1 /* Various utility functions.
2 Copyright (C) 1996-2006 Free Software Foundation, Inc.
4 This file is part of GNU Wget.
6 GNU Wget is free software; you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation; either version 3 of the License, or
9 (at your option) any later version.
11 GNU Wget is distributed in the hope that it will be useful,
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 GNU General Public License for more details.
16 You should have received a copy of the GNU General Public License
17 along with Wget. If not, see <http://www.gnu.org/licenses/>.
19 In addition, as a special exception, the Free Software Foundation
20 gives permission to link the code of its release of Wget with the
21 OpenSSL project's "OpenSSL" library (or with modified versions of it
22 that use the same license as the "OpenSSL" library), and distribute
23 the linked executables. You must obey the GNU General Public License
24 in all respects for all of the code used other than "OpenSSL". If you
25 modify this file, you may extend this exception to your version of the
26 file, but you are not obligated to do so. If you do not wish to do
27 so, delete this exception statement from your version. */
35 #ifdef HAVE_SYS_TIME_H
36 # include <sys/time.h>
42 # include <sys/mman.h>
47 #ifdef HAVE_SYS_UTIME_H
48 # include <sys/utime.h>
56 /* For TIOCGWINSZ and friends: */
57 #ifdef HAVE_SYS_IOCTL_H
58 # include <sys/ioctl.h>
64 /* Needed for Unix version of run_with_timeout. */
68 #ifndef HAVE_SIGSETJMP
69 /* If sigsetjmp is a macro, configure won't pick it up. */
71 # define HAVE_SIGSETJMP
75 #if defined HAVE_SIGSETJMP || defined HAVE_SIGBLOCK
76 # define USE_SIGNAL_TIMEOUT
87 /* Utility function: like xstrdup(), but also lowercases S. */
90 xstrdup_lower (const char *s)
92 char *copy = xstrdup (s);
99 /* Copy the string formed by two pointers (one on the beginning, other
100 on the char after the last char) to a new, malloc-ed location.
103 strdupdelim (const char *beg, const char *end)
105 char *res = xmalloc (end - beg + 1);
106 memcpy (res, beg, end - beg);
107 res[end - beg] = '\0';
111 /* Parse a string containing comma-separated elements, and return a
112 vector of char pointers with the elements. Spaces following the
113 commas are ignored. */
115 sepstring (const char *s)
129 res = xrealloc (res, (i + 2) * sizeof (char *));
130 res[i] = strdupdelim (p, s);
133 /* Skip the blanks following the ','. */
141 res = xrealloc (res, (i + 2) * sizeof (char *));
142 res[i] = strdupdelim (p, s);
147 /* Like sprintf, but prints into a string of sufficient size freshly
148 allocated with malloc, which is returned. If unable to print due
149 to invalid format, returns NULL. Inability to allocate needed
150 memory results in abort, as with xmalloc. This is in spirit
151 similar to the GNU/BSD extension asprintf, but somewhat easier to
154 Internally the function either calls vasprintf or loops around
155 vsnprintf until the correct size is found. Since Wget also ships a
156 fallback implementation of vsnprintf, this should be portable. */
159 aprintf (const char *fmt, ...)
161 #if defined HAVE_VASPRINTF && !defined DEBUG_MALLOC
166 va_start (args, fmt);
167 ret = vasprintf (&str, fmt, args);
169 if (ret < 0 && errno == ENOMEM)
170 abort (); /* for consistency with xmalloc/xrealloc */
174 #else /* not HAVE_VASPRINTF */
176 /* vasprintf is unavailable. snprintf into a small buffer and
177 resize it as necessary. */
179 char *str = xmalloc (size);
181 /* #### This code will infloop and eventually abort in xrealloc if
182 passed a FMT that causes snprintf to consistently return -1. */
189 va_start (args, fmt);
190 n = vsnprintf (str, size, fmt, args);
193 /* If the printing worked, return the string. */
194 if (n > -1 && n < size)
197 /* Else try again with a larger buffer. */
198 if (n > -1) /* C99 */
199 size = n + 1; /* precisely what is needed */
201 size <<= 1; /* twice the old size */
202 str = xrealloc (str, size);
204 #endif /* not HAVE_VASPRINTF */
207 /* Concatenate the NULL-terminated list of string arguments into
208 freshly allocated space. */
211 concat_strings (const char *str0, ...)
214 int saved_lengths[5]; /* inspired by Apache's apr_pstrcat */
217 const char *next_str;
218 int total_length = 0;
221 /* Calculate the length of and allocate the resulting string. */
224 va_start (args, str0);
225 for (next_str = str0; next_str != NULL; next_str = va_arg (args, char *))
227 int len = strlen (next_str);
228 if (argcount < countof (saved_lengths))
229 saved_lengths[argcount++] = len;
233 p = ret = xmalloc (total_length + 1);
235 /* Copy the strings into the allocated space. */
238 va_start (args, str0);
239 for (next_str = str0; next_str != NULL; next_str = va_arg (args, char *))
242 if (argcount < countof (saved_lengths))
243 len = saved_lengths[argcount++];
245 len = strlen (next_str);
246 memcpy (p, next_str, len);
255 /* Format the provided time according to the specified format. The
256 format is a string with format elements supported by strftime. */
259 fmttime (time_t t, const char *fmt)
261 static char output[32];
262 struct tm *tm = localtime(&t);
265 if (!strftime(output, sizeof(output), fmt, tm))
270 /* Return pointer to a static char[] buffer in which zero-terminated
271 string-representation of TM (in form hh:mm:ss) is printed.
273 If TM is NULL, the current time will be used. */
278 return fmttime(t, "%H:%M:%S");
281 /* Like the above, but include the date: YYYY-MM-DD hh:mm:ss. */
284 datetime_str (time_t t)
286 return fmttime(t, "%Y-%m-%d %H:%M:%S");
289 /* The Windows versions of the following two functions are defined in
294 fork_to_background (void)
297 /* Whether we arrange our own version of opt.lfilename here. */
298 bool logfile_changed = false;
302 /* We must create the file immediately to avoid either a race
303 condition (which arises from using unique_name and failing to
304 use fopen_excl) or lying to the user about the log file name
305 (which arises from using unique_name, printing the name, and
306 using fopen_excl later on.) */
307 FILE *new_log_fp = unique_create (DEFAULT_LOGFILE, false, &opt.lfilename);
310 logfile_changed = true;
323 /* parent, no error */
324 printf (_("Continuing in background, pid %d.\n"), (int) pid);
326 printf (_("Output will be written to `%s'.\n"), opt.lfilename);
327 exit (0); /* #### should we use _exit()? */
330 /* child: give up the privileges and keep running. */
332 freopen ("/dev/null", "r", stdin);
333 freopen ("/dev/null", "w", stdout);
334 freopen ("/dev/null", "w", stderr);
336 #endif /* not WINDOWS */
338 /* "Touch" FILE, i.e. make its mtime ("modified time") equal the time
339 specified with TM. The atime ("access time") is set to the current
343 touch (const char *file, time_t tm)
345 #ifdef HAVE_STRUCT_UTIMBUF
346 struct utimbuf times;
354 times.actime = time (NULL);
355 if (utime (file, ×) == -1)
356 logprintf (LOG_NOTQUIET, "utime(%s): %s\n", file, strerror (errno));
359 /* Checks if FILE is a symbolic link, and removes it if it is. Does
360 nothing under MS-Windows. */
362 remove_link (const char *file)
367 if (lstat (file, &st) == 0 && S_ISLNK (st.st_mode))
369 DEBUGP (("Unlinking %s (symlink).\n", file));
372 logprintf (LOG_VERBOSE, _("Failed to unlink symlink `%s': %s\n"),
373 file, strerror (errno));
378 /* Does FILENAME exist? This is quite a lousy implementation, since
379 it supplies no error codes -- only a yes-or-no answer. Thus it
380 will return that a file does not exist if, e.g., the directory is
381 unreadable. I don't mind it too much currently, though. The
382 proper way should, of course, be to have a third, error state,
383 other than true/false, but that would introduce uncalled-for
384 additional complexity to the callers. */
386 file_exists_p (const char *filename)
389 return access (filename, F_OK) >= 0;
392 return stat (filename, &buf) >= 0;
396 /* Returns 0 if PATH is a directory, 1 otherwise (any kind of file).
