2 This file is part of GNUnet.
3 Copyright (C) 2001-2013 GNUnet e.V.
5 GNUnet is free software: you can redistribute it and/or modify it
6 under the terms of the GNU Affero General Public License as published
7 by the Free Software Foundation, either version 3 of the License,
8 or (at your 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 Affero General Public License for more details.
15 You should have received a copy of the GNU Affero General Public License
16 along with this program. If not, see <http://www.gnu.org/licenses/>.
18 SPDX-License-Identifier: AGPL3.0-or-later
22 * @author Christian Grothoff
23 * @author Krista Bennett
24 * @author Gerd Knorr <kraxel@bytesex.org>
25 * @author Ioana Patrascu
26 * @author Tzvetan Horozov
29 * Strings and string handling functions
31 * @defgroup strings Strings library
32 * Strings and string handling functions, including malloc and string tokenizing.
36 #ifndef GNUNET_STRINGS_LIB_H
37 #define GNUNET_STRINGS_LIB_H
39 /* we need size_t, and since it can be both unsigned int
40 or unsigned long long, this IS platform dependent;
41 but "stdlib.h" should be portable 'enough' to be
42 unconditionally available... */
44 #include <sys/socket.h>
45 #include <netinet/in.h>
46 #include <netinet/ip.h>
51 #if 0 /* keep Emacsens' auto-indent happy */
56 #include "gnunet_time_lib.h"
60 * Convert a given fancy human-readable size to bytes.
62 * @param fancy_size human readable string (i.e. 1 MB)
63 * @param size set to the size in bytes
64 * @return #GNUNET_OK on success, #GNUNET_SYSERR on error
67 GNUNET_STRINGS_fancy_size_to_bytes (const char *fancy_size,
68 unsigned long long *size);
72 * Convert a given fancy human-readable time to our internal
75 * @param fancy_time human readable string (i.e. 1 minute)
76 * @param rtime set to the relative time
77 * @return #GNUNET_OK on success, #GNUNET_SYSERR on error
80 GNUNET_STRINGS_fancy_time_to_relative (const char *fancy_time,
81 struct GNUNET_TIME_Relative *rtime);
86 * Convert a given fancy human-readable time to our internal
87 * representation. The human-readable time is expected to be
88 * in local time, whereas the returned value will be in UTC.
90 * @param fancy_time human readable string (i.e. %Y-%m-%d %H:%M:%S)
91 * @param atime set to the absolute time
92 * @return #GNUNET_OK on success, #GNUNET_SYSERR on error
95 GNUNET_STRINGS_fancy_time_to_absolute (const char *fancy_time,
96 struct GNUNET_TIME_Absolute *atime);
100 * Convert a given filesize into a fancy human-readable format.
102 * @param size number of bytes
103 * @return fancy representation of the size (possibly rounded) for humans
106 GNUNET_STRINGS_byte_size_fancy (unsigned long long size);
110 * Convert the len characters long character sequence
111 * given in input that is in the given input charset
112 * to a string in given output charset.
114 * @param input input string
115 * @param len number of bytes in @a input
116 * @param input_charset character set used for @a input
117 * @param output_charset desired character set for the return value
118 * @return the converted string (0-terminated),
119 * if conversion fails, a copy of the orignal
120 * string is returned.
123 GNUNET_STRINGS_conv (const char *input, size_t len,
124 const char *input_charset,
125 const char *output_charset);
129 * Convert the len characters long character sequence
130 * given in input that is in the given charset
133 * @param input the input string (not necessarily 0-terminated)
134 * @param len the number of bytes in the @a input
135 * @param charset character set to convert from
136 * @return the converted string (0-terminated)
139 GNUNET_STRINGS_to_utf8 (const char *input,
141 const char *charset);
145 * Convert the len bytes-long UTF-8 string
146 * given in input to the given charset.
148 * @param input the input string (not necessarily 0-terminated)
149 * @param len the number of bytes in the @a input
150 * @param charset character set to convert to
151 * @return the converted string (0-terminated),
152 * if conversion fails, a copy of the orignal
153 * string is returned.
156 GNUNET_STRINGS_from_utf8 (const char *input,
158 const char *charset);
162 * Convert the utf-8 input string to lower case.
163 * Output needs to be allocated appropriately.
165 * @param input input string
166 * @param output output buffer
169 GNUNET_STRINGS_utf8_tolower (const char *input,
174 * Convert the utf-8 input string to upper case.
175 * Output needs to be allocated appropriately.
177 * @param input input string
178 * @param output output buffer
181 GNUNET_STRINGS_utf8_toupper (const char *input,
186 * Complete filename (a la shell) from abbrevition.
