit contains a ISO-8859-1 encoded author name, which
85 the locale aware vsnprintf(3) won't interpolate in the format
86 argument, due to mismatch between the data encoding and the
87 locale.
88
89 Even if it wasn't for that bug we wouldn't want to use LC_CTYPE at
90 this point, because it'd require auditing all the code that uses C
91 functions whose semantics are modified by LC_CTYPE.
92
93 But only setting LC_MESSAGES as we do creates a problem, since
94 we declare the encoding of our PO files[2] the gettext
95 implementation will try to recode it to the user's locale, but
96 without LC_CTYPE it'll emit something like this on 'git init'
97 under the Icelandic locale:
98
99 Bj? til t?ma Git lind ? /hlagh/.git/
100
101 Gettext knows about the encoding of our PO file, but we haven't
102 told it about the user's encoding, so all the non-US-ASCII
103 characters get encoded to question marks.
104
105 But we're in luck! We can set LC_CTYPE from the environment
106 only while we call nl_langinfo and
107 bind_textdomain_codeset. That suffices to tell gettext what
108 encoding it should emit in, so it'll now say:
109
110 Bjó til tóma Git lind í /hlagh/.git/
111
112 And the equivalent ISO-8859-1 string will be emitted under a
113 ISO-8859-1 locale.
114
115 With this change way we get the advantages of setting LC_CTYPE
116 (talk to the user in his language/encoding), without the major
117 drawbacks (changed semantics for C functions we rely on).
118
119 However foreign functions using other message catalogs that
120 aren't using our neat trick will still have a problem, e.g. if
121 we have to call perror(3):
122
123 #include <stdio.h>
124 #include <locale.h>
125 #include <errno.h>
126
127 int main(void)
128 {
129 setlocale(LC_MESSAGES, "");
130 setlocale(LC_CTYPE, "C");
131 errno = ENODEV;
132 perror("test");
133 return 0;
134 }
135
136 Running that will give you a message with question marks:
137
138 $ LANGUAGE= LANG=de_DE.utf8 ./test
139 test: Kein passendes Ger?t gefunden
140
141 The vsnprintf bug has been fixed since glibc 2.17.
142
143 Then we could simply set LC_CTYPE from the environment, which would
144 make things like the external perror(3) messages work.
145
146 See t/t0203-gettext-setlocale-sanity.sh's "gettext.c" tests for
147 regression tests.
148
149 1. http://sourceware.org/bugzilla/show_bug.cgi?id=6530
150 2. E.g. "Content-Type: text/plain; charset=UTF-8\n" in po/is.po
151 */
152 setlocale(LC_CTYPE, "");
153 charset = locale_charset();
154 bind_textdomain_codeset(domain, charset);
155 /* the string is taken from v0.99.6~1 */
156 if (test_vsnprintf("%.*s", 13, "David_K\345gedal") < 0)
157 setlocale(LC_CTYPE, "C");
158 }
159
160 void git_setup_gettext(void)
161 {
162 const char *podir = getenv(GIT_TEXT_DOMAIN_DIR_ENVIRONMENT);
163 char *p = NULL;
164
165 if (!podir)
166 podir = p = system_path(GIT_LOCALE_PATH);
167
168 use_gettext_poison(); /* getenv() reentrancy paranoia */
169
170 if (!is_directory(podir)) {
171 free(p);
172 return;
173 }
174
175 bindtextdomain("git", podir);
176 setlocale(LC_MESSAGES, "");
177 setlocale(LC_TIME, "");
178 init_gettext_charset("git");
179 textdomain("git");
180
181 free(p);
182 }
183
184 /* return the number of columns of string 's' in current locale */
185 int gettext_width(const char *s)
186 {
187 static int is_utf8 = -1;
188 if (is_utf8 == -1)
189 is_utf8 = is_utf8_locale();
190
191 return is_utf8 ? utf8_strwidth(s) : strlen(s);
192 }
193 #endif
194
195 int is_utf8_locale(void)
196 {
197 #ifdef NO_GETTEXT
198 if (!charset) {
199 const char *env = getenv("LC_ALL");
200 if (!env || !*env)
201 env = getenv("LC_CTYPE");
202 if (!env || !*env)
203 env = getenv("LANG");
204 if (!env)
205 env = "";
206 if (strchr(env, '.'))
207 env = strchr(env, '.') + 1;
208 charset = xstrdup(env);
209 }
210 #endif
211 return is_encoding_utf8(charset);
212 }