397 Returns 0 on error. */
399 file_non_directory_p (const char *path)
402 /* Use lstat() rather than stat() so that symbolic links pointing to
403 directories can be identified correctly. */
404 if (lstat (path, &buf) != 0)
406 return S_ISDIR (buf.st_mode) ? false : true;
409 /* Return the size of file named by FILENAME, or -1 if it cannot be
410 opened or seeked into. */
412 file_size (const char *filename)
414 #if defined(HAVE_FSEEKO) && defined(HAVE_FTELLO)
416 /* We use fseek rather than stat to determine the file size because
417 that way we can also verify that the file is readable without
418 explicitly checking for permissions. Inspired by the POST patch
420 FILE *fp = fopen (filename, "rb");
423 fseeko (fp, 0, SEEK_END);
429 if (stat (filename, &st) < 0)
435 /* stat file names named PREFIX.1, PREFIX.2, etc., until one that
436 doesn't exist is found. Return a freshly allocated copy of the
440 unique_name_1 (const char *prefix)
443 int plen = strlen (prefix);
444 char *template = (char *)alloca (plen + 1 + 24);
445 char *template_tail = template + plen;
447 memcpy (template, prefix, plen);
448 *template_tail++ = '.';
451 number_to_string (template_tail, count++);
452 while (file_exists_p (template));
454 return xstrdup (template);
457 /* Return a unique file name, based on FILE.
459 More precisely, if FILE doesn't exist, it is returned unmodified.
460 If not, FILE.1 is tried, then FILE.2, etc. The first FILE.<number>
461 file name that doesn't exist is returned.
463 The resulting file is not created, only verified that it didn't
464 exist at the point in time when the function was called.
465 Therefore, where security matters, don't rely that the file created
466 by this function exists until you open it with O_EXCL or
469 If ALLOW_PASSTHROUGH is 0, it always returns a freshly allocated
470 string. Otherwise, it may return FILE if the file doesn't exist
471 (and therefore doesn't need changing). */
474 unique_name (const char *file, bool allow_passthrough)
476 /* If the FILE itself doesn't exist, return it without
478 if (!file_exists_p (file))
479 return allow_passthrough ? (char *)file : xstrdup (file);
481 /* Otherwise, find a numeric suffix that results in unused file name
483 return unique_name_1 (file);
486 /* Create a file based on NAME, except without overwriting an existing
487 file with that name. Providing O_EXCL is correctly implemented,
488 this function does not have the race condition associated with
489 opening the file returned by unique_name. */
492 unique_create (const char *name, bool binary, char **opened_name)
494 /* unique file name, based on NAME */
495 char *uname = unique_name (name, false);
497 while ((fp = fopen_excl (uname, binary)) == NULL && errno == EEXIST)
500 uname = unique_name (name, false);
502 if (opened_name && fp != NULL)
505 *opened_name = uname;
517 /* Open the file for writing, with the addition that the file is
518 opened "exclusively". This means that, if the file already exists,
519 this function will *fail* and errno will be set to EEXIST. If
520 BINARY is set, the file will be opened in binary mode, equivalent
523 If opening the file fails for any reason, including the file having
524 previously existed, this function returns NULL and sets errno
528 fopen_excl (const char *fname, bool binary)
532 int flags = O_WRONLY | O_CREAT | O_EXCL;
537 fd = open (fname, flags, 0666);
540 return fdopen (fd, binary ? "wb" : "w");
541 #else /* not O_EXCL */
542 /* Manually check whether the file exists. This is prone to race
543 conditions, but systems without O_EXCL haven't deserved
545 if (file_exists_p (fname))
550 return fopen (fname, binary ? "wb" : "w");
551 #endif /* not O_EXCL */
554 /* Create DIRECTORY. If some of the pathname components of DIRECTORY
555 are missing, create them first. In case any mkdir() call fails,
556 return its error status. Returns 0 on successful completion.
558 The behaviour of this function should be identical to the behaviour
559 of `mkdir -p' on systems where mkdir supports the `-p' option. */
561 make_directory (const char *directory)
563 int i, ret, quit = 0;
566 /* Make a copy of dir, to be able to write to it. Otherwise, the
567 function is unsafe if called with a read-only char *argument. */
568 STRDUP_ALLOCA (dir, directory);
570 /* If the first character of dir is '/', skip it (and thus enable
571 creation of absolute-pathname directories. */
572 for (i = (*dir == '/'); 1; ++i)
574 for (; dir[i] && dir[i] != '/'; i++)
579 /* Check whether the directory already exists. Allow creation of
580 of intermediate directories to fail, as the initial path components
581 are not necessarily directories! */
582 if (!file_exists_p (dir))
583 ret = mkdir (dir, 0777);
594 /* Merge BASE with FILE. BASE can be a directory or a file name, FILE
595 should be a file name.
597 file_merge("/foo/bar", "baz") => "/foo/baz"
598 file_merge("/foo/bar/", "baz") => "/foo/bar/baz"
599 file_merge("foo", "bar") => "bar"
601 In other words, it's a simpler and gentler version of uri_merge. */
604 file_merge (const char *base, const char *file)
607 const char *cut = (const char *)strrchr (base, '/');
610 return xstrdup (file);
612 result = xmalloc (cut - base + 1 + strlen (file) + 1);
613 memcpy (result, base, cut - base);
614 result[cut - base] = '/';
615 strcpy (result + (cut - base) + 1, file);
620 /* Like fnmatch, but performs a case-insensitive match. */
623 fnmatch_nocase (const char *pattern, const char *string, int flags)
626 /* The FNM_CASEFOLD flag started as a GNU extension, but it is now
627 also present on *BSD platforms, and possibly elsewhere. */
628 return fnmatch (pattern, string, flags | FNM_CASEFOLD);
630 /* Turn PATTERN and STRING to lower case and call fnmatch on them. */
631 char *patcopy = (char *) alloca (strlen (pattern) + 1);
632 char *strcopy = (char *) alloca (strlen (string) + 1);
634 for (p = patcopy; *pattern; pattern++, p++)
635 *p = TOLOWER (*pattern);
637 for (p = strcopy; *string; string++, p++)
638 *p = TOLOWER (*string);
640 return fnmatch (patcopy, strcopy, flags);
644 static bool in_acclist (const char *const *, const char *, bool);
646 /* Determine whether a file is acceptable to be followed, according to
647 lists of patterns to accept/reject. */
649 acceptable (const char *s)
653 while (l && s[l] != '/')
660 return (in_acclist ((const char *const *)opt.accepts, s, true)
661 && !in_acclist ((const char *const *)opt.rejects, s, true));
663 return in_acclist ((const char *const *)opt.accepts, s, true);
665 else if (opt.rejects)
666 return !in_acclist ((const char *const *)opt.rejects, s, true);
670 /* Check if D2 is a subdirectory of D1. E.g. if D1 is `/something', subdir_p()
671 will return true if and only if D2 begins with `/something/' or is exactly
674 subdir_p (const char *d1, const char *d2)
676 if (!opt.ignore_case)
677 for (; *d1 && *d2 && (*d1 == *d2); ++d1, ++d2)
680 for (; *d1 && *d2 && (TOLOWER (*d1) == TOLOWER (*d2)); ++d1, ++d2)
683 return *d1 == '\0' && (*d2 == '\0' || *d2 == '/');
686 /* Iterate through DIRLIST (which must be NULL-terminated), and return the
687 first element that matches DIR, through wildcards or front comparison (as
690 dir_matches_p (char **dirlist, const char *dir)
693 int (*matcher) (const char *, const char *, int)
694 = opt.ignore_case ? fnmatch_nocase : fnmatch;
696 for (x = dirlist; *x; x++)
698 /* Remove leading '/' */
699 char *p = *x + (**x == '/');
700 if (has_wildcards_p (p))
702 if (matcher (p, dir, FNM_PATHNAME) == 0)
707 if (subdir_p (p, dir))
712 return *x ? true : false;
715 /* Returns whether DIRECTORY is acceptable for download, wrt the
716 include/exclude lists.
