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 3, 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
80 * representation. The human-readable time is expected to be
81 * in local time, whereas the returned value will be in UTC.
83 * @param fancy_time human readable string (i.e. %Y-%m-%d %H:%M:%S)
84 * @param atime set to the absolute time
85 * @return #GNUNET_OK on success, #GNUNET_SYSERR on error
88 GNUNET_STRINGS_fancy_time_to_absolute (const char *fancy_time,
89 struct GNUNET_TIME_Absolute *atime);
93 * Convert a given filesize into a fancy human-readable format.
95 * @param size number of bytes
96 * @return fancy representation of the size (possibly rounded) for humans
99 GNUNET_STRINGS_byte_size_fancy (unsigned long long size);
103 * Convert the len characters long character sequence
104 * given in input that is in the given input charset
105 * to a string in given output charset.
106 * @return the converted string (0-terminated),
107 * if conversion fails, a copy of the orignal
108 * string is returned.
111 GNUNET_STRINGS_conv (const char *input, size_t len,
112 const char *input_charset,
113 const char *output_charset);
117 * Convert the len characters long character sequence
118 * given in input that is in the given charset
121 * @param input the input string (not necessarily 0-terminated)
122 * @param len the number of bytes in the input
123 * @param charset character set to convert from
124 * @return the converted string (0-terminated)
127 GNUNET_STRINGS_to_utf8 (const char *input,
129 const char *charset);
133 * Convert the len bytes-long UTF-8 string
134 * given in input to the given charset.
136 * @return the converted string (0-terminated),
137 * if conversion fails, a copy of the orignal
138 * string is returned.
141 GNUNET_STRINGS_from_utf8 (const char *input,
143 const char *charset);
147 * Convert the utf-8 input string to lowercase
148 * Output needs to be allocated appropriately
150 * @param input input string
151 * @param output output buffer
154 GNUNET_STRINGS_utf8_tolower (const char* input,
159 * Convert the utf-8 input string to lowercase
160 * Output needs to be allocated appropriately
162 * @param input input string
163 * @param output output buffer
166 GNUNET_STRINGS_utf8_toupper (const char* input,
171 * Complete filename (a la shell) from abbrevition.
173 * @param fil the name of the file, may contain ~/ or
174 * be relative to the current directory
175 * @return the full file name,
176 * NULL is returned on error
179 GNUNET_STRINGS_filename_expand (const char *fil);
183 * Fill a buffer of the given size with count 0-terminated strings
184 * (given as varargs). If "buffer" is NULL, only compute the amount
185 * of space required (sum of "strlen(arg)+1").
187 * Unlike using "snprintf" with "%s", this function will add
188 * 0-terminators after each string. The
189 * "GNUNET_string_buffer_tokenize" function can be used to parse the
190 * buffer back into individual strings.
192 * @param buffer the buffer to fill with strings, can
193 * be NULL in which case only the necessary
194 * amount of space will be calculated
195 * @param size number of bytes available in buffer
196 * @param count number of strings that follow
197 * @param ... count 0-terminated strings to copy to buffer
198 * @return number of bytes written to the buffer
199 * (or number of bytes that would have been written)
202 GNUNET_STRINGS_buffer_fill (char *buffer,
209 * Given a buffer of a given size, find "count" 0-terminated strings
210 * in the buffer and assign the count (varargs) of type "const char**"
211 * to the locations of the respective strings in the buffer.
213 * @param buffer the buffer to parse
214 * @param size size of the buffer
215 * @param count number of strings to locate
216 * @param ... pointers to where to store the strings
217 * @return offset of the character after the last 0-termination
218 * in the buffer, or 0 on error.
221 GNUNET_STRINGS_buffer_tokenize (const char *buffer, size_t size,
222 unsigned int count, ...);
227 * "asctime", except for GNUnet time. Converts a GNUnet internal
228 * absolute time (which is in UTC) to a string in local time.
