1 /* Reading/parsing the initialization file.
2 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
3 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012 Free Software Foundation,
6 This file is part of GNU Wget.
8 GNU Wget is free software; you can redistribute it and/or modify
9 it under the terms of the GNU General Public License as published by
10 the Free Software Foundation; either version 3 of the License, or
11 (at your option) any later version.
13 GNU Wget is distributed in the hope that it will be useful,
14 but WITHOUT ANY WARRANTY; without even the implied warranty of
15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 GNU General Public License for more details.
18 You should have received a copy of the GNU General Public License
19 along with Wget. If not, see <http://www.gnu.org/licenses/>.
21 Additional permission under GNU GPL version 3 section 7
23 If you modify this program, or any covered work, by linking or
24 combining it with the OpenSSL project's OpenSSL library (or a
25 modified version of that library), containing parts covered by the
26 terms of the OpenSSL or SSLeay licenses, the Free Software Foundation
27 grants you additional permission to convey the resulting work.
28 Corresponding Source for a non-source form of such a combination
29 shall include the source code for the parts of OpenSSL used as well
30 as that of the covered work. */
42 /* not all systems provide PATH_MAX in limits.h */
44 # include <sys/param.h>
46 # define PATH_MAX MAXPATHLEN
65 #include "recur.h" /* for INFINITE_RECURSION */
66 #include "convert.h" /* for convert_cleanup */
67 #include "res.h" /* for res_cleanup */
68 #include "http.h" /* for http_cleanup */
69 #include "retr.h" /* for output_stream */
70 #include "warc.h" /* for warc_close */
78 #define CMD_DECLARE(func) static bool func (const char *, const char *, void *)
80 CMD_DECLARE (cmd_boolean);
81 CMD_DECLARE (cmd_bytes);
82 CMD_DECLARE (cmd_bytes_sum);
84 CMD_DECLARE (cmd_cert_type);
86 CMD_DECLARE (cmd_directory_vector);
87 CMD_DECLARE (cmd_number);
88 CMD_DECLARE (cmd_number_inf);
89 CMD_DECLARE (cmd_string);
90 CMD_DECLARE (cmd_string_uppercase);
91 CMD_DECLARE (cmd_file);
92 CMD_DECLARE (cmd_directory);
93 CMD_DECLARE (cmd_time);
94 CMD_DECLARE (cmd_vector);
96 CMD_DECLARE (cmd_spec_dirstruct);
97 CMD_DECLARE (cmd_spec_header);
98 CMD_DECLARE (cmd_spec_warc_header);
99 CMD_DECLARE (cmd_spec_htmlify);
100 CMD_DECLARE (cmd_spec_mirror);
101 CMD_DECLARE (cmd_spec_prefer_family);
102 CMD_DECLARE (cmd_spec_progress);
103 CMD_DECLARE (cmd_spec_recursive);
104 CMD_DECLARE (cmd_spec_regex_type);
105 CMD_DECLARE (cmd_spec_restrict_file_names);
106 CMD_DECLARE (cmd_spec_report_speed);
108 CMD_DECLARE (cmd_spec_secure_protocol);
110 CMD_DECLARE (cmd_spec_timeout);
111 CMD_DECLARE (cmd_spec_useragent);
112 CMD_DECLARE (cmd_spec_verbose);
114 /* List of recognized commands, each consisting of name, place and
115 function. When adding a new command, simply add it to the list,
116 but be sure to keep the list sorted alphabetically, as
117 command_by_name's binary search depends on it. Also, be sure to
118 add any entries that allocate memory (e.g. cmd_string and
119 cmd_vector) to the cleanup() function below. */
121 static const struct {
124 bool (*action) (const char *, const char *, void *);
126 /* KEEP THIS LIST ALPHABETICALLY SORTED */
127 { "accept", &opt.accepts, cmd_vector },
128 { "acceptregex", &opt.acceptregex_s, cmd_string },
129 { "addhostdir", &opt.add_hostdir, cmd_boolean },
130 { "adjustextension", &opt.adjust_extension, cmd_boolean },
131 { "alwaysrest", &opt.always_rest, cmd_boolean }, /* deprecated */
132 { "askpassword", &opt.ask_passwd, cmd_boolean },
133 { "authnochallenge", &opt.auth_without_challenge,
135 { "background", &opt.background, cmd_boolean },
136 { "backupconverted", &opt.backup_converted, cmd_boolean },
137 { "backups", &opt.backups, cmd_number },
138 { "base", &opt.base_href, cmd_string },
