2 This file is part of GNUnet.
3 (C) 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2009 Christian Grothoff (and other contributing authors)
5 GNUnet is free software; you can redistribute it and/or modify
6 it under the terms of the GNU General Public License as published
7 by the Free Software Foundation; either version 2, or (at your
8 option) any later version.
10 GNUnet is distributed in the hope that it will be useful, but
11 WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 General Public License for more details.
15 You should have received a copy of the GNU General Public License
16 along with GNUnet; see the file COPYING. If not, write to the
17 Free Software Foundation, Inc., 59 Temple Place - Suite 330,
18 Boston, MA 02111-1307, USA.
22 * @file include/gnunet_strings_lib.h
23 * @brief strings and string handling functions (including malloc
24 * and string tokenizing)
26 * @author Christian Grothoff
27 * @author Krista Bennett
28 * @author Gerd Knorr <kraxel@bytesex.org>
29 * @author Ioana Patrascu
30 * @author Tzvetan Horozov
33 #ifndef GNUNET_STRINGS_LIB_H
34 #define GNUNET_STRINGS_LIB_H
36 /* we need size_t, and since it can be both unsigned int
37 or unsigned long long, this IS platform dependent;
38 but "stdlib.h" should be portable 'enough' to be
39 unconditionally available... */
45 #if 0 /* keep Emacsens' auto-indent happy */
50 #include "gnunet_time_lib.h"
54 * Convert a given fancy human-readable size to bytes.
56 * @param fancy_size human readable string (i.e. 1 MB)
57 * @param size set to the size in bytes
58 * @return GNUNET_OK on success, GNUNET_SYSERR on error
61 GNUNET_STRINGS_fancy_size_to_bytes (const char *fancy_size,
62 unsigned long long *size);
66 * Convert a given fancy human-readable time to our internal
69 * @param fancy_time human readable string (i.e. 1 minute)
70 * @param rtime set to the relative time
71 * @return GNUNET_OK on success, GNUNET_SYSERR on error
74 GNUNET_STRINGS_fancy_time_to_relative (const char *fancy_time,
75 struct GNUNET_TIME_Relative *rtime);
79 * Convert a given fancy human-readable time to our internal
82 * @param fancy_time human readable string (i.e. %Y-%m-%d %H:%M:%S)
83 * @param atime set to the absolute time
84 * @return GNUNET_OK on success, GNUNET_SYSERR on error
87 GNUNET_STRINGS_fancy_time_to_absolute (const char *fancy_time,
88 struct GNUNET_TIME_Absolute *atime);
92 * Convert a given filesize into a fancy human-readable format.
94 * @param size number of bytes
95 * @return fancy representation of the size (possibly rounded) for humans
98 GNUNET_STRINGS_byte_size_fancy (unsigned long long size);
102 * Convert the len characters long character sequence
103 * given in input that is in the given input charset
104 * to a string in given output charset.
105 * @return the converted string (0-terminated),
106 * if conversion fails, a copy of the orignal
107 * string is returned.
110 GNUNET_STRINGS_conv (const char *input, size_t len,
111 const char *input_charset, const char *output_charset);
114 * Convert the len characters long character sequence
115 * given in input that is in the given charset
118 * @param input the input string (not necessarily 0-terminated)
119 * @param len the number of bytes in the input
120 * @param charset character set to convert from
121 * @return the converted string (0-terminated)
124 GNUNET_STRINGS_to_utf8 (const char *input, size_t len, const char *charset);
127 * Convert the len bytes-long UTF-8 string
128 * given in input to the given charset.
130 * @return the converted string (0-terminated),
131 * if conversion fails, a copy of the orignal
132 * string is returned.
135 GNUNET_STRINGS_from_utf8 (const char *input, size_t len, const char *charset);
138 * Convert the utf-8 input string to lowercase
139 * Output needs to be allocated appropriately
141 * @param input input string
142 * @param output output buffer
145 GNUNET_STRINGS_utf8_tolower(const char* input, char** output);
149 * Convert the utf-8 input string to lowercase
150 * Output needs to be allocated appropriately
152 * @param input input string
153 * @param output output buffer
156 GNUNET_STRINGS_utf8_toupper(const char* input, char** output);
160 * Complete filename (a la shell) from abbrevition.
162 * @param fil the name of the file, may contain ~/ or
163 * be relative to the current directory
164 * @return the full file name,
165 * NULL is returned on error
168 GNUNET_STRINGS_filename_expand (const char *fil);
172 * Fill a buffer of the given size with
173 * count 0-terminated strings (given as varargs).