188 * @param fil the name of the file, may contain ~/ or
189 * be relative to the current directory
190 * @return the full file name,
191 * NULL is returned on error
194 GNUNET_STRINGS_filename_expand (const char *fil);
198 * Fill a buffer of the given size with count 0-terminated strings
199 * (given as varargs). If "buffer" is NULL, only compute the amount
200 * of space required (sum of "strlen(arg)+1").
202 * Unlike using "snprintf" with "%s", this function will add
203 * 0-terminators after each string. The
204 * "GNUNET_string_buffer_tokenize" function can be used to parse the
205 * buffer back into individual strings.
207 * @param buffer the buffer to fill with strings, can
208 * be NULL in which case only the necessary
209 * amount of space will be calculated
210 * @param size number of bytes available in buffer
211 * @param count number of strings that follow
212 * @param ... count 0-terminated strings to copy to buffer
213 * @return number of bytes written to the buffer
214 * (or number of bytes that would have been written)
217 GNUNET_STRINGS_buffer_fill (char *buffer,
224 * Given a buffer of a given size, find "count" 0-terminated strings
225 * in the buffer and assign the count (varargs) of type "const char**"
226 * to the locations of the respective strings in the buffer.
228 * @param buffer the buffer to parse
229 * @param size size of the @a buffer
230 * @param count number of strings to locate
231 * @param ... pointers to where to store the strings
232 * @return offset of the character after the last 0-termination
233 * in the buffer, or 0 on error.
236 GNUNET_STRINGS_buffer_tokenize (const char *buffer,
238 unsigned int count, ...);
243 * Like `asctime`, except for GNUnet time. Converts a GNUnet internal
244 * absolute time (which is in UTC) to a string in local time.
245 * Note that the returned value will be overwritten if this function
248 * @param t the absolute time to convert
249 * @return timestamp in human-readable form in local time
252 GNUNET_STRINGS_absolute_time_to_string (struct GNUNET_TIME_Absolute t);
257 * Give relative time in human-readable fancy format.
258 * This is one of the very few calls in the entire API that is
261 * @param delta time in milli seconds
262 * @param do_round are we allowed to round a bit?
263 * @return string in human-readable form
266 GNUNET_STRINGS_relative_time_to_string (struct GNUNET_TIME_Relative delta,
272 * Returns a pointer to a part of filename (allocates nothing)!
274 * @param filename filename to extract basename from
275 * @return short (base) name of the file (that is, everything following the
276 * last directory separator in filename. If filename ends with a
277 * directory separator, the result will be a zero-length string.
278 * If filename has no directory separators, the result is filename
282 GNUNET_STRINGS_get_short_name (const char *filename);
286 * Convert binary data to ASCII encoding using CrockfordBase32.
287 * Does not append 0-terminator, but returns a pointer to the place where
288 * it should be placed, if needed.
290 * @param data data to encode
291 * @param size size of data (in bytes)
292 * @param out buffer to fill
293 * @param out_size size of the buffer. Must be large enough to hold
294 * ((size*8) + (((size*8) % 5) > 0 ? 5 - ((size*8) % 5) : 0)) / 5
295 * @return pointer to the next byte in 'out' or NULL on error.
298 GNUNET_STRINGS_data_to_string (const void *data,
305 * Return the base32crockford encoding of the given buffer.
307 * The returned string will be freshly allocated, and must be free'd
308 * with #GNUNET_free().
310 * @param buf buffer with data
311 * @param size size of the buffer @a buf
312 * @return freshly allocated, null-terminated string
315 GNUNET_STRINGS_data_to_string_alloc (const void *buf,
320 * Convert CrockfordBase32 encoding back to data.
321 * @a out_size must match exactly the size of the data before it was encoded.
323 * @param enc the encoding
324 * @param enclen number of characters in @a enc (without 0-terminator, which can be missing)
325 * @param out location where to store the decoded data
326 * @param out_size size of the output buffer @a out
327 * @return #GNUNET_OK on success, #GNUNET_SYSERR if result has the wrong encoding
330 GNUNET_STRINGS_string_to_data (const char *enc,
337 * Encode into Base64.
339 * @param data the data to encode
340 * @param len the length of the input
341 * @param output where to write the output (*output should be NULL,
343 * @return the size of the output
346 GNUNET_STRINGS_base64_encode (const void *in,
352 * Encode into Base64url. RFC7515
354 * @param in the data to encode
355 * @param len the length of the input
356 * @param output where to write the output (*output should be NULL,
358 * @return the size of the output
361 GNUNET_STRINGS_base64url_encode (const void *in, size_t len, char **output);
365 * Decode from Base64.
