2 * charset.h - header file for general character set conversion
6 #ifndef charset_charset_h
7 #define charset_charset_h
12 * Enumeration that lists all the multibyte or single-byte
13 * character sets known to this library.
16 CS_NONE
, /* used for reporting errors, etc */
17 CS_ASCII
, /* ordinary US-ASCII is worth having! */
19 CS_ISO8859_1_X11
, /* X font encoding with VT100 glyphs */
96 * This macro is used to initialise a charset_state structure:
98 * charset_state mystate = CHARSET_INIT_STATE;
100 #define CHARSET_INIT_STATE { 0L, 0L } /* a suitable initialiser */
103 * This external variable contains the same data, but is provided
104 * for easy structure-copy assignment:
106 * mystate = charset_init_state;
108 extern const charset_state charset_init_state
;
111 * Routine to convert a MB/SB character set to Unicode.
113 * This routine accepts some number of bytes, updates a state
114 * variable, and outputs some number of Unicode characters. There
115 * are no guarantees. You can't even guarantee that at most one
116 * Unicode character will be output per byte you feed in; for
117 * example, suppose you're reading UTF-8, you've seen E1 80, and
118 * then you suddenly see FE. Now you need to output _two_ error
119 * characters - one for the incomplete sequence E1 80, and one for
120 * the completely invalid UTF-8 byte FE.
122 * Returns the number of wide characters output; will never output
123 * more than the size of the buffer (as specified on input).
124 * Advances the `input' pointer and decrements `inlen', to indicate
125 * how far along the input string it got.
127 * The sequence of `errlen' wide characters pointed to by `errstr'
128 * will be used to indicate a conversion error. If `errstr' is
129 * NULL, `errlen' will be ignored, and the library will choose
130 * something sensible to do on its own. For Unicode, this will be
131 * U+FFFD (REPLACEMENT CHARACTER).
134 int charset_to_unicode(const char **input
, int *inlen
,
135 wchar_t *output
, int outlen
,
136 int charset
, charset_state
*state
,
137 const wchar_t *errstr
, int errlen
);
140 * Routine to convert Unicode to an MB/SB character set.
142 * This routine accepts some number of Unicode characters, updates
143 * a state variable, and outputs some number of bytes.
145 * Returns the number of bytes output; will never output more than
146 * the size of the buffer (as specified on input), and will never
147 * output a partial MB character. Advances the `input' pointer and
148 * decrements `inlen', to indicate how far along the input string
151 * If `error' is non-NULL and a character is found which cannot be
152 * expressed in the output charset, conversion will terminate at
153 * that character (so `input' points to the offending character)
154 * and `*error' will be set to TRUE; if `error' is non-NULL and no
155 * difficult characters are encountered, `*error' will be set to
156 * FALSE. If `error' is NULL, difficult characters will simply be
159 * If `input' is NULL, this routine will output the necessary bytes
160 * to reset the encoding state in any way which might be required
161 * at the end of an output piece of text.
164 int charset_from_unicode(const wchar_t **input
, int *inlen
,
165 char *output
, int outlen
,
166 int charset
, charset_state
*state
, int *error
);
169 * Convert X11 encoding names to and from our charset identifiers.
171 const char *charset_to_xenc(int charset
);
172 int charset_from_xenc(const char *name
);
175 * Convert MIME encoding names to and from our charset identifiers.
177 const char *charset_to_mimeenc(int charset
);
178 int charset_from_mimeenc(const char *name
);
181 * Convert our own encoding names to and from our charset
184 const char *charset_to_localenc(int charset
);
185 int charset_from_localenc(const char *name
);
186 int charset_localenc_nth(int n
);
189 * Convert Mac OS script/region/font to our charset identifiers.
191 int charset_from_macenc(int script
, int region
, int sysvers
,
192 const char *fontname
);
195 * Upgrade a charset identifier to a superset charset which is
196 * often confused with it. For example, people whose MUAs report
197 * their mail as ASCII or ISO8859-1 often in practice turn out to
198 * be using CP1252 quote characters, so when parsing incoming mail
199 * it is prudent to treat ASCII and ISO8859-1 as aliases for CP1252
200 * - and since it's a superset of both, this will cause no
201 * genuinely correct mail to be parsed wrongly.
203 int charset_upgrade(int charset
);
206 * This function returns TRUE if the input charset is a vaguely
207 * sensible superset of ASCII. That is, it returns FALSE for 7-bit
208 * encoding formats such as HZ and UTF-7.
210 int charset_contains_ascii(int charset
);
213 * This function tries to deduce the CS_* identifier of the charset
214 * used in the current C locale. It falls back to CS_ASCII if it
215 * can't figure it out at all, so it will always return a valid
218 * (Note that you should have already called setlocale(LC_CTYPE,
219 * "") to guarantee that this function will do the right thing.)
221 int charset_from_locale(void);
223 #endif /* charset_charset_h */