718 The leading `/' is ignored in paths; relative and absolute paths
719 may be freely intermixed. */
722 accdir (const char *directory)
724 /* Remove starting '/'. */
725 if (*directory == '/')
729 if (!dir_matches_p (opt.includes, directory))
734 if (dir_matches_p (opt.excludes, directory))
740 /* Return true if STRING ends with TAIL. For instance:
742 match_tail ("abc", "bc", false) -> 1
743 match_tail ("abc", "ab", false) -> 0
744 match_tail ("abc", "abc", false) -> 1
746 If FOLD_CASE is true, the comparison will be case-insensitive. */
749 match_tail (const char *string, const char *tail, bool fold_case)
753 /* We want this to be fast, so we code two loops, one with
754 case-folding, one without. */
758 for (i = strlen (string), j = strlen (tail); i >= 0 && j >= 0; i--, j--)
759 if (string[i] != tail[j])
764 for (i = strlen (string), j = strlen (tail); i >= 0 && j >= 0; i--, j--)
765 if (TOLOWER (string[i]) != TOLOWER (tail[j]))
769 /* If the tail was exhausted, the match was succesful. */
776 /* Checks whether string S matches each element of ACCEPTS. A list
777 element are matched either with fnmatch() or match_tail(),
778 according to whether the element contains wildcards or not.
780 If the BACKWARD is false, don't do backward comparison -- just compare
783 in_acclist (const char *const *accepts, const char *s, bool backward)
785 for (; *accepts; accepts++)
787 if (has_wildcards_p (*accepts))
789 int res = opt.ignore_case
790 ? fnmatch_nocase (*accepts, s, 0) : fnmatch (*accepts, s, 0);
791 /* fnmatch returns 0 if the pattern *does* match the string. */
799 if (match_tail (s, *accepts, opt.ignore_case))
804 int cmp = opt.ignore_case
805 ? strcasecmp (s, *accepts) : strcmp (s, *accepts);
814 /* Return the location of STR's suffix (file extension). Examples:
815 suffix ("foo.bar") -> "bar"
816 suffix ("foo.bar.baz") -> "baz"
817 suffix ("/foo/bar") -> NULL
818 suffix ("/foo.bar/baz") -> NULL */
820 suffix (const char *str)
824 for (i = strlen (str); i && str[i] != '/' && str[i] != '.'; i--)
828 return (char *)str + i;
833 /* Return true if S contains globbing wildcards (`*', `?', `[' or
837 has_wildcards_p (const char *s)
840 if (*s == '*' || *s == '?' || *s == '[' || *s == ']')
845 /* Return true if FNAME ends with a typical HTML suffix. The
846 following (case-insensitive) suffixes are presumed to be HTML
851 ?html (`?' matches one character)
853 #### CAVEAT. This is not necessarily a good indication that FNAME
854 refers to a file that contains HTML! */
856 has_html_suffix_p (const char *fname)
860 if ((suf = suffix (fname)) == NULL)
862 if (!strcasecmp (suf, "html"))
864 if (!strcasecmp (suf, "htm"))
866 if (suf[0] && !strcasecmp (suf + 1, "html"))
871 /* Read a line from FP and return the pointer to freshly allocated
872 storage. The storage space is obtained through malloc() and should
873 be freed with free() when it is no longer needed.
875 The length of the line is not limited, except by available memory.
876 The newline character at the end of line is retained. The line is
877 terminated with a zero character.
879 After end-of-file is encountered without anything being read, NULL
880 is returned. NULL is also returned on error. To distinguish
881 between these two cases, use the stdio function ferror(). */
884 read_whole_line (FILE *fp)
888 char *line = xmalloc (bufsize);
890 while (fgets (line + length, bufsize - length, fp))
892 length += strlen (line + length);
894 /* Possible for example when reading from a binary file where
895 a line begins with \0. */
898 if (line[length - 1] == '\n')
901 /* fgets() guarantees to read the whole line, or to use up the
902 space we've given it. We can double the buffer
905 line = xrealloc (line, bufsize);
907 if (length == 0 || ferror (fp))
912 if (length + 1 < bufsize)
913 /* Relieve the memory from our exponential greediness. We say
914 `length + 1' because the terminating \0 is not included in
915 LENGTH. We don't need to zero-terminate the string ourselves,
916 though, because fgets() does that. */
917 line = xrealloc (line, length + 1);
921 /* Read FILE into memory. A pointer to `struct file_memory' are
922 returned; use struct element `content' to access file contents, and
923 the element `length' to know the file length. `content' is *not*
924 zero-terminated, and you should *not* read or write beyond the [0,
925 length) range of characters.
927 After you are done with the file contents, call read_file_free to
930 Depending on the operating system and the type of file that is
931 being read, read_file() either mmap's the file into memory, or
932 reads the file into the core using read().
