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 */
100 * This macro is used to initialise a charset_state structure:
102 * charset_state mystate = CHARSET_INIT_STATE;
104 #define CHARSET_INIT_STATE { 0L, 0L } /* a suitable initialiser */
107 * This external variable contains the same data, but is provided
108 * for easy structure-copy assignment:
110 * mystate = charset_init_state;
112 extern const charset_state charset_init_state
;
115 * Routine to convert a MB/SB character set to Unicode.
117 * This routine accepts some number of bytes, updates a state
118 * variable, and outputs some number of Unicode characters. There
119 * are no guarantees. You can't even guarantee that at most one
120 * Unicode character will be output per byte you feed in; for
121 * example, suppose you're reading UTF-8, you've seen E1 80, and
122 * then you suddenly see FE. Now you need to output _two_ error
123 * characters - one for the incomplete sequence E1 80, and one for
124 * the completely invalid UTF-8 byte FE.
126 * Returns the number of wide characters output; will never output
127 * more than the size of the buffer (as specified on input).
128 * Advances the `input' pointer and decrements `inlen', to indicate
129 * how far along the input string it got.
131 * The sequence of `errlen' wide characters pointed to by `errstr'
132 * will be used to indicate a conversion error. If `errstr' is
133 * NULL, `errlen' will be ignored, and the library will choose
134 * something sensible to do on its own. For Unicode, this will be
135 * U+FFFD (REPLACEMENT CHARACTER).
138 int charset_to_unicode(const char **input
, int *inlen
,
139 wchar_t *output
, int outlen
,
140 int charset
, charset_state
*state
,
141 const wchar_t *errstr
, int errlen
);
144 * Routine to convert Unicode to an MB/SB character set.
146 * This routine accepts some number of Unicode characters, updates
147 * a state variable, and outputs some number of bytes.
149 * Returns the number of bytes output; will never output more than
150 * the size of the buffer (as specified on input), and will never
151 * output a partial MB character. Advances the `input' pointer and
152 * decrements `inlen', to indicate how far along the input string
155 * If `error' is non-NULL and a character is found which cannot be
156 * expressed in the output charset, conversion will terminate at
157 * that character (so `input' points to the offending character)
158 * and `*error' will be set to TRUE; if `error' is non-NULL and no
159 * difficult characters are encountered, `*error' will be set to
160 * FALSE. If `error' is NULL, difficult characters will simply be
163 * If `input' is NULL, this routine will output the necessary bytes
164 * to reset the encoding state in any way which might be required
165 * at the end of an output piece of text.
168 int charset_from_unicode(const wchar_t **input
, int *inlen
,
169 char *output
, int outlen
,
170 int charset
, charset_state
*state
, int *error
);
173 * Convert X11 encoding names to and from our charset identifiers.
175 const char *charset_to_xenc(int charset
);
176 int charset_from_xenc(const char *name
);
179 * Convert MIME encoding names to and from our charset identifiers.
181 const char *charset_to_mimeenc(int charset
);
182 int charset_from_mimeenc(const char *name
);
185 * Convert our own encoding names to and from our charset
188 const char *charset_to_localenc(int charset
);
189 int charset_from_localenc(const char *name
);
190 int charset_localenc_nth(int n
);
193 * Convert Mac OS script/region/font to our charset identifiers.
195 int charset_from_macenc(int script
, int region
, int sysvers
,
196 const char *fontname
);
199 * Upgrade a charset identifier to a superset charset which is
200 * often confused with it. For example, people whose MUAs report
201 * their mail as ASCII or ISO8859-1 often in practice turn out to
202 * be using CP1252 quote characters, so when parsing incoming mail
203 * it is prudent to treat ASCII and ISO8859-1 as aliases for CP1252
204 * - and since it's a superset of both, this will cause no
205 * genuinely correct mail to be parsed wrongly.
207 int charset_upgrade(int charset
);
210 * This function returns TRUE if the input charset is a vaguely
211 * sensible superset of ASCII. That is, it returns FALSE for 7-bit
212 * encoding formats such as HZ and UTF-7.
214 int charset_contains_ascii(int charset
);
217 * This function tries to deduce the CS_* identifier of the charset
218 * used in the current C locale. It falls back to CS_ASCII if it
219 * can't figure it out at all, so it will always return a valid
222 * (Note that you should have already called setlocale(LC_CTYPE,
223 * "") to guarantee that this function will do the right thing.)
225 int charset_from_locale(void);
227 #endif /* charset_charset_h */