174 * If "buffer" is NULL, only compute the amount of
175 * space required (sum of "strlen(arg)+1").
177 * Unlike using "snprintf" with "%s", this function
178 * will add 0-terminators after each string. The
179 * "GNUNET_string_buffer_tokenize" function can be
180 * used to parse the buffer back into individual
183 * @param buffer the buffer to fill with strings, can
184 * be NULL in which case only the necessary
185 * amount of space will be calculated
186 * @param size number of bytes available in buffer
187 * @param count number of strings that follow
188 * @param ... count 0-terminated strings to copy to buffer
189 * @return number of bytes written to the buffer
190 * (or number of bytes that would have been written)
193 GNUNET_STRINGS_buffer_fill (char *buffer, size_t size, unsigned int count, ...);
197 * Given a buffer of a given size, find "count"
198 * 0-terminated strings in the buffer and assign
199 * the count (varargs) of type "const char**" to the
200 * locations of the respective strings in the
203 * @param buffer the buffer to parse
204 * @param size size of the buffer
205 * @param count number of strings to locate
206 * @param ... pointers to where to store the strings
207 * @return offset of the character after the last 0-termination
208 * in the buffer, or 0 on error.
211 GNUNET_STRINGS_buffer_tokenize (const char *buffer, size_t size,
212 unsigned int count, ...);
217 * "man ctime_r", except for GNUnet time; also, unlike ctime, the
218 * return value does not include the newline character.
220 * @param t the absolute time to convert
221 * @return timestamp in human-readable form
224 GNUNET_STRINGS_absolute_time_to_string (struct GNUNET_TIME_Absolute t);
228 * Give relative time in human-readable fancy format.
230 * @param delta time in milli seconds
231 * @return string in human-readable form
234 GNUNET_STRINGS_relative_time_to_string (struct GNUNET_TIME_Relative delta);
238 * Returns a pointer to a part of filename (allocates nothing)!
240 * @param filename filename to extract basename from
241 * @return short (base) name of the file (that is, everything following the
242 * last directory separator in filename. If filename ends with a
243 * directory separator, the result will be a zero-length string.
244 * If filename has no directory separators, the result is filename
248 GNUNET_STRINGS_get_short_name (const char *filename);
252 * Convert binary data to ASCII encoding. The ASCII encoding is rather
253 * GNUnet specific. It was chosen such that it only uses characters
254 * in [0-9A-V], can be produced without complex arithmetics and uses a
255 * small number of characters. The GNUnet encoding uses 103 characters.
256 * Does not append 0-terminator, but returns a pointer to the place where
257 * it should be placed, if needed.
259 * @param data data to encode
260 * @param size size of data (in bytes)
261 * @param out buffer to fill
262 * @param out_size size of the buffer. Must be large enough to hold
263 * ((size*8) + (((size*8) % 5) > 0 ? 5 - ((size*8) % 5) : 0)) / 5
264 * @return pointer to the next byte in 'out' or NULL on error.
267 GNUNET_STRINGS_data_to_string (const unsigned char *data, size_t size,
268 char *out, size_t out_size);
272 * Convert ASCII encoding back to data
273 * out_size must match exactly the size of the data before it was encoded.
275 * @param enc the encoding
276 * @param enclen number of characters in 'enc' (without 0-terminator, which can be missing)
277 * @param out location where to store the decoded data
278 * @param out_size sizeof the output buffer
279 * @return GNUNET_OK on success, GNUNET_SYSERR if result has the wrong encoding
282 GNUNET_STRINGS_string_to_data (const char *enc, size_t enclen,
283 unsigned char *out, size_t out_size);
286 #if 0 /* keep Emacsens' auto-indent happy */
293 enum GNUNET_STRINGS_FilenameCheck
295 GNUNET_STRINGS_CHECK_EXISTS = 0x00000001,
296 GNUNET_STRINGS_CHECK_IS_DIRECTORY = 0x00000002,
297 GNUNET_STRINGS_CHECK_IS_LINK = 0x00000004,
298 GNUNET_STRINGS_CHECK_IS_ABSOLUTE = 0x00000008
302 * Parse a path that might be an URI.
304 * @param path path to parse. Must be NULL-terminated.
305 * @param scheme_part a pointer to 'char *' where a pointer to a string that
306 * represents the URI scheme will be stored. Can be NULL. The string is
307 * allocated by the function, and should be freed by GNUNET_free() when
308 * it is no longer needed.