139 { "bindaddress", &opt.bind_address, cmd_string },
140 { "bodydata", &opt.body_data, cmd_string },
141 { "bodyfile", &opt.body_file, cmd_string },
143 { "cacertificate", &opt.ca_cert, cmd_file },
145 { "cache", &opt.allow_cache, cmd_boolean },
147 { "cadirectory", &opt.ca_directory, cmd_directory },
148 { "certificate", &opt.cert_file, cmd_file },
149 { "certificatetype", &opt.cert_type, cmd_cert_type },
150 { "checkcertificate", &opt.check_cert, cmd_boolean },
152 { "chooseconfig", &opt.choose_config, cmd_file },
153 { "connecttimeout", &opt.connect_timeout, cmd_time },
154 { "contentdisposition", &opt.content_disposition, cmd_boolean },
155 { "contentonerror", &opt.content_on_error, cmd_boolean },
156 { "continue", &opt.always_rest, cmd_boolean },
157 { "convertlinks", &opt.convert_links, cmd_boolean },
158 { "cookies", &opt.cookies, cmd_boolean },
159 { "cutdirs", &opt.cut_dirs, cmd_number },
161 { "debug", &opt.debug, cmd_boolean },
163 { "defaultpage", &opt.default_page, cmd_string },
164 { "deleteafter", &opt.delete_after, cmd_boolean },
165 { "dirprefix", &opt.dir_prefix, cmd_directory },
166 { "dirstruct", NULL, cmd_spec_dirstruct },
167 { "dnscache", &opt.dns_cache, cmd_boolean },
168 { "dnstimeout", &opt.dns_timeout, cmd_time },
169 { "domains", &opt.domains, cmd_vector },
170 { "dotbytes", &opt.dot_bytes, cmd_bytes },
171 { "dotsinline", &opt.dots_in_line, cmd_number },
172 { "dotspacing", &opt.dot_spacing, cmd_number },
173 { "dotstyle", &opt.dot_style, cmd_string }, /* deprecated */
175 { "egdfile", &opt.egd_file, cmd_file },
177 { "excludedirectories", &opt.excludes, cmd_directory_vector },
178 { "excludedomains", &opt.exclude_domains, cmd_vector },
179 { "followftp", &opt.follow_ftp, cmd_boolean },
180 { "followtags", &opt.follow_tags, cmd_vector },
181 { "forcehtml", &opt.force_html, cmd_boolean },
182 { "ftppasswd", &opt.ftp_passwd, cmd_string }, /* deprecated */
183 { "ftppassword", &opt.ftp_passwd, cmd_string },
184 { "ftpproxy", &opt.ftp_proxy, cmd_string },
186 { "ftpstmlf", &opt.ftp_stmlf, cmd_boolean },
187 #endif /* def __VMS */
188 { "ftpuser", &opt.ftp_user, cmd_string },
189 { "glob", &opt.ftp_glob, cmd_boolean },
190 { "header", NULL, cmd_spec_header },
191 { "htmlextension", &opt.adjust_extension, cmd_boolean }, /* deprecated */
192 { "htmlify", NULL, cmd_spec_htmlify },
193 { "httpkeepalive", &opt.http_keep_alive, cmd_boolean },
194 { "httppasswd", &opt.http_passwd, cmd_string }, /* deprecated */
195 { "httppassword", &opt.http_passwd, cmd_string },
196 { "httpproxy", &opt.http_proxy, cmd_string },
198 { "httpsonly", &opt.https_only, cmd_boolean },
200 { "httpsproxy", &opt.https_proxy, cmd_string },
201 { "httpuser", &opt.http_user, cmd_string },
202 { "ignorecase", &opt.ignore_case, cmd_boolean },
203 { "ignorelength", &opt.ignore_length, cmd_boolean },
204 { "ignoretags", &opt.ignore_tags, cmd_vector },
205 { "includedirectories", &opt.includes, cmd_directory_vector },
207 { "inet4only", &opt.ipv4_only, cmd_boolean },
208 { "inet6only", &opt.ipv6_only, cmd_boolean },
210 { "input", &opt.input_filename, cmd_file },
211 { "iri", &opt.enable_iri, cmd_boolean },
212 { "keepsessioncookies", &opt.keep_session_cookies, cmd_boolean },
213 { "limitrate", &opt.limit_rate, cmd_bytes },
214 { "loadcookies", &opt.cookies_input, cmd_file },
215 { "localencoding", &opt.locale, cmd_string },
216 { "logfile", &opt.lfilename, cmd_file },
217 { "login", &opt.ftp_user, cmd_string },/* deprecated*/
218 { "maxredirect", &opt.max_redirect, cmd_number },
219 { "method", &opt.method, cmd_string_uppercase },
220 { "mirror", NULL, cmd_spec_mirror },
221 { "netrc", &opt.netrc, cmd_boolean },
222 { "noclobber", &opt.noclobber, cmd_boolean },
223 { "noparent", &opt.no_parent, cmd_boolean },
224 { "noproxy", &opt.no_proxy, cmd_vector },
225 { "numtries", &opt.ntry, cmd_number_inf },/* deprecated*/
226 { "outputdocument", &opt.output_document, cmd_file },
227 { "pagerequisites", &opt.page_requisites, cmd_boolean },