367 * @param data the data to encode
368 * @param len the length of the input
369 * @param[out] output where to write the output (*output should be NULL,
371 * @return the size of the output
374 GNUNET_STRINGS_base64_decode (const char *data,
380 * Decode from Base64url. RFC7515
382 * @param data the data to decode
383 * @param len the length of the input
384 * @param output where to write the output (*output should be NULL,
386 * @return the size of the output
389 GNUNET_STRINGS_base64url_decode (const char *data, size_t len, void **out);
393 * Convert a peer path to a human-readable string.
395 * @param pids array of PIDs to convert to a string
396 * @param num_pids length of the @a pids array
397 * @return string representing the array of @a pids
400 GNUNET_STRINGS_pp2s (const struct GNUNET_PeerIdentity *pids,
401 unsigned int num_pids);
405 * Parse a path that might be an URI.
407 * @param path path to parse. Must be NULL-terminated.
408 * @param[out] scheme_part pointer to a string that
409 * represents the URI scheme will be stored. Can be NULL. The string is
410 * allocated by the function, and should be freed by GNUNET_free() when
411 * it is no longer needed.
412 * @param path_part a pointer to 'const char *' where a pointer to the path
413 * part of the URI will be stored. Can be NULL. Points to the same block
414 * of memory as @a path, and thus must not be freed. Might point to '\0',
415 * if path part is zero-length.
416 * @return #GNUNET_YES if it's an URI, #GNUNET_NO otherwise. If 'path' is not
417 * an URI, '* scheme_part' and '*path_part' will remain unchanged
418 * (if they weren't NULL).
421 GNUNET_STRINGS_parse_uri (const char *path,
423 const char **path_part);
427 * Check whether filename is absolute or not, and if it's an URI
429 * @param filename filename to check
430 * @param can_be_uri #GNUNET_YES to check for being URI, #GNUNET_NO - to
431 * assume it's not URI
432 * @param r_is_uri a pointer to an int that is set to #GNUNET_YES if 'filename'
433 * is URI and to GNUNET_NO otherwise. Can be NULL. If 'can_be_uri' is
434 * not #GNUNET_YES, *r_is_uri is set to #GNUNET_NO.
435 * @param r_uri_scheme a pointer to a char * that is set to a pointer to URI scheme.
436 * The string is allocated by the function, and should be freed with
437 * GNUNET_free (). Can be NULL.
438 * @return #GNUNET_YES if 'filename' is absolute, #GNUNET_NO otherwise.
441 GNUNET_STRINGS_path_is_absolute (const char *filename,
444 char **r_uri_scheme);
448 * Flags for what we should check a file for.
450 enum GNUNET_STRINGS_FilenameCheck
453 * Check that it exists.
455 GNUNET_STRINGS_CHECK_EXISTS = 0x00000001,
458 * Check that it is a directory.
460 GNUNET_STRINGS_CHECK_IS_DIRECTORY = 0x00000002,
463 * Check that it is a link.
465 GNUNET_STRINGS_CHECK_IS_LINK = 0x00000004,
468 * Check that the path is an absolute path.
470 GNUNET_STRINGS_CHECK_IS_ABSOLUTE = 0x00000008
475 * Perform checks on @a filename. FIXME: some duplication with
476 * "GNUNET_DISK_"-APIs. We should unify those.
478 * @param filename file to check
479 * @param checks checks to perform
480 * @return #GNUNET_YES if all checks pass, #GNUNET_NO if at least one of them
481 * fails, #GNUNET_SYSERR when a check can't be performed
484 GNUNET_STRINGS_check_filename (const char *filename,
485 enum GNUNET_STRINGS_FilenameCheck checks);
489 * Tries to convert @a zt_addr string to an IPv6 address.
490 * The string is expected to have the format "[ABCD::01]:80".
492 * @param zt_addr 0-terminated string. May be mangled by the function.
493 * @param addrlen length of zt_addr (not counting 0-terminator).
494 * @param r_buf a buffer to fill. Initially gets filled with zeroes,
495 * then its sin6_port, sin6_family and sin6_addr are set appropriately.
496 * @return #GNUNET_OK if conversion succeded. #GNUNET_SYSERR otherwise, in which
497 * case the contents of r_buf are undefined.
500 GNUNET_STRINGS_to_address_ipv6 (const char *zt_addr,
502 struct sockaddr_in6 *r_buf);
506 * Tries to convert @a zt_addr string to an IPv4 address.
507 * The string is expected to have the format "1.2.3.4:80".
509 * @param zt_addr 0-terminated string. May be mangled by the function.
510 * @param addrlen length of zt_addr (not counting 0-terminator).
511 * @param r_buf a buffer to fill.
512 * @return #GNUNET_OK if conversion succeded. #GNUNET_SYSERR otherwise, in which case
513 * the contents of r_buf are undefined.