934 If file is named "-", fileno(stdin) is used for reading instead.
935 If you want to read from a real file named "-", use "./-" instead. */
938 read_file (const char *file)
941 struct file_memory *fm;
943 bool inhibit_close = false;
945 /* Some magic in the finest tradition of Perl and its kin: if FILE
946 is "-", just use stdin. */
950 inhibit_close = true;
951 /* Note that we don't inhibit mmap() in this case. If stdin is
952 redirected from a regular file, mmap() will still work. */
955 fd = open (file, O_RDONLY);
958 fm = xnew (struct file_memory);
963 if (fstat (fd, &buf) < 0)
965 fm->length = buf.st_size;
966 /* NOTE: As far as I know, the callers of this function never
967 modify the file text. Relying on this would enable us to
968 specify PROT_READ and MAP_SHARED for a marginal gain in
969 efficiency, but at some cost to generality. */
970 fm->content = mmap (NULL, fm->length, PROT_READ | PROT_WRITE,
972 if (fm->content == (char *)MAP_FAILED)
982 /* The most common reason why mmap() fails is that FD does not point
983 to a plain file. However, it's also possible that mmap() doesn't
984 work for a particular type of file. Therefore, whenever mmap()
985 fails, we just fall back to the regular method. */
986 #endif /* HAVE_MMAP */
989 size = 512; /* number of bytes fm->contents can
990 hold at any given time. */
991 fm->content = xmalloc (size);
995 if (fm->length > size / 2)
997 /* #### I'm not sure whether the whole exponential-growth
998 thing makes sense with kernel read. On Linux at least,
999 read() refuses to read more than 4K from a file at a
1000 single chunk anyway. But other Unixes might optimize it
1001 better, and it doesn't *hurt* anything, so I'm leaving
1004 /* Normally, we grow SIZE exponentially to make the number
1005 of calls to read() and realloc() logarithmic in relation
1006 to file size. However, read() can read an amount of data
1007 smaller than requested, and it would be unreasonable to
1008 double SIZE every time *something* was read. Therefore,
1009 we double SIZE only when the length exceeds half of the
1010 entire allocated size. */
1012 fm->content = xrealloc (fm->content, size);
1014 nread = read (fd, fm->content + fm->length, size - fm->length);
1016 /* Successful read. */
1017 fm->length += nread;
1027 if (size > fm->length && fm->length != 0)
1028 /* Due to exponential growth of fm->content, the allocated region
1029 might be much larger than what is actually needed. */
1030 fm->content = xrealloc (fm->content, fm->length);
1037 xfree (fm->content);
1042 /* Release the resources held by FM. Specifically, this calls
1043 munmap() or xfree() on fm->content, depending whether mmap or
1044 malloc/read were used to read in the file. It also frees the
1045 memory needed to hold the FM structure itself. */
1048 read_file_free (struct file_memory *fm)
1053 munmap (fm->content, fm->length);
1058 xfree (fm->content);
1063 /* Free the pointers in a NULL-terminated vector of pointers, then
1064 free the pointer itself. */
1066 free_vec (char **vec)
1077 /* Append vector V2 to vector V1. The function frees V2 and
1078 reallocates V1 (thus you may not use the contents of neither
1079 pointer after the call). If V1 is NULL, V2 is returned. */
1081 merge_vecs (char **v1, char **v2)
1091 /* To avoid j == 0 */
1096 for (i = 0; v1[i]; i++)
1099 for (j = 0; v2[j]; j++)
1101 /* Reallocate v1. */
1102 v1 = xrealloc (v1, (i + j + 1) * sizeof (char **));
1103 memcpy (v1 + i, v2, (j + 1) * sizeof (char *));
1108 /* Append a freshly allocated copy of STR to VEC. If VEC is NULL, it
1109 is allocated as needed. Return the new value of the vector. */
1112 vec_append (char **vec, const char *str)
1114 int cnt; /* count of vector elements, including
1115 the one we're about to append */
1118 for (cnt = 0; vec[cnt]; cnt++)
1124 /* Reallocate the array to fit the new element and the NULL. */
1125 vec = xrealloc (vec, (cnt + 1) * sizeof (char *));
1126 /* Append a copy of STR to the vector. */
1127 vec[cnt - 1] = xstrdup (str);
1132 /* Sometimes it's useful to create "sets" of strings, i.e. special
1133 hash tables where you want to store strings as keys and merely
1134 query for their existence. Here is a set of utility routines that
1135 makes that transparent. */
1138 string_set_add (struct hash_table *ht, const char *s)
1140 /* First check whether the set element already exists. If it does,
1141 do nothing so that we don't have to free() the old element and
1142 then strdup() a new one. */
1143 if (hash_table_contains (ht, s))
1146 /* We use "1" as value. It provides us a useful and clear arbitrary
1147 value, and it consumes no memory -- the pointers to the same
1148 string "1" will be shared by all the key-value pairs in all `set'
1150 hash_table_put (ht, xstrdup (s), "1");
1153 /* Synonym for hash_table_contains... */
1156 string_set_contains (struct hash_table *ht, const char *s)
1158 return hash_table_contains (ht, s);
1161 /* Convert the specified string set to array. ARRAY should be large
1162 enough to hold hash_table_count(ht) char pointers. */
1164 void string_set_to_array (struct hash_table *ht, char **array)
1166 hash_table_iterator iter;
1167 for (hash_table_iterate (ht, &iter); hash_table_iter_next (&iter); )
1168 *array++ = iter.key;
1171 /* Free the string set. This frees both the storage allocated for
1172 keys and the actual hash table. (hash_table_destroy would only
1173 destroy the hash table.) */
1176 string_set_free (struct hash_table *ht)
1178 hash_table_iterator iter;
1179 for (hash_table_iterate (ht, &iter); hash_table_iter_next (&iter); )
1181 hash_table_destroy (ht);
1184 /* Utility function: simply call xfree() on all keys and values of HT. */
1187 free_keys_and_values (struct hash_table *ht)
1189 hash_table_iterator iter;
1190 for (hash_table_iterate (ht, &iter); hash_table_iter_next (&iter); )
1197 /* Get digit grouping data for thousand separors by calling
1198 localeconv(). The data includes separator string and grouping info
1199 and is cached after the first call to the function.
1201 In locales that don't set a thousand separator (such as the "C"
1202 locale), this forces it to be ",". We are now only showing
1203 thousand separators in one place, so this shouldn't be a problem in
1207 get_grouping_data (const char **sep, const char **grouping)
1209 static const char *cached_sep;
1210 static const char *cached_grouping;
1211 static bool initialized;
1214 /* Get the grouping info from the locale. */
1215 struct lconv *lconv = localeconv ();
1216 cached_sep = lconv->thousands_sep;
1217 cached_grouping = lconv->grouping;
1220 /* Many locales (such as "C" or "hr_HR") don't specify
1221 grouping, which we still want to use it for legibility.
1222 In those locales set the sep char to ',', unless that
1223 character is used for decimal point, in which case set it
1225 if (*lconv->decimal_point != ',')
1229 cached_grouping = "\x03";
1234 *grouping = cached_grouping;
1237 /* Return a printed representation of N with thousand separators.
1238 This should respect locale settings, with the exception of the "C"
1239 locale which mandates no separator, but we use one anyway.
1241 Unfortunately, we cannot use %'d (in fact it would be %'j) to get
1242 the separators because it's too non-portable, and it's hard to test
1243 for this feature at configure time. Besides, it wouldn't display
1244 separators in the "C" locale, still used by many Unix users. */
1247 with_thousand_seps (wgint n)
1249 static char outbuf[48];
1250 char *p = outbuf + sizeof outbuf;
1252 /* Info received from locale */
1253 const char *grouping, *sep;
1256 /* State information */
1257 int i = 0, groupsize;
1258 const char *atgroup;
1260 bool negative = n < 0;
1262 /* Initialize grouping data. */
1263 get_grouping_data (&sep, &grouping);
1264 seplen = strlen (sep);
1266 groupsize = *atgroup++;
1268 /* This would overflow on WGINT_MIN, but printing negative numbers
1269 is not an important goal of this fuinction. */
1273 /* Write the number into the buffer, backwards, inserting the
1274 separators as necessary. */
1278 *--p = n % 10 + '0';
1282 /* Prepend SEP to every groupsize'd digit and get new groupsize. */
1283 if (++i == groupsize)
1288 memcpy (p -= seplen, sep, seplen);
1291 groupsize = *atgroup++;
1300 /* N, a byte quantity, is converted to a human-readable abberviated
1301 form a la sizes printed by `ls -lh'. The result is written to a
1302 static buffer, a pointer to which is returned.