229 * This is one of the very few calls in the entire API that is
232 * @param t the absolute time to convert
233 * @return timestamp in human-readable form in local time
236 GNUNET_STRINGS_absolute_time_to_string (struct GNUNET_TIME_Absolute t);
240 * Give relative time in human-readable fancy format.
241 * This is one of the very few calls in the entire API that is
244 * @param delta time in milli seconds
245 * @param do_round are we allowed to round a bit?
246 * @return string in human-readable form
249 GNUNET_STRINGS_relative_time_to_string (struct GNUNET_TIME_Relative delta,
255 * Returns a pointer to a part of filename (allocates nothing)!
257 * @param filename filename to extract basename from
258 * @return short (base) name of the file (that is, everything following the
259 * last directory separator in filename. If filename ends with a
260 * directory separator, the result will be a zero-length string.
261 * If filename has no directory separators, the result is filename
265 GNUNET_STRINGS_get_short_name (const char *filename);
269 * Convert binary data to ASCII encoding. The ASCII encoding is rather
270 * GNUnet specific. It was chosen such that it only uses characters
271 * in [0-9A-V], can be produced without complex arithmetics and uses a
272 * small number of characters. The GNUnet encoding uses 103 characters.
273 * Does not append 0-terminator, but returns a pointer to the place where
274 * it should be placed, if needed.
276 * @param data data to encode
277 * @param size size of data (in bytes)
278 * @param out buffer to fill
279 * @param out_size size of the buffer. Must be large enough to hold
280 * ((size*8) + (((size*8) % 5) > 0 ? 5 - ((size*8) % 5) : 0)) / 5
281 * @return pointer to the next byte in 'out' or NULL on error.
284 GNUNET_STRINGS_data_to_string (const void *data,
291 * Convert ASCII encoding back to data
292 * out_size must match exactly the size of the data before it was encoded.
294 * @param enc the encoding
295 * @param enclen number of characters in 'enc' (without 0-terminator, which can be missing)
296 * @param out location where to store the decoded data
297 * @param out_size size of the output buffer
298 * @return GNUNET_OK on success, GNUNET_SYSERR if result has the wrong encoding
301 GNUNET_STRINGS_string_to_data (const char *enc,
308 * Parse a path that might be an URI.
310 * @param path path to parse. Must be NULL-terminated.
311 * @param scheme_part a pointer to 'char *' where a pointer to a string that
312 * represents the URI scheme will be stored. Can be NULL. The string is
313 * allocated by the function, and should be freed by GNUNET_free() when
314 * it is no longer needed.
315 * @param path_part a pointer to 'const char *' where a pointer to the path
316 * part of the URI will be stored. Can be NULL. Points to the same block
317 * of memory as 'path', and thus must not be freed. Might point to '\0',
318 * if path part is zero-length.
319 * @return GNUNET_YES if it's an URI, GNUNET_NO otherwise. If 'path' is not
320 * an URI, '* scheme_part' and '*path_part' will remain unchanged
321 * (if they weren't NULL).
324 GNUNET_STRINGS_parse_uri (const char *path,
326 const char **path_part);
330 * Check whether filename is absolute or not, and if it's an URI
332 * @param filename filename to check
333 * @param can_be_uri GNUNET_YES to check for being URI, GNUNET_NO - to
334 * assume it's not URI
335 * @param r_is_uri a pointer to an int that is set to GNUNET_YES if 'filename'
336 * is URI and to GNUNET_NO otherwise. Can be NULL. If 'can_be_uri' is
337 * not GNUNET_YES, *r_is_uri is set to GNUNET_NO.
338 * @param r_uri_scheme a pointer to a char * that is set to a pointer to URI scheme.
339 * The string is allocated by the function, and should be freed with
340 * GNUNET_free (). Can be NULL.
341 * @return GNUNET_YES if 'filename' is absolute, GNUNET_NO otherwise.