228 { "passiveftp", &opt.ftp_pasv, cmd_boolean },
229 { "passwd", &opt.ftp_passwd, cmd_string },/* deprecated*/
230 { "password", &opt.passwd, cmd_string },
231 { "postdata", &opt.post_data, cmd_string },
232 { "postfile", &opt.post_file_name, cmd_file },
233 { "preferfamily", NULL, cmd_spec_prefer_family },
234 { "preservepermissions", &opt.preserve_perm, cmd_boolean },
236 { "privatekey", &opt.private_key, cmd_file },
237 { "privatekeytype", &opt.private_key_type, cmd_cert_type },
239 { "progress", &opt.progress_type, cmd_spec_progress },
240 { "protocoldirectories", &opt.protocol_directories, cmd_boolean },
241 { "proxypasswd", &opt.proxy_passwd, cmd_string }, /* deprecated */
242 { "proxypassword", &opt.proxy_passwd, cmd_string },
243 { "proxyuser", &opt.proxy_user, cmd_string },
244 { "quiet", &opt.quiet, cmd_boolean },
245 { "quota", &opt.quota, cmd_bytes_sum },
247 { "randomfile", &opt.random_file, cmd_file },
249 { "randomwait", &opt.random_wait, cmd_boolean },
250 { "readtimeout", &opt.read_timeout, cmd_time },
251 { "reclevel", &opt.reclevel, cmd_number_inf },
252 { "recursive", NULL, cmd_spec_recursive },
253 { "referer", &opt.referer, cmd_string },
254 { "regextype", &opt.regex_type, cmd_spec_regex_type },
255 { "reject", &opt.rejects, cmd_vector },
256 { "rejectregex", &opt.rejectregex_s, cmd_string },
257 { "relativeonly", &opt.relative_only, cmd_boolean },
258 { "remoteencoding", &opt.encoding_remote, cmd_string },
259 { "removelisting", &opt.remove_listing, cmd_boolean },
260 { "reportspeed", &opt.report_bps, cmd_spec_report_speed},
261 { "restrictfilenames", NULL, cmd_spec_restrict_file_names },
262 { "retrsymlinks", &opt.retr_symlinks, cmd_boolean },
263 { "retryconnrefused", &opt.retry_connrefused, cmd_boolean },
264 { "robots", &opt.use_robots, cmd_boolean },
265 { "savecookies", &opt.cookies_output, cmd_file },
266 { "saveheaders", &opt.save_headers, cmd_boolean },
268 { "secureprotocol", &opt.secure_protocol, cmd_spec_secure_protocol },
270 { "serverresponse", &opt.server_response, cmd_boolean },
271 { "showalldnsentries", &opt.show_all_dns_entries, cmd_boolean },
272 { "spanhosts", &opt.spanhost, cmd_boolean },
273 { "spider", &opt.spider, cmd_boolean },
274 { "strictcomments", &opt.strict_comments, cmd_boolean },
275 { "timeout", NULL, cmd_spec_timeout },
276 { "timestamping", &opt.timestamping, cmd_boolean },
277 { "tries", &opt.ntry, cmd_number_inf },
278 { "trustservernames", &opt.trustservernames, cmd_boolean },
279 { "unlink", &opt.unlink, cmd_boolean },
280 { "useproxy", &opt.use_proxy, cmd_boolean },
281 { "user", &opt.user, cmd_string },
282 { "useragent", NULL, cmd_spec_useragent },
283 { "useservertimestamps", &opt.useservertimestamps, cmd_boolean },
284 { "verbose", NULL, cmd_spec_verbose },
285 { "wait", &opt.wait, cmd_time },
286 { "waitretry", &opt.waitretry, cmd_time },
287 { "warccdx", &opt.warc_cdx_enabled, cmd_boolean },
288 { "warccdxdedup", &opt.warc_cdx_dedup_filename, cmd_file },
290 { "warccompression", &opt.warc_compression_enabled, cmd_boolean },
292 { "warcdigests", &opt.warc_digests_enabled, cmd_boolean },
293 { "warcfile", &opt.warc_filename, cmd_file },
294 { "warcheader", NULL, cmd_spec_warc_header },
295 { "warckeeplog", &opt.warc_keep_log, cmd_boolean },
296 { "warcmaxsize", &opt.warc_maxsize, cmd_bytes },
297 { "warctempdir", &opt.warc_tempdir, cmd_directory },
299 { "wdebug", &opt.wdebug, cmd_boolean },
303 /* Look up CMDNAME in the commands[] and return its position in the
304 array. If CMDNAME is not found, return -1. */
307 command_by_name (const char *cmdname)
309 /* Use binary search for speed. Wget has ~100 commands, which
310 guarantees a worst case performance of 7 string comparisons. */
311 int lo = 0, hi = countof (commands) - 1;
315 int mid = (lo + hi) >> 1;
316 int cmp = strcasecmp (cmdname, commands[mid].name);
327 /* Reset the variables to default values. */
333 /* Most of the default values are 0 (and 0.0, NULL, and false).