1304 Unlike `with_thousand_seps', this approximates to the nearest unit.
1305 Quoting GNU libit: "Most people visually process strings of 3-4
1306 digits effectively, but longer strings of digits are more prone to
1307 misinterpretation. Hence, converting to an abbreviated form
1308 usually improves readability."
1310 This intentionally uses kilobyte (KB), megabyte (MB), etc. in their
1311 original computer-related meaning of "powers of 1024". We don't
1312 use the "*bibyte" names invented in 1998, and seldom used in
1313 practice. Wikipedia's entry on "binary prefix" discusses this in
1317 human_readable (HR_NUMTYPE n)
1319 /* These suffixes are compatible with those of GNU `ls -lh'. */
1320 static char powers[] =
1322 'K', /* kilobyte, 2^10 bytes */
1323 'M', /* megabyte, 2^20 bytes */
1324 'G', /* gigabyte, 2^30 bytes */
1325 'T', /* terabyte, 2^40 bytes */
1326 'P', /* petabyte, 2^50 bytes */
1327 'E', /* exabyte, 2^60 bytes */
1332 /* If the quantity is smaller than 1K, just print it. */
1335 snprintf (buf, sizeof (buf), "%d", (int) n);
1339 /* Loop over powers, dividing N with 1024 in each iteration. This
1340 works unchanged for all sizes of wgint, while still avoiding
1341 non-portable `long double' arithmetic. */
1342 for (i = 0; i < countof (powers); i++)
1344 /* At each iteration N is greater than the *subsequent* power.
1345 That way N/1024.0 produces a decimal number in the units of
1347 if ((n / 1024) < 1024 || i == countof (powers) - 1)
1349 double val = n / 1024.0;
1350 /* Print values smaller than 10 with one decimal digits, and
1351 others without any decimals. */
1352 snprintf (buf, sizeof (buf), "%.*f%c",
1353 val < 10 ? 1 : 0, val, powers[i]);
1358 return NULL; /* unreached */
1361 /* Count the digits in the provided number. Used to allocate space
1362 when printing numbers. */
1365 numdigit (wgint number)
1369 ++cnt; /* accomodate '-' */
1370 while ((number /= 10) != 0)
1375 #define PR(mask) *p++ = n / (mask) + '0'
1377 /* DIGITS_<D> is used to print a D-digit number and should be called
1378 with mask==10^(D-1). It prints n/mask (the first digit), reducing
1379 n to n%mask (the remaining digits), and calling DIGITS_<D-1>.
1380 Recursively this continues until DIGITS_1 is invoked. */
1382 #define DIGITS_1(mask) PR (mask)
1383 #define DIGITS_2(mask) PR (mask), n %= (mask), DIGITS_1 ((mask) / 10)
1384 #define DIGITS_3(mask) PR (mask), n %= (mask), DIGITS_2 ((mask) / 10)
1385 #define DIGITS_4(mask) PR (mask), n %= (mask), DIGITS_3 ((mask) / 10)
1386 #define DIGITS_5(mask) PR (mask), n %= (mask), DIGITS_4 ((mask) / 10)
1387 #define DIGITS_6(mask) PR (mask), n %= (mask), DIGITS_5 ((mask) / 10)
1388 #define DIGITS_7(mask) PR (mask), n %= (mask), DIGITS_6 ((mask) / 10)
1389 #define DIGITS_8(mask) PR (mask), n %= (mask), DIGITS_7 ((mask) / 10)
1390 #define DIGITS_9(mask) PR (mask), n %= (mask), DIGITS_8 ((mask) / 10)
1391 #define DIGITS_10(mask) PR (mask), n %= (mask), DIGITS_9 ((mask) / 10)
1393 /* DIGITS_<11-20> are only used on machines with 64-bit wgints. */
1395 #define DIGITS_11(mask) PR (mask), n %= (mask), DIGITS_10 ((mask) / 10)
1396 #define DIGITS_12(mask) PR (mask), n %= (mask), DIGITS_11 ((mask) / 10)
1397 #define DIGITS_13(mask) PR (mask), n %= (mask), DIGITS_12 ((mask) / 10)
1398 #define DIGITS_14(mask) PR (mask), n %= (mask), DIGITS_13 ((mask) / 10)
1399 #define DIGITS_15(mask) PR (mask), n %= (mask), DIGITS_14 ((mask) / 10)
1400 #define DIGITS_16(mask) PR (mask), n %= (mask), DIGITS_15 ((mask) / 10)
1401 #define DIGITS_17(mask) PR (mask), n %= (mask), DIGITS_16 ((mask) / 10)
1402 #define DIGITS_18(mask) PR (mask), n %= (mask), DIGITS_17 ((mask) / 10)
1403 #define DIGITS_19(mask) PR (mask), n %= (mask), DIGITS_18 ((mask) / 10)
1405 /* Shorthand for casting to wgint. */
1408 /* Print NUMBER to BUFFER in base 10. This is equivalent to
1409 `sprintf(buffer, "%lld", (long long) number)', only typically much
1410 faster and portable to machines without long long.
1412 The speedup may make a difference in programs that frequently
1413 convert numbers to strings. Some implementations of sprintf,
1414 particularly the one in some versions of GNU libc, have been known
1415 to be quite slow when converting integers to strings.
1417 Return the pointer to the location where the terminating zero was
1418 printed. (Equivalent to calling buffer+strlen(buffer) after the
1421 BUFFER should be large enough to accept as many bytes as you expect
1422 the number to take up. On machines with 64-bit wgints the maximum
1423 needed size is 24 bytes. That includes the digits needed for the
1424 largest 64-bit number, the `-' sign in case it's negative, and the
1425 terminating '\0'. */
1428 number_to_string (char *buffer, wgint number)
1433 int last_digit_char = 0;
1435 #if (SIZEOF_WGINT != 4) && (SIZEOF_WGINT != 8)
1436 /* We are running in a very strange environment. Leave the correct
1437 printing to sprintf. */
1438 p += sprintf (buf, "%j", (intmax_t) (n));
1439 #else /* (SIZEOF_WGINT == 4) || (SIZEOF_WGINT == 8) */
1445 /* n = -n would overflow because -n would evaluate to a
1446 wgint value larger than WGINT_MAX. Need to make n
1447 smaller and handle the last digit separately. */
1448 int last_digit = n % 10;
1449 /* The sign of n%10 is implementation-defined. */
1451 last_digit_char = '0' - last_digit;
1453 last_digit_char = '0' + last_digit;
1454 /* After n is made smaller, -n will not overflow. */
1462 /* Use the DIGITS_ macro appropriate for N's number of digits. That
1463 way printing any N is fully open-coded without a loop or jump.
1464 (Also see description of DIGITS_*.) */
1466 if (n < 10) DIGITS_1 (1);
1467 else if (n < 100) DIGITS_2 (10);
1468 else if (n < 1000) DIGITS_3 (100);
1469 else if (n < 10000) DIGITS_4 (1000);
1470 else if (n < 100000) DIGITS_5 (10000);
1471 else if (n < 1000000) DIGITS_6 (100000);
1472 else if (n < 10000000) DIGITS_7 (1000000);
1473 else if (n < 100000000) DIGITS_8 (10000000);
1474 else if (n < 1000000000) DIGITS_9 (100000000);
1475 #if SIZEOF_WGINT == 4
1476 /* wgint is 32 bits wide: no number has more than 10 digits. */
1477 else DIGITS_10 (1000000000);
1479 /* wgint is 64 bits wide: handle numbers with 9-19 decimal digits.