344 GNUNET_STRINGS_path_is_absolute (const char *filename,
347 char **r_uri_scheme);
351 * Flags for what we should check a file for.
353 enum GNUNET_STRINGS_FilenameCheck
356 * Check that it exists.
358 GNUNET_STRINGS_CHECK_EXISTS = 0x00000001,
361 * Check that it is a directory.
363 GNUNET_STRINGS_CHECK_IS_DIRECTORY = 0x00000002,
366 * Check that it is a link.
368 GNUNET_STRINGS_CHECK_IS_LINK = 0x00000004,
371 * Check that the path is an absolute path.
373 GNUNET_STRINGS_CHECK_IS_ABSOLUTE = 0x00000008
378 * Perform checks on 'filename'. FIXME: some duplication with
379 * "GNUNET_DISK_"-APIs. We should unify those.
381 * @param filename file to check
382 * @param checks checks to perform
383 * @return GNUNET_YES if all checks pass, GNUNET_NO if at least one of them
384 * fails, GNUNET_SYSERR when a check can't be performed
387 GNUNET_STRINGS_check_filename (const char *filename,
388 enum GNUNET_STRINGS_FilenameCheck checks);
392 * Tries to convert 'zt_addr' string to an IPv6 address.
393 * The string is expected to have the format "[ABCD::01]:80".
395 * @param zt_addr 0-terminated string. May be mangled by the function.
396 * @param addrlen length of zt_addr (not counting 0-terminator).
397 * @param r_buf a buffer to fill. Initially gets filled with zeroes,
398 * then its sin6_port, sin6_family and sin6_addr are set appropriately.
399 * @return GNUNET_OK if conversion succeded. GNUNET_SYSERR otherwise, in which
400 * case the contents of r_buf are undefined.
403 GNUNET_STRINGS_to_address_ipv6 (const char *zt_addr,
405 struct sockaddr_in6 *r_buf);
409 * Tries to convert 'zt_addr' string to an IPv4 address.
410 * The string is expected to have the format "1.2.3.4:80".
412 * @param zt_addr 0-terminated string. May be mangled by the function.
413 * @param addrlen length of zt_addr (not counting 0-terminator).
414 * @param r_buf a buffer to fill.
415 * @return GNUNET_OK if conversion succeded. GNUNET_SYSERR otherwise, in which case
416 * the contents of r_buf are undefined.
419 GNUNET_STRINGS_to_address_ipv4 (const char *zt_addr,
421 struct sockaddr_in *r_buf);
425 * Tries to convert 'addr' string to an IP (v4 or v6) address.
426 * Will automatically decide whether to treat 'addr' as v4 or v6 address.
428 * @param addr a string, may not be 0-terminated.
429 * @param addrlen number of bytes in addr (if addr is 0-terminated,
430 * 0-terminator should not be counted towards addrlen).
431 * @param r_buf a buffer to fill.
432 * @return GNUNET_OK if conversion succeded. GNUNET_SYSERR otherwise, in which
433 * case the contents of r_buf are undefined.
436 GNUNET_STRINGS_to_address_ip (const char *addr,
438 struct sockaddr_storage *r_buf);
442 * Returns utf-8 encoded arguments.
443 * Does nothing (returns a copy of argc and argv) on any platform
445 * Returned argv has u8argv[u8argc] == NULL.
446 * Returned argv is a single memory block, and can be freed with a single
447 * GNUNET_free () call.
449 * @param argc argc (as given by main())
450 * @param argv argv (as given by main())
451 * @param u8argc a location to store new argc in (though it's th same as argc)
452 * @param u8argv a location to store new argv in
453 * @return GNUNET_OK on success, GNUNET_SYSERR on failure
456 GNUNET_STRINGS_get_utf8_args (int argc,
459 char *const **u8argv);
462 #if 0 /* keep Emacsens' auto-indent happy */
470 /* ifndef GNUNET_UTIL_STRING_H */
472 /* end of gnunet_util_string.h */