334 Just reset everything, and fill in the non-zero values. Note
335 that initializing pointers to NULL this way is technically
336 illegal, but porting Wget to a machine where NULL is not all-zero
337 bit pattern will be the least of the implementors' worries. */
344 opt.add_hostdir = true;
348 opt.http_keep_alive = true;
349 opt.use_proxy = true;
350 tmp = getenv ("no_proxy");
352 opt.no_proxy = sepstring (tmp);
353 opt.prefer_family = prefer_none;
354 opt.allow_cache = true;
356 opt.read_timeout = 900;
357 opt.use_robots = true;
359 opt.remove_listing = true;
361 opt.dot_bytes = 1024;
362 opt.dot_spacing = 10;
363 opt.dots_in_line = 50;
365 opt.dns_cache = true;
369 opt.check_cert = true;
372 /* The default for file name restriction defaults to the OS type. */
373 #if defined(WINDOWS) || defined(MSDOS) || defined(__CYGWIN__)
374 opt.restrict_files_os = restrict_windows;
376 opt.restrict_files_os = restrict_unix;
378 opt.restrict_files_ctrl = true;
379 opt.restrict_files_nonascii = false;
380 opt.restrict_files_case = restrict_no_case_restriction;
382 opt.regex_type = regex_type_posix;
384 opt.max_redirect = 20;
389 opt.enable_iri = true;
391 opt.enable_iri = false;
394 opt.encoding_remote = NULL;
396 opt.useservertimestamps = true;
397 opt.show_all_dns_entries = false;
399 opt.warc_maxsize = 0; /* 1024 * 1024 * 1024; */
401 opt.warc_compression_enabled = true;
403 opt.warc_compression_enabled = false;
405 opt.warc_digests_enabled = true;
406 opt.warc_cdx_enabled = false;
407 opt.warc_cdx_dedup_filename = NULL;
408 opt.warc_tempdir = NULL;
409 opt.warc_keep_log = true;
412 /* Return the user's home directory (strdup-ed), or NULL if none is
417 static char *buf = NULL;
418 static char *home, *ret;
422 home = getenv ("HOME");
428 /* Under MSDOS, if $HOME isn't defined, use the directory where
429 `wget.exe' resides. */
430 const char *_w32_get_argv0 (void); /* in libwatt.a/pcconfig.c */
433 buff = _w32_get_argv0 ();
435 p = strrchr (buf, '/'); /* djgpp */
437 p = strrchr (buf, '\\'); /* others */
441 buff = malloc (len + 1);
445 strncpy (buff, _w32_get_argv0 (), len);
449 #elif !defined(WINDOWS)
450 /* If HOME is not defined, try getting it from the password
452 struct passwd *pwd = getpwuid (getuid ());
453 if (!pwd || !pwd->pw_dir)
457 /* Under Windows, if $HOME isn't defined, use the directory where
458 `wget.exe' resides. */
464 ret = home ? xstrdup (home) : NULL;
470 /* Check the 'WGETRC' environment variable and return the file name
471 if 'WGETRC' is set and is a valid file.