309 * @param path_part a pointer to 'const char *' where a pointer to the path
310 * part of the URI will be stored. Can be NULL. Points to the same block
311 * of memory as 'path', and thus must not be freed. Might point to '\0',
312 * if path part is zero-length.
313 * @return GNUNET_YES if it's an URI, GNUNET_NO otherwise. If 'path' is not
314 * an URI, '* scheme_part' and '*path_part' will remain unchanged
315 * (if they weren't NULL).
318 GNUNET_STRINGS_parse_uri (const char *path, char **scheme_part,
319 const char **path_part);
323 * Check whether filename is absolute or not, and if it's an URI
325 * @param filename filename to check
326 * @param can_be_uri GNUNET_YES to check for being URI, GNUNET_NO - to
327 * assume it's not URI
328 * @param r_is_uri a pointer to an int that is set to GNUNET_YES if 'filename'
329 * is URI and to GNUNET_NO otherwise. Can be NULL. If 'can_be_uri' is
330 * not GNUNET_YES, *r_is_uri is set to GNUNET_NO.
331 * @param r_uri_scheme a pointer to a char * that is set to a pointer to URI scheme.
332 * The string is allocated by the function, and should be freed with
333 * GNUNET_free (). Can be NULL.
334 * @return GNUNET_YES if 'filename' is absolute, GNUNET_NO otherwise.
337 GNUNET_STRINGS_path_is_absolute (const char *filename,
340 char **r_uri_scheme);
344 * Perform checks on 'filename;
346 * @param filename file to check
347 * @param checks checks to perform
348 * @return GNUNET_YES if all checks pass, GNUNET_NO if at least one of them
349 * fails, GNUNET_SYSERR when a check can't be performed
352 GNUNET_STRINGS_check_filename (const char *filename,
353 enum GNUNET_STRINGS_FilenameCheck checks);
357 * Tries to convert 'zt_addr' string to an IPv6 address.
358 * The string is expected to have the format "[ABCD::01]:80".
360 * @param zt_addr 0-terminated string. May be mangled by the function.
361 * @param addrlen length of zt_addr (not counting 0-terminator).
362 * @param r_buf a buffer to fill. Initially gets filled with zeroes,
363 * then its sin6_port, sin6_family and sin6_addr are set appropriately.
364 * @return GNUNET_OK if conversion succeded. GNUNET_SYSERR otherwise, in which
365 * case the contents of r_buf are undefined.
368 GNUNET_STRINGS_to_address_ipv6 (const char *zt_addr,
370 struct sockaddr_in6 *r_buf);
374 * Tries to convert 'zt_addr' string to an IPv4 address.
375 * The string is expected to have the format "1.2.3.4:80".
377 * @param zt_addr 0-terminated string. May be mangled by the function.
378 * @param addrlen length of zt_addr (not counting 0-terminator).
379 * @param r_buf a buffer to fill.
380 * @return GNUNET_OK if conversion succeded. GNUNET_SYSERR otherwise, in which case
381 * the contents of r_buf are undefined.
384 GNUNET_STRINGS_to_address_ipv4 (const char *zt_addr,
386 struct sockaddr_in *r_buf);
390 * Tries to convert 'addr' string to an IP (v4 or v6) address.
391 * Will automatically decide whether to treat 'addr' as v4 or v6 address.
393 * @param addr a string, may not be 0-terminated.
394 * @param addrlen number of bytes in addr (if addr is 0-terminated,
395 * 0-terminator should not be counted towards addrlen).
396 * @param r_buf a buffer to fill.
397 * @return GNUNET_OK if conversion succeded. GNUNET_SYSERR otherwise, in which
398 * case the contents of r_buf are undefined.
401 GNUNET_STRINGS_to_address_ip (const char *addr,
403 struct sockaddr_storage *r_buf);
407 * Returns utf-8 encoded arguments.
408 * Does nothing (returns a copy of argc and argv) on any platform
410 * Returned argv has u8argv[u8argc] == NULL.
411 * Returned argv is a single memory block, and can be freed with a single
412 * GNUNET_free () call.
414 * @param argc argc (as given by main())
415 * @param argv argv (as given by main())
416 * @param u8argc a location to store new argc in (though it's th same as argc)
417 * @param u8argv a location to store new argv in
418 * @return GNUNET_OK on success, GNUNET_SYSERR on failure
421 GNUNET_STRINGS_get_utf8_args (int argc, char *const *argv, int *u8argc,
422 char *const **u8argv);
424 /* ifndef GNUNET_UTIL_STRING_H */
426 /* end of gnunet_util_string.h */