1480 Constants are constructed by compile-time multiplication to avoid
1481 dealing with different notations for 64-bit constants
1482 (nL/nLL/nI64, depending on the compiler and architecture). */
1483 else if (n < 10*(W)1000000000) DIGITS_10 (1000000000);
1484 else if (n < 100*(W)1000000000) DIGITS_11 (10*(W)1000000000);
1485 else if (n < 1000*(W)1000000000) DIGITS_12 (100*(W)1000000000);
1486 else if (n < 10000*(W)1000000000) DIGITS_13 (1000*(W)1000000000);
1487 else if (n < 100000*(W)1000000000) DIGITS_14 (10000*(W)1000000000);
1488 else if (n < 1000000*(W)1000000000) DIGITS_15 (100000*(W)1000000000);
1489 else if (n < 10000000*(W)1000000000) DIGITS_16 (1000000*(W)1000000000);
1490 else if (n < 100000000*(W)1000000000) DIGITS_17 (10000000*(W)1000000000);
1491 else if (n < 1000000000*(W)1000000000) DIGITS_18 (100000000*(W)1000000000);
1492 else DIGITS_19 (1000000000*(W)1000000000);
1495 if (last_digit_char)
1496 *p++ = last_digit_char;
1499 #endif /* (SIZEOF_WGINT == 4) || (SIZEOF_WGINT == 8) */
1506 #undef SPRINTF_WGINT
1529 /* Print NUMBER to a statically allocated string and return a pointer
1530 to the printed representation.
1532 This function is intended to be used in conjunction with printf.
1533 It is hard to portably print wgint values:
1534 a) you cannot use printf("%ld", number) because wgint can be long
1535 long on 32-bit machines with LFS.
1536 b) you cannot use printf("%lld", number) because NUMBER could be
1537 long on 32-bit machines without LFS, or on 64-bit machines,
1538 which do not require LFS. Also, Windows doesn't support %lld.
1539 c) you cannot use printf("%j", (int_max_t) number) because not all
1540 versions of printf support "%j", the most notable being the one
1542 d) you cannot #define WGINT_FMT to the appropriate format and use
1543 printf(WGINT_FMT, number) because that would break translations
1544 for user-visible messages, such as printf("Downloaded: %d
1547 What you should use instead is printf("%s", number_to_static_string
1550 CAVEAT: since the function returns pointers to static data, you
1551 must be careful to copy its result before calling it again.
1552 However, to make it more useful with printf, the function maintains
1553 an internal ring of static buffers to return. That way things like
1554 printf("%s %s", number_to_static_string (num1),
1555 number_to_static_string (num2)) work as expected. Three buffers
1556 are currently used, which means that "%s %s %s" will work, but "%s
1557 %s %s %s" won't. If you need to print more than three wgints,
1558 bump the RING_SIZE (or rethink your message.) */
1561 number_to_static_string (wgint number)
1563 static char ring[RING_SIZE][24];
1565 char *buf = ring[ringpos];
1566 number_to_string (buf, number);
1567 ringpos = (ringpos + 1) % RING_SIZE;
1571 /* Determine the width of the terminal we're running on. If that's
1572 not possible, return 0. */
1575 determine_screen_width (void)
1577 /* If there's a way to get the terminal size using POSIX
1578 tcgetattr(), somebody please tell me. */
1583 if (opt.lfilename != NULL)
1586 fd = fileno (stderr);
1587 if (ioctl (fd, TIOCGWINSZ, &wsz) < 0)
1588 return 0; /* most likely ENOTTY */
1591 #elif defined(WINDOWS)
1592 CONSOLE_SCREEN_BUFFER_INFO csbi;
1593 if (!GetConsoleScreenBufferInfo (GetStdHandle (STD_ERROR_HANDLE), &csbi))
1595 return csbi.dwSize.X;
1596 #else /* neither TIOCGWINSZ nor WINDOWS */
1598 #endif /* neither TIOCGWINSZ nor WINDOWS */
1601 /* Whether the rnd system (either rand or [dl]rand48) has been
1603 static int rnd_seeded;
1605 /* Return a random number between 0 and MAX-1, inclusive.
1607 If the system does not support lrand48 and MAX is greater than the
1608 value of RAND_MAX+1 on the system, the returned value will be in
1609 the range [0, RAND_MAX]. This may be fixed in a future release.
1610 The random number generator is seeded automatically the first time
1613 This uses lrand48 where available, rand elsewhere. DO NOT use it
1614 for cryptography. It is only meant to be used in situations where
1615 quality of the random numbers returned doesn't really matter. */
1618 random_number (int max)
1623 srand48 ((long) time (NULL) ^ (long) getpid ());
1626 return lrand48 () % max;
1627 #else /* not HAVE_DRAND48 */
1633 srand ((unsigned) time (NULL) ^ (unsigned) getpid ());
1638 /* Like rand() % max, but uses the high-order bits for better
1639 randomness on architectures where rand() is implemented using a
1640 simple congruential generator. */
1642 bounded = (double) max * rnd / (RAND_MAX + 1.0);
1643 return (int) bounded;
1645 #endif /* not HAVE_DRAND48 */
1648 /* Return a random uniformly distributed floating point number in the
1649 [0, 1) range. Uses drand48 where available, and a really lame
1650 kludge elsewhere. */
1658 srand48 ((long) time (NULL) ^ (long) getpid ());
1662 #else /* not HAVE_DRAND48 */
1663 return ( random_number (10000) / 10000.0
1664 + random_number (10000) / (10000.0 * 10000.0)
1665 + random_number (10000) / (10000.0 * 10000.0 * 10000.0)
1666 + random_number (10000) / (10000.0 * 10000.0 * 10000.0 * 10000.0));
1667 #endif /* not HAVE_DRAND48 */
1670 /* Implementation of run_with_timeout, a generic timeout-forcing
1671 routine for systems with Unix-like signal handling. */
1673 #ifdef USE_SIGNAL_TIMEOUT
1674 # ifdef HAVE_SIGSETJMP
1675 # define SETJMP(env) sigsetjmp (env, 1)
1677 static sigjmp_buf run_with_timeout_env;
1680 abort_run_with_timeout (int sig)
1682 assert (sig == SIGALRM);
1683 siglongjmp (run_with_timeout_env, -1);
1685 # else /* not HAVE_SIGSETJMP */
1686 # define SETJMP(env) setjmp (env)
1688 static jmp_buf run_with_timeout_env;
1691 abort_run_with_timeout (int sig)
1693 assert (sig == SIGALRM);
1694 /* We don't have siglongjmp to preserve the set of blocked signals;
1695 if we longjumped out of the handler at this point, SIGALRM would
1696 remain blocked. We must unblock it manually. */
1697 int mask = siggetmask ();
1698 mask &= ~sigmask (SIGALRM);
1701 /* Now it's safe to longjump. */
1702 longjmp (run_with_timeout_env, -1);
1704 # endif /* not HAVE_SIGSETJMP */
1706 /* Arrange for SIGALRM to be delivered in TIMEOUT seconds. This uses
1707 setitimer where available, alarm otherwise.