472 If the `WGETRC' variable exists but the file does not exist, the
473 function will exit(). */
475 wgetrc_env_file_name (void)
477 char *env = getenv ("WGETRC");
480 if (!file_exists_p (env))
482 fprintf (stderr, _("%s: WGETRC points to %s, which doesn't exist.\n"),
486 return xstrdup (env);
491 /* Check for the existance of '$HOME/.wgetrc' and return its path
492 if it exists and is set. */
494 wgetrc_user_file_name (void)
498 /* If that failed, try $HOME/.wgetrc (or equivalent). */
501 file = "SYS$LOGIN:.wgetrc";
502 #else /* def __VMS */
505 file = aprintf ("%s/.wgetrc", home);
507 #endif /* def __VMS [else] */
511 if (!file_exists_p (file))
519 /* Return the path to the user's .wgetrc. This is either the value of
520 `WGETRC' environment variable, or `$HOME/.wgetrc'.
522 Additionally, for windows, look in the directory where wget.exe
525 wgetrc_file_name (void)
527 char *file = wgetrc_env_file_name ();
531 file = wgetrc_user_file_name ();
534 /* Under Windows, if we still haven't found .wgetrc, look for the file
535 `wget.ini' in the directory where `wget.exe' resides; we do this for
536 backward compatibility with previous versions of Wget.
537 SYSTEM_WGETRC should not be defined under WINDOWS. */
540 char *home = home_dir ();
546 file = aprintf ("%s/wget.ini", home);
547 if (!file_exists_p (file))
560 /* Return values of parse_line. */
568 static enum parse_line parse_line (const char *, char **, char **, int *);
569 static bool setval_internal (int, const char *, const char *);
570 static bool setval_internal_tilde (int, const char *, const char *);
572 /* Initialize variables from a wgetrc file. Returns zero (failure) if
573 there were errors in the file. */
576 run_wgetrc (const char *file)
584 fp = fopen (file, "r");
587 fprintf (stderr, _("%s: Cannot read %s (%s).\n"), exec_name,
588 file, strerror (errno));
589 return true; /* not a fatal error */
592 while (getline (&line, &bufsize, fp) > 0)
594 char *com = NULL, *val = NULL;
597 /* Parse the line. */
598 switch (parse_line (line, &com, &val, &comind))
601 /* If everything is OK, set the value. */
602 if (!setval_internal_tilde (comind, com, val))
604 fprintf (stderr, _("%s: Error in %s at line %d.\n"),
605 exec_name, file, ln);
609 case line_syntax_error:
610 fprintf (stderr, _("%s: Syntax error in %s at line %d.\n"),
611 exec_name, file, ln);
614 case line_unknown_command:
615 fprintf (stderr, _("%s: Unknown command %s in %s at line %d.\n"),
616 exec_name, quote (com), file, ln);
634 /* Initialize the defaults and run the system wgetrc and user's own
639 char *file, *env_sysrc;
642 /* Run a non-standard system rc file when the according environment
643 variable has been set. For internal testing purposes only! */
644 env_sysrc = getenv ("SYSTEM_WGETRC");
645 if (env_sysrc && file_exists_p (env_sysrc))
647 ok &= run_wgetrc (env_sysrc);
648 /* If there are any problems parsing the system wgetrc file, tell
652 fprintf (stderr, _("\
653 Parsing system wgetrc file (env SYSTEM_WGETRC) failed. Please check\n\
655 or specify a different file using --config.\n"), env_sysrc);
659 /* Otherwise, if SYSTEM_WGETRC is defined, use it. */
661 else if (file_exists_p (SYSTEM_WGETRC))
662 ok &= run_wgetrc (SYSTEM_WGETRC);
663 /* If there are any problems parsing the system wgetrc file, tell
667 fprintf (stderr, _("\
668 Parsing system wgetrc file failed. Please check\n\
670 or specify a different file using --config.\n"), SYSTEM_WGETRC);
674 /* Override it with your own, if one exists. */
675 file = wgetrc_file_name ();
678 /* #### We should canonicalize `file' and SYSTEM_WGETRC with
679 something like realpath() before comparing them with `strcmp' */
681 if (!strcmp (file, SYSTEM_WGETRC))
683 fprintf (stderr, _("\
684 %s: Warning: Both system and user wgetrc point to %s.\n"),
685 exec_name, quote (file));
689 ok &= run_wgetrc (file);
691 /* If there were errors processing either `.wgetrc', abort. */
699 /* Remove dashes and underscores from S, modifying S in the
705 char *t = s; /* t - tortoise */
706 char *h = s; /* h - hare */
708 if (*h == '_' || *h == '-')
715 /* Parse the line pointed by line, with the syntax:
716 <sp>* command <sp>* = <sp>* value <sp>*
717 Uses malloc to allocate space for command and value.
719 Returns one of line_ok, line_empty, line_syntax_error, or
720 line_unknown_command.
