2 This file is part of GNUnet.
3 (C) 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2009, 2012 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,
112 const char *output_charset);
116 * Convert the len characters long character sequence
117 * given in input that is in the given charset
120 * @param input the input string (not necessarily 0-terminated)
121 * @param len the number of bytes in the input
122 * @param charset character set to convert from
123 * @return the converted string (0-terminated)
126 GNUNET_STRINGS_to_utf8 (const char *input,
128 const char *charset);
132 * Convert the len bytes-long UTF-8 string
133 * given in input to the given charset.
135 * @return the converted string (0-terminated),
136 * if conversion fails, a copy of the orignal
137 * string is returned.
140 GNUNET_STRINGS_from_utf8 (const char *input,
142 const char *charset);
146 * Convert the utf-8 input string to lowercase
147 * Output needs to be allocated appropriately
149 * @param input input string
150 * @param output output buffer
153 GNUNET_STRINGS_utf8_tolower (const char* input,
158 * Convert the utf-8 input string to lowercase
159 * Output needs to be allocated appropriately
161 * @param input input string
162 * @param output output buffer
165 GNUNET_STRINGS_utf8_toupper (const char* input,
170 * Complete filename (a la shell) from abbrevition.
172 * @param fil the name of the file, may contain ~/ or
173 * be relative to the current directory
174 * @return the full file name,
175 * NULL is returned on error
178 GNUNET_STRINGS_filename_expand (const char *fil);
182 * Fill a buffer of the given size with count 0-terminated strings
183 * (given as varargs). If "buffer" is NULL, only compute the amount
184 * of space required (sum of "strlen(arg)+1").
186 * Unlike using "snprintf" with "%s", this function will add
187 * 0-terminators after each string. The
188 * "GNUNET_string_buffer_tokenize" function can be used to parse the
189 * buffer back into individual strings.
191 * @param buffer the buffer to fill with strings, can
192 * be NULL in which case only the necessary
193 * amount of space will be calculated
194 * @param size number of bytes available in buffer
195 * @param count number of strings that follow
196 * @param ... count 0-terminated strings to copy to buffer
197 * @return number of bytes written to the buffer
198 * (or number of bytes that would have been written)
201 GNUNET_STRINGS_buffer_fill (char *buffer,
208 * Given a buffer of a given size, find "count" 0-terminated strings
209 * in the buffer and assign the count (varargs) of type "const char**"
210 * to the locations of the respective strings in the buffer.
212 * @param buffer the buffer to parse
213 * @param size size of the buffer
214 * @param count number of strings to locate
215 * @param ... pointers to where to store the strings
216 * @return offset of the character after the last 0-termination
217 * in the buffer, or 0 on error.
220 GNUNET_STRINGS_buffer_tokenize (const char *buffer, size_t size,
221 unsigned int count, ...);
226 * "asctime", except for GNUnet time.
228 * @param t the absolute time to convert
229 * @return timestamp in human-readable form
232 GNUNET_STRINGS_absolute_time_to_string (struct GNUNET_TIME_Absolute t);
236 * Give relative time in human-readable fancy format.
238 * @param delta time in milli seconds
239 * @return string in human-readable form
242 GNUNET_STRINGS_relative_time_to_string (struct GNUNET_TIME_Relative delta);
247 * Returns a pointer to a part of filename (allocates nothing)!
249 * @param filename filename to extract basename from
250 * @return short (base) name of the file (that is, everything following the
251 * last directory separator in filename. If filename ends with a
252 * directory separator, the result will be a zero-length string.
253 * If filename has no directory separators, the result is filename
257 GNUNET_STRINGS_get_short_name (const char *filename);
261 * Convert binary data to ASCII encoding. The ASCII encoding is rather
262 * GNUnet specific. It was chosen such that it only uses characters
263 * in [0-9A-V], can be produced without complex arithmetics and uses a
264 * small number of characters. The GNUnet encoding uses 103 characters.
265 * Does not append 0-terminator, but returns a pointer to the place where
266 * it should be placed, if needed.
268 * @param data data to encode
269 * @param size size of data (in bytes)
270 * @param out buffer to fill
271 * @param out_size size of the buffer. Must be large enough to hold
272 * ((size*8) + (((size*8) % 5) > 0 ? 5 - ((size*8) % 5) : 0)) / 5
273 * @return pointer to the next byte in 'out' or NULL on error.
276 GNUNET_STRINGS_data_to_string (const unsigned char *data,
283 * Convert ASCII encoding back to data
284 * out_size must match exactly the size of the data before it was encoded.
286 * @param enc the encoding
287 * @param enclen number of characters in 'enc' (without 0-terminator, which can be missing)
288 * @param out location where to store the decoded data
289 * @param out_size sizeof the output buffer
290 * @return GNUNET_OK on success, GNUNET_SYSERR if result has the wrong encoding
293 GNUNET_STRINGS_string_to_data (const char *enc,
300 * Parse a path that might be an URI.
302 * @param path path to parse. Must be NULL-terminated.
303 * @param scheme_part a pointer to 'char *' where a pointer to a string that
304 * represents the URI scheme will be stored. Can be NULL. The string is
305 * allocated by the function, and should be freed by GNUNET_free() when
306 * it is no longer needed.