1709 TIMEOUT should be non-zero. If the timeout value is so small that
1710 it would be rounded to zero, it is rounded to the least legal value
1711 instead (1us for setitimer, 1s for alarm). That ensures that
1712 SIGALRM will be delivered in all cases. */
1715 alarm_set (double timeout)
1718 /* Use the modern itimer interface. */
1719 struct itimerval itv;
1721 itv.it_value.tv_sec = (long) timeout;
1722 itv.it_value.tv_usec = 1000000 * (timeout - (long)timeout);
1723 if (itv.it_value.tv_sec == 0 && itv.it_value.tv_usec == 0)
1724 /* Ensure that we wait for at least the minimum interval.
1725 Specifying zero would mean "wait forever". */
1726 itv.it_value.tv_usec = 1;
1727 setitimer (ITIMER_REAL, &itv, NULL);
1728 #else /* not ITIMER_REAL */
1729 /* Use the old alarm() interface. */
1730 int secs = (int) timeout;
1732 /* Round TIMEOUTs smaller than 1 to 1, not to zero. This is
1733 because alarm(0) means "never deliver the alarm", i.e. "wait
1734 forever", which is not what someone who specifies a 0.5s
1735 timeout would expect. */
1738 #endif /* not ITIMER_REAL */
1741 /* Cancel the alarm set with alarm_set. */
1747 struct itimerval disable;
1749 setitimer (ITIMER_REAL, &disable, NULL);
1750 #else /* not ITIMER_REAL */
1752 #endif /* not ITIMER_REAL */
1755 /* Call FUN(ARG), but don't allow it to run for more than TIMEOUT
1756 seconds. Returns true if the function was interrupted with a
1757 timeout, false otherwise.
1759 This works by setting up SIGALRM to be delivered in TIMEOUT seconds
1760 using setitimer() or alarm(). The timeout is enforced by
1761 longjumping out of the SIGALRM handler. This has several
1762 advantages compared to the traditional approach of relying on
1763 signals causing system calls to exit with EINTR:
1765 * The callback function is *forcibly* interrupted after the
1766 timeout expires, (almost) regardless of what it was doing and
1767 whether it was in a syscall. For example, a calculation that
1768 takes a long time is interrupted as reliably as an IO
1771 * It works with both SYSV and BSD signals because it doesn't
1772 depend on the default setting of SA_RESTART.
1774 * It doesn't require special handler setup beyond a simple call
1775 to signal(). (It does use sigsetjmp/siglongjmp, but they're
1778 The only downside is that, if FUN allocates internal resources that
1779 are normally freed prior to exit from the functions, they will be
1780 lost in case of timeout. */
1783 run_with_timeout (double timeout, void (*fun) (void *), void *arg)
1793 signal (SIGALRM, abort_run_with_timeout);
1794 if (SETJMP (run_with_timeout_env) != 0)
1796 /* Longjumped out of FUN with a timeout. */
1797 signal (SIGALRM, SIG_DFL);
1800 alarm_set (timeout);
1803 /* Preserve errno in case alarm() or signal() modifies it. */
1804 saved_errno = errno;
1806 signal (SIGALRM, SIG_DFL);
1807 errno = saved_errno;
1812 #else /* not USE_SIGNAL_TIMEOUT */
1815 /* A stub version of run_with_timeout that just calls FUN(ARG). Don't
1816 define it under Windows, because Windows has its own version of
1817 run_with_timeout that uses threads. */
1820 run_with_timeout (double timeout, void (*fun) (void *), void *arg)
1825 #endif /* not WINDOWS */
1826 #endif /* not USE_SIGNAL_TIMEOUT */
1830 /* Sleep the specified amount of seconds. On machines without
1831 nanosleep(), this may sleep shorter if interrupted by signals. */
1834 xsleep (double seconds)
1836 #ifdef HAVE_NANOSLEEP
1837 /* nanosleep is the preferred interface because it offers high
1838 accuracy and, more importantly, because it allows us to reliably
1839 restart receiving a signal such as SIGWINCH. (There was an
1840 actual Debian bug report about --limit-rate malfunctioning while
1841 the terminal was being resized.) */
1842 struct timespec sleep, remaining;
1843 sleep.tv_sec = (long) seconds;
1844 sleep.tv_nsec = 1000000000 * (seconds - (long) seconds);
1845 while (nanosleep (&sleep, &remaining) < 0 && errno == EINTR)
1846 /* If nanosleep has been interrupted by a signal, adjust the
1847 sleeping period and return to sleep. */
1849 #elif defined(HAVE_USLEEP)
1850 /* If usleep is available, use it in preference to select. */
1853 /* On some systems, usleep cannot handle values larger than
1854 1,000,000. If the period is larger than that, use sleep
1855 first, then add usleep for subsecond accuracy. */
1857 seconds -= (long) seconds;
1859 usleep (seconds * 1000000);
1860 #else /* fall back select */
1861 /* Note that, although Windows supports select, it can't be used to
1862 implement sleeping because Winsock's select doesn't implement
1863 timeout when it is passed NULL pointers for all fd sets. (But it
1864 does under Cygwin, which implements Unix-compatible select.) */
1865 struct timeval sleep;
1866 sleep.tv_sec = (long) seconds;
1867 sleep.tv_usec = 1000000 * (seconds - (long) seconds);
1868 select (0, NULL, NULL, NULL, &sleep);
1869 /* If select returns -1 and errno is EINTR, it means we were
1870 interrupted by a signal. But without knowing how long we've
1871 actually slept, we can't return to sleep. Using gettimeofday to
1872 track sleeps is slow and unreliable due to clock skew. */
1876 #endif /* not WINDOWS */
1878 /* Encode the octets in DATA of length LENGTH to base64 format,
1879 storing the result to DEST. The output will be zero-terminated,
1880 and must point to a writable buffer of at least
1881 1+BASE64_LENGTH(length) bytes. The function returns the length of
1882 the resulting base64 data, not counting the terminating zero.