722 In case of line_ok, *COM and *VAL point to freshly allocated
723 strings, and *COMIND points to com's index. In case of error or
724 empty line, their values are unmodified. */
726 static enum parse_line
727 parse_line (const char *line, char **com, char **val, int *comind)
730 const char *end = line + strlen (line);
731 const char *cmdstart, *cmdend;
732 const char *valstart, *valend;
737 /* Skip leading and trailing whitespace. */
738 while (*line && c_isspace (*line))
740 while (end > line && c_isspace (end[-1]))
743 /* Skip empty lines and comments. */
744 if (!*line || *line == '#')
750 while (p < end && (c_isalnum (*p) || *p == '_' || *p == '-'))
754 /* Skip '=', as well as any space before or after it. */
755 while (p < end && c_isspace (*p))
757 if (p == end || *p != '=')
758 return line_syntax_error;
760 while (p < end && c_isspace (*p))
766 /* The syntax is valid (even though the command might not be). Fill
767 in the command name and value. */
768 *com = strdupdelim (cmdstart, cmdend);
769 *val = strdupdelim (valstart, valend);
771 /* The line now known to be syntactically correct. Check whether
772 the command is valid. */
773 BOUNDED_TO_ALLOCA (cmdstart, cmdend, cmdcopy);
775 ind = command_by_name (cmdcopy);
777 return line_unknown_command;
779 /* Report success to the caller. */
784 #if defined(WINDOWS) || defined(MSDOS)
785 # define ISSEP(c) ((c) == '/' || (c) == '\\')
787 # define ISSEP(c) ((c) == '/')
790 /* Run commands[comind].action. */
793 setval_internal (int comind, const char *com, const char *val)
795 assert (0 <= comind && ((size_t) comind) < countof (commands));
796 DEBUGP (("Setting %s (%s) to %s\n", com, commands[comind].name, val));
797 return commands[comind].action (com, val, commands[comind].place);
801 setval_internal_tilde (int comind, const char *com, const char *val)
807 ret = setval_internal (comind, com, val);
809 /* We make tilde expansion for cmd_file and cmd_directory */
810 if (((commands[comind].action == cmd_file) ||
811 (commands[comind].action == cmd_directory))
812 && ret && (*val == '~' && ISSEP (val[1])))
814 pstring = commands[comind].place;
818 homelen = strlen (home);
819 while (homelen && ISSEP (home[homelen - 1]))
820 home[--homelen] = '\0';
822 /* Skip the leading "~/". */
823 for (++val; ISSEP (*val); val++)
825 *pstring = concat_strings (home, "/", val, (char *)0);
831 /* Run command COM with value VAL. If running the command produces an
832 error, report the error and exit.
834 This is intended to be called from main() to modify Wget's behavior
835 through command-line switches. Since COM is hard-coded in main(),
836 it is not canonicalized, and this aborts when COM is not found.
838 If COMIND's are exported to init.h, this function will be changed
839 to accept COMIND directly. */
842 setoptval (const char *com, const char *val, const char *optname)
844 /* Prepend "--" to OPTNAME. */
845 char *dd_optname = (char *) alloca (2 + strlen (optname) + 1);
848 strcpy (dd_optname + 2, optname);
850 assert (val != NULL);
851 if (!setval_internal (command_by_name (com), dd_optname, val))
855 /* Parse OPT into command and value and run it. For example,
856 run_command("foo=bar") is equivalent to setoptval("foo", "bar").