307 * @param path_part a pointer to 'const char *' where a pointer to the path
308 * part of the URI will be stored. Can be NULL. Points to the same block
309 * of memory as 'path', and thus must not be freed. Might point to '\0',
310 * if path part is zero-length.
311 * @return GNUNET_YES if it's an URI, GNUNET_NO otherwise. If 'path' is not
312 * an URI, '* scheme_part' and '*path_part' will remain unchanged
313 * (if they weren't NULL).
316 GNUNET_STRINGS_parse_uri (const char *path,
318 const char **path_part);
322 * Check whether filename is absolute or not, and if it's an URI
324 * @param filename filename to check
325 * @param can_be_uri GNUNET_YES to check for being URI, GNUNET_NO - to
326 * assume it's not URI
327 * @param r_is_uri a pointer to an int that is set to GNUNET_YES if 'filename'
328 * is URI and to GNUNET_NO otherwise. Can be NULL. If 'can_be_uri' is
329 * not GNUNET_YES, *r_is_uri is set to GNUNET_NO.
330 * @param r_uri_scheme a pointer to a char * that is set to a pointer to URI scheme.
331 * The string is allocated by the function, and should be freed with
332 * GNUNET_free (). Can be NULL.
333 * @return GNUNET_YES if 'filename' is absolute, GNUNET_NO otherwise.
336 GNUNET_STRINGS_path_is_absolute (const char *filename,
339 char **r_uri_scheme);
343 * Flags for what we should check a file for.
345 enum GNUNET_STRINGS_FilenameCheck
348 * Check that it exists.
350 GNUNET_STRINGS_CHECK_EXISTS = 0x00000001,
353 * Check that it is a directory.
355 GNUNET_STRINGS_CHECK_IS_DIRECTORY = 0x00000002,
358 * Check that it is a link.
360 GNUNET_STRINGS_CHECK_IS_LINK = 0x00000004,
363 * Check that the path is an absolute path.
365 GNUNET_STRINGS_CHECK_IS_ABSOLUTE = 0x00000008
370 * Perform checks on 'filename'. FIXME: some duplication with
371 * "GNUNET_DISK_"-APIs. We should unify those.
373 * @param filename file to check
374 * @param checks checks to perform
375 * @return GNUNET_YES if all checks pass, GNUNET_NO if at least one of them
376 * fails, GNUNET_SYSERR when a check can't be performed
379 GNUNET_STRINGS_check_filename (const char *filename,
380 enum GNUNET_STRINGS_FilenameCheck checks);
384 * Tries to convert 'zt_addr' string to an IPv6 address.
385 * The string is expected to have the format "[ABCD::01]:80".
387 * @param zt_addr 0-terminated string. May be mangled by the function.
388 * @param addrlen length of zt_addr (not counting 0-terminator).
389 * @param r_buf a buffer to fill. Initially gets filled with zeroes,
390 * then its sin6_port, sin6_family and sin6_addr are set appropriately.
391 * @return GNUNET_OK if conversion succeded. GNUNET_SYSERR otherwise, in which
392 * case the contents of r_buf are undefined.
395 GNUNET_STRINGS_to_address_ipv6 (const char *zt_addr,
397 struct sockaddr_in6 *r_buf);
401 * Tries to convert 'zt_addr' string to an IPv4 address.
402 * The string is expected to have the format "1.2.3.4:80".
404 * @param zt_addr 0-terminated string. May be mangled by the function.
405 * @param addrlen length of zt_addr (not counting 0-terminator).
406 * @param r_buf a buffer to fill.
407 * @return GNUNET_OK if conversion succeded. GNUNET_SYSERR otherwise, in which case
408 * the contents of r_buf are undefined.
411 GNUNET_STRINGS_to_address_ipv4 (const char *zt_addr,
413 struct sockaddr_in *r_buf);
417 * Tries to convert 'addr' string to an IP (v4 or v6) address.
418 * Will automatically decide whether to treat 'addr' as v4 or v6 address.
420 * @param addr a string, may not be 0-terminated.
421 * @param addrlen number of bytes in addr (if addr is 0-terminated,
422 * 0-terminator should not be counted towards addrlen).
423 * @param r_buf a buffer to fill.
424 * @return GNUNET_OK if conversion succeded. GNUNET_SYSERR otherwise, in which
425 * case the contents of r_buf are undefined.
428 GNUNET_STRINGS_to_address_ip (const char *addr,
430 struct sockaddr_storage *r_buf);
434 * Returns utf-8 encoded arguments.
435 * Does nothing (returns a copy of argc and argv) on any platform
437 * Returned argv has u8argv[u8argc] == NULL.
438 * Returned argv is a single memory block, and can be freed with a single
439 * GNUNET_free () call.
441 * @param argc argc (as given by main())
442 * @param argv argv (as given by main())
443 * @param u8argc a location to store new argc in (though it's th same as argc)
444 * @param u8argv a location to store new argv in
445 * @return GNUNET_OK on success, GNUNET_SYSERR on failure
448 GNUNET_STRINGS_get_utf8_args (int argc,
451 char *const **u8argv);
454 #if 0 /* keep Emacsens' auto-indent happy */
462 /* ifndef GNUNET_UTIL_STRING_H */
464 /* end of gnunet_util_string.h */