516 GNUNET_STRINGS_to_address_ipv4 (const char *zt_addr,
518 struct sockaddr_in *r_buf);
522 * Parse an address given as a string into a
525 * @param addr the address
526 * @param[out] af set to the parsed address family (i.e. AF_INET)
527 * @param[out] sa set to the parsed address
528 * @return 0 on error, otherwise number of bytes in @a sa
531 GNUNET_STRINGS_parse_socket_addr (const char *addr,
533 struct sockaddr **sa);
537 * Tries to convert @a addr string to an IP (v4 or v6) address.
538 * Will automatically decide whether to treat 'addr' as v4 or v6 address.
540 * @param addr a string, may not be 0-terminated.
541 * @param addrlen number of bytes in @a addr (if addr is 0-terminated,
542 * 0-terminator should not be counted towards addrlen).
543 * @param r_buf a buffer to fill.
544 * @return #GNUNET_OK if conversion succeded. #GNUNET_SYSERR otherwise, in which
545 * case the contents of r_buf are undefined.
548 GNUNET_STRINGS_to_address_ip (const char *addr,
550 struct sockaddr_storage *r_buf);
554 * Returns utf-8 encoded arguments. Does nothing (returns a copy of
555 * @a argc and @a argv) on any platform other than W32. Returned @a
556 * argv has `u8argv[u8argc] == NULL`. Returned @a argv is a single
557 * memory block, and can be freed with a single GNUNET_free() call.
559 * @param argc argc (as given by main())
560 * @param argv argv (as given by main())
561 * @param u8argc a location to store new argc in (though it's th same as argc)
562 * @param u8argv a location to store new argv in
563 * @return #GNUNET_OK on success, #GNUNET_SYSERR on failure
566 GNUNET_STRINGS_get_utf8_args (int argc,
569 char *const **u8argv);
573 * Like strlcpy but portable. The given string @a src is copied in full length
574 * (until its null byte). The destination buffer is guaranteed to be
577 * to a destination buffer
578 * and ensures that the destination string is null-terminated.
580 * @param dst destination of the copy
581 * @param src source of the copy, must be null-terminated
582 * @param n the length of the string to copy, including its terminating null
584 * @return the length of the string that was copied, excluding the terminating
588 GNUNET_strlcpy (char *dst, const char *src, size_t n);
591 /* ***************** IPv4/IPv6 parsing ****************** */
593 struct GNUNET_STRINGS_PortPolicy
596 * Starting port range (0 if none given).
601 * End of port range (0 if none given).
606 * #GNUNET_YES if the port range should be negated
609 int negate_portrange;
614 * @brief IPV4 network in CIDR notation.
616 struct GNUNET_STRINGS_IPv4NetworkPolicy
621 struct in_addr network;
626 struct in_addr netmask;
629 * Policy for port access.
631 struct GNUNET_STRINGS_PortPolicy pp;
636 * @brief network in CIDR notation for IPV6.
638 struct GNUNET_STRINGS_IPv6NetworkPolicy
643 struct in6_addr network;
648 struct in6_addr netmask;
651 * Policy for port access.
653 struct GNUNET_STRINGS_PortPolicy pp;
658 * Parse an IPv4 network policy. The argument specifies a list of
659 * subnets. The format is <tt>(network[/netmask][:[!]SPORT-DPORT];)*</tt>
660 * (no whitespace, must be terminated with a semicolon). The network
661 * must be given in dotted-decimal notation. The netmask can be given
662 * in CIDR notation (/16) or in dotted-decimal (/255.255.0.0).
664 * @param routeListX a string specifying the IPv4 subnets
665 * @return the converted list, terminated with all zeros;
666 * NULL if the synatx is flawed
668 struct GNUNET_STRINGS_IPv4NetworkPolicy *
669 GNUNET_STRINGS_parse_ipv4_policy (const char *routeListX);
673 * Parse an IPv6 network policy. The argument specifies a list of
674 * subnets. The format is <tt>(network[/netmask[:[!]SPORT[-DPORT]]];)*</tt>
675 * (no whitespace, must be terminated with a semicolon). The network
676 * must be given in colon-hex notation. The netmask must be given in
677 * CIDR notation (/16) or can be omitted to specify a single host.
678 * Note that the netmask is mandatory if ports are specified.
680 * @param routeListX a string specifying the policy
681 * @return the converted list, 0-terminated, NULL if the synatx is flawed
683 struct GNUNET_STRINGS_IPv6NetworkPolicy *
684 GNUNET_STRINGS_parse_ipv6_policy (const char *routeListX);
687 #if 0 /* keep Emacsens' auto-indent happy */
694 /* ifndef GNUNET_UTIL_STRING_H */
697 /** @} */ /* end of group */
699 /* end of gnunet_util_string.h */