857 This is used by the `--execute' flag in main.c. */
860 run_command (const char *opt)
864 switch (parse_line (opt, &com, &val, &comind))
867 if (!setval_internal (comind, com, val))
873 fprintf (stderr, _("%s: Invalid --execute command %s\n"),
874 exec_name, quote (opt));
879 /* Generic helper functions, for use with `commands'. */
881 /* Forward declarations: */
886 static bool decode_string (const char *, const struct decode_item *, int, int *);
887 static bool simple_atoi (const char *, const char *, int *);
888 static bool simple_atof (const char *, const char *, double *);
890 #define CMP1(p, c0) (c_tolower((p)[0]) == (c0) && (p)[1] == '\0')
892 #define CMP2(p, c0, c1) (c_tolower((p)[0]) == (c0) \
893 && c_tolower((p)[1]) == (c1) \
896 #define CMP3(p, c0, c1, c2) (c_tolower((p)[0]) == (c0) \
897 && c_tolower((p)[1]) == (c1) \
898 && c_tolower((p)[2]) == (c2) \
902 /* Store the boolean value from VAL to PLACE. COM is ignored,
903 except for error messages. */
905 cmd_boolean (const char *com, const char *val, void *place)
909 if (CMP2 (val, 'o', 'n') || CMP3 (val, 'y', 'e', 's') || CMP1 (val, '1'))
910 /* "on", "yes" and "1" mean true. */
912 else if (CMP3 (val, 'o', 'f', 'f') || CMP2 (val, 'n', 'o') || CMP1 (val, '0'))
913 /* "off", "no" and "0" mean false. */
918 _("%s: %s: Invalid boolean %s; use `on' or `off'.\n"),
919 exec_name, com, quote (val));
923 *(bool *) place = value;
927 /* Set the non-negative integer value from VAL to PLACE. With
928 incorrect specification, the number remains unchanged. */
930 cmd_number (const char *com, const char *val, void *place)
932 if (!simple_atoi (val, val + strlen (val), place)
933 || *(int *) place < 0)
935 fprintf (stderr, _("%s: %s: Invalid number %s.\n"),
936 exec_name, com, quote (val));
942 /* Similar to cmd_number(), only accepts `inf' as a synonym for 0. */
944 cmd_number_inf (const char *com, const char *val, void *place)
946 if (!strcasecmp (val, "inf"))
951 return cmd_number (com, val, place);
954 /* Copy (strdup) the string at COM to a new location and place a
955 pointer to *PLACE. */
957 cmd_string (const char *com, const char *val, void *place)
959 char **pstring = (char **)place;
961 xfree_null (*pstring);
962 *pstring = xstrdup (val);
966 /* Like cmd_string but ensure the string is upper case. */
968 cmd_string_uppercase (const char *com, const char *val, void *place)
971 pstring = (char **)place;
972 xfree_null (*pstring);
974 *pstring = xmalloc (strlen (val) + 1);
976 for (q = *pstring; *val; val++, q++)
977 *q = c_toupper (*val);
984 /* Like cmd_string, but handles tilde-expansion when reading a user's
985 `.wgetrc'. In that case, and if VAL begins with `~', the tilde
986 gets expanded to the user's home directory. */
988 cmd_file (const char *com, const char *val, void *place)
990 char **pstring = (char **)place;
992 xfree_null (*pstring);
994 /* #### If VAL is empty, perhaps should set *PLACE to NULL. */
996 *pstring = xstrdup (val);
998 #if defined(WINDOWS) || defined(MSDOS)
999 /* Convert "\" to "/". */
1002 for (s = *pstring; *s; s++)
1010 /* Like cmd_file, but strips trailing '/' characters. */
1012 cmd_directory (const char *com, const char *val, void *place)
1016 /* Call cmd_file() for tilde expansion and separator
1017 canonicalization (backslash -> slash under Windows). These
1018 things should perhaps be in a separate function. */
1019 if (!cmd_file (com, val, place))
1022 s = *(char **)place;
1024 while (t > s && *--t == '/')
1030 /* Split VAL by space to a vector of values, and append those values
1031 to vector pointed to by the PLACE argument. If VAL is empty, the
1032 PLACE vector is cleared instead. */
1035 cmd_vector (const char *com, const char *val, void *place)
1037 char ***pvec = (char ***)place;
1040 *pvec = merge_vecs (*pvec, sepstring (val));
1050 cmd_directory_vector (const char *com, const char *val, void *place)
1052 char ***pvec = (char ***)place;
1056 /* Strip the trailing slashes from directories. */
1059 seps = sepstring (val);
1060 for (t = seps; t && *t; t++)
1062 int len = strlen (*t);
1063 /* Skip degenerate case of root directory. */
1066 if ((*t)[len - 1] == '/')
1067 (*t)[len - 1] = '\0';
1070 *pvec = merge_vecs (*pvec, seps);
1080 /* Engine for cmd_bytes and cmd_bytes_sum: converts a string such as
1081 "100k" or "2.5G" to a floating point number. */
1084 parse_bytes_helper (const char *val, double *result)
1086 double number, mult;
1087 const char *end = val + strlen (val);
1089 /* Check for "inf". */
1090 if (0 == strcmp (val, "inf"))
1096 /* Strip trailing whitespace. */
1097 while (val < end && c_isspace (end[-1]))
1102 switch (c_tolower (end[-1]))
1105 --end, mult = 1024.0;
1108 --end, mult = 1048576.0;
1111 --end, mult = 1073741824.0;
1114 --end, mult = 1099511627776.0;
1117 /* Not a recognized suffix: assume it's a digit. (If not,
1118 simple_atof will raise an error.) */
1122 /* Skip leading and trailing whitespace. */
1123 while (val < end && c_isspace (*val))
1125 while (val < end && c_isspace (end[-1]))
1130 if (!simple_atof (val, end, &number) || number < 0)
1133 *result = number * mult;
1137 /* Parse VAL as a number and set its value to PLACE (which should
1140 By default, the value is assumed to be in bytes. If "K", "M", or
1141 "G" are appended, the value is multiplied with 1<<10, 1<<20, or
1142 1<<30, respectively. Floating point values are allowed and are
1143 cast to integer before use. The idea is to be able to use things
1144 like 1.5k instead of "1536".