1884 This implementation does not emit newlines after 76 characters of
1888 base64_encode (const void *data, int length, char *dest)
1890 /* Conversion table. */
1891 static const char tbl[64] = {
1892 'A','B','C','D','E','F','G','H','I','J','K','L','M','N','O','P',
1893 'Q','R','S','T','U','V','W','X','Y','Z','a','b','c','d','e','f',
1894 'g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v',
1895 'w','x','y','z','0','1','2','3','4','5','6','7','8','9','+','/'
1897 /* Access bytes in DATA as unsigned char, otherwise the shifts below
1898 don't work for data with MSB set. */
1899 const unsigned char *s = data;
1900 /* Theoretical ANSI violation when length < 3. */
1901 const unsigned char *end = (const unsigned char *) data + length - 2;
1904 /* Transform the 3x8 bits to 4x6 bits, as required by base64. */
1905 for (; s < end; s += 3)
1907 *p++ = tbl[s[0] >> 2];
1908 *p++ = tbl[((s[0] & 3) << 4) + (s[1] >> 4)];
1909 *p++ = tbl[((s[1] & 0xf) << 2) + (s[2] >> 6)];
1910 *p++ = tbl[s[2] & 0x3f];
1913 /* Pad the result if necessary... */
1917 *p++ = tbl[s[0] >> 2];
1918 *p++ = tbl[(s[0] & 3) << 4];
1923 *p++ = tbl[s[0] >> 2];
1924 *p++ = tbl[((s[0] & 3) << 4) + (s[1] >> 4)];
1925 *p++ = tbl[((s[1] & 0xf) << 2)];
1929 /* ...and zero-terminate it. */
1935 /* Store in C the next non-whitespace character from the string, or \0
1936 when end of string is reached. */
1937 #define NEXT_CHAR(c, p) do { \
1938 c = (unsigned char) *p++; \
1939 } while (ISSPACE (c))
1941 #define IS_ASCII(c) (((c) & 0x80) == 0)
1943 /* Decode data from BASE64 (a null-terminated string) into memory
1944 pointed to by DEST. DEST is assumed to be large enough to
1945 accomodate the decoded data, which is guaranteed to be no more than
1948 Since DEST is assumed to contain binary data, it is not
1949 NUL-terminated. The function returns the length of the data
1950 written to TO. -1 is returned in case of error caused by malformed
1953 This function originates from Free Recode. */
1956 base64_decode (const char *base64, void *dest)
1958 /* Table of base64 values for first 128 characters. Note that this
1959 assumes ASCII (but so does Wget in other places). */
1960 static const signed char base64_char_to_value[128] =
1962 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, /* 0- 9 */
1963 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, /* 10- 19 */
1964 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, /* 20- 29 */
1965 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, /* 30- 39 */
1966 -1, -1, -1, 62, -1, -1, -1, 63, 52, 53, /* 40- 49 */
1967 54, 55, 56, 57, 58, 59, 60, 61, -1, -1, /* 50- 59 */
1968 -1, -1, -1, -1, -1, 0, 1, 2, 3, 4, /* 60- 69 */
1969 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, /* 70- 79 */
1970 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, /* 80- 89 */
1971 25, -1, -1, -1, -1, -1, -1, 26, 27, 28, /* 90- 99 */
1972 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, /* 100-109 */
1973 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, /* 110-119 */
1974 49, 50, 51, -1, -1, -1, -1, -1 /* 120-127 */
1976 #define BASE64_CHAR_TO_VALUE(c) ((int) base64_char_to_value[c])
1977 #define IS_BASE64(c) ((IS_ASCII (c) && BASE64_CHAR_TO_VALUE (c) >= 0) || c == '=')
1979 const char *p = base64;
1985 unsigned long value;
1987 /* Process first byte of a quadruplet. */
1991 if (c == '=' || !IS_BASE64 (c))
1992 return -1; /* illegal char while decoding base64 */
1993 value = BASE64_CHAR_TO_VALUE (c) << 18;
1995 /* Process second byte of a quadruplet. */
1998 return -1; /* premature EOF while decoding base64 */
1999 if (c == '=' || !IS_BASE64 (c))
2000 return -1; /* illegal char while decoding base64 */
2001 value |= BASE64_CHAR_TO_VALUE (c) << 12;
2004 /* Process third byte of a quadruplet. */
2007 return -1; /* premature EOF while decoding base64 */
2009 return -1; /* illegal char while decoding base64 */
2015 return -1; /* premature EOF while decoding base64 */
2017 return -1; /* padding `=' expected but not found */
2021 value |= BASE64_CHAR_TO_VALUE (c) << 6;
2022 *q++ = 0xff & value >> 8;
2024 /* Process fourth byte of a quadruplet. */
2027 return -1; /* premature EOF while decoding base64 */
2031 return -1; /* illegal char while decoding base64 */
2033 value |= BASE64_CHAR_TO_VALUE (c);
2034 *q++ = 0xff & value;
2037 #undef BASE64_CHAR_TO_VALUE
2039 return q - (char *) dest;
2045 /* Simple merge sort for use by stable_sort. Implementation courtesy
2046 Zeljko Vrba with additional debugging by Nenad Barbutov. */
2049 mergesort_internal (void *base, void *temp, size_t size, size_t from, size_t to,
2050 int (*cmpfun) (const void *, const void *))
2052 #define ELT(array, pos) ((char *)(array) + (pos) * size)
2056 size_t mid = (to + from) / 2;
2057 mergesort_internal (base, temp, size, from, mid, cmpfun);
2058 mergesort_internal (base, temp, size, mid + 1, to, cmpfun);
2061 for (k = from; (i <= mid) && (j <= to); k++)
2062 if (cmpfun (ELT (base, i), ELT (base, j)) <= 0)
2063 memcpy (ELT (temp, k), ELT (base, i++), size);
2065 memcpy (ELT (temp, k), ELT (base, j++), size);
2067 memcpy (ELT (temp, k++), ELT (base, i++), size);
2069 memcpy (ELT (temp, k++), ELT (base, j++), size);
2070 for (k = from; k <= to; k++)
2071 memcpy (ELT (base, k), ELT (temp, k), size);
2076 /* Stable sort with interface exactly like standard library's qsort.
2077 Uses mergesort internally, allocating temporary storage with
2081 stable_sort (void *base, size_t nmemb, size_t size,
2082 int (*cmpfun) (const void *, const void *))
2086 void *temp = alloca (nmemb * size * sizeof (void *));
2087 mergesort_internal (base, temp, size, 0, nmemb - 1, cmpfun);
2091 /* Print a decimal number. If it is equal to or larger than ten, the
2092 number is rounded. Otherwise it is printed with one significant
2093 digit without trailing zeros and with no more than three fractional
2094 digits total. For example, 0.1 is printed as "0.1", 0.035 is
2095 printed as "0.04", 0.0091 as "0.009", and 0.0003 as simply "0".
2097 This is useful for displaying durations because it provides
2098 order-of-magnitude information without unnecessary clutter --
2099 long-running downloads are shown without the fractional part, and
2100 short ones still retain one significant digit. */
2103 print_decimal (double number)
2105 static char buf[32];
2106 double n = number >= 0 ? number : -number;
2109 /* Cut off at 9.95 because the below %.1f would round 9.96 to
2110 "10.0" instead of "10". OTOH 9.94 will print as "9.9". */
2111 snprintf (buf, sizeof buf, "%.0f", number);
2113 snprintf (buf, sizeof buf, "%.1f", number);
2114 else if (n >= 0.001)
2115 snprintf (buf, sizeof buf, "%.1g", number);
2116 else if (n >= 0.0005)
2117 /* round [0.0005, 0.001) to 0.001 */
2118 snprintf (buf, sizeof buf, "%.3f", number);
2120 /* print numbers close to 0 as 0, not 0.000 */
2137 { "/somedir", "/somedir", true },
2138 { "/somedir", "/somedir/d2", true },
2139 { "/somedir/d1", "/somedir", false },
2142 for (i = 0; i < countof(test_array); ++i)
2144 bool res = subdir_p (test_array[i].d1, test_array[i].d2);
2146 mu_assert ("test_subdir_p: wrong result",
2147 res == test_array[i].result);
2154 test_dir_matches_p()
2162 { { "/somedir", "/someotherdir", NULL }, "somedir", true },
2163 { { "/somedir", "/someotherdir", NULL }, "anotherdir", false },
2164 { { "/somedir", "/*otherdir", NULL }, "anotherdir", true },
2165 { { "/somedir/d1", "/someotherdir", NULL }, "somedir/d1", true },
2166 { { "/somedir/d1", "/someotherdir", NULL }, "d1", false },
2169 for (i = 0; i < countof(test_array); ++i)
2171 bool res = dir_matches_p (test_array[i].dirlist, test_array[i].dir);
2173 mu_assert ("test_dir_matches_p: wrong result",
2174 res == test_array[i].result);
2180 #endif /* TESTING */