1146 The string "inf" is returned as 0.
1148 In case of error, false is returned and memory pointed to by PLACE
1149 remains unmodified. */
1152 cmd_bytes (const char *com, const char *val, void *place)
1155 if (!parse_bytes_helper (val, &byte_value))
1157 fprintf (stderr, _("%s: %s: Invalid byte value %s\n"),
1158 exec_name, com, quote (val));
1161 *(wgint *)place = (wgint)byte_value;
1165 /* Like cmd_bytes, but PLACE is interpreted as a pointer to
1166 SIZE_SUM. It works by converting the string to double, therefore
1167 working with values up to 2^53-1 without loss of precision. This
1168 value (8192 TB) is large enough to serve for a while. */
1171 cmd_bytes_sum (const char *com, const char *val, void *place)
1174 if (!parse_bytes_helper (val, &byte_value))
1176 fprintf (stderr, _("%s: %s: Invalid byte value %s\n"),
1177 exec_name, com, quote (val));
1180 *(SUM_SIZE_INT *) place = (SUM_SIZE_INT) byte_value;
1184 /* Store the value of VAL to *OUT. The value is a time period, by
1185 default expressed in seconds, but also accepting suffixes "m", "h",
1186 "d", and "w" for minutes, hours, days, and weeks respectively. */
1189 cmd_time (const char *com, const char *val, void *place)
1191 double number, mult;
1192 const char *end = val + strlen (val);
1194 /* Strip trailing whitespace. */
1195 while (val < end && c_isspace (end[-1]))
1201 fprintf (stderr, _("%s: %s: Invalid time period %s\n"),
1202 exec_name, com, quote (val));
1206 switch (c_tolower (end[-1]))
1209 --end, mult = 1; /* seconds */
1212 --end, mult = 60; /* minutes */
1215 --end, mult = 3600; /* hours */
1218 --end, mult = 86400.0; /* days */
1221 --end, mult = 604800.0; /* weeks */
1224 /* Not a recognized suffix: assume it belongs to the number.
1225 (If not, simple_atof will raise an error.) */
1229 /* Skip leading and trailing whitespace. */
1230 while (val < end && c_isspace (*val))
1232 while (val < end && c_isspace (end[-1]))
1237 if (!simple_atof (val, end, &number))
1240 *(double *)place = number * mult;
1246 cmd_cert_type (const char *com, const char *val, void *place)
1248 static const struct decode_item choices[] = {
1249 { "pem", keyfile_pem },
1250 { "der", keyfile_asn1 },
1251 { "asn1", keyfile_asn1 },
1253 int ok = decode_string (val, choices, countof (choices), place);
1255 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1260 /* Specialized helper functions, used by `commands' to handle some
1261 options specially. */
1263 static bool check_user_specified_header (const char *);
1266 cmd_spec_dirstruct (const char *com, const char *val, void *place_ignored)
1268 if (!cmd_boolean (com, val, &opt.dirstruct))
1270 /* Since dirstruct behaviour is explicitly changed, no_dirstruct
1271 must be affected inversely. */
1273 opt.no_dirstruct = false;
1275 opt.no_dirstruct = true;
1280 cmd_spec_header (const char *com, const char *val, void *place_ignored)
1282 /* Empty value means reset the list of headers. */
1285 free_vec (opt.user_headers);
1286 opt.user_headers = NULL;
1290 if (!check_user_specified_header (val))
1292 fprintf (stderr, _("%s: %s: Invalid header %s.\n"),
1293 exec_name, com, quote (val));
1296 opt.user_headers = vec_append (opt.user_headers, val);
1301 cmd_spec_warc_header (const char *com, const char *val, void *place_ignored)
1303 /* Empty value means reset the list of headers. */
1306 free_vec (opt.warc_user_headers);
1307 opt.warc_user_headers = NULL;
1311 if (!check_user_specified_header (val))
1313 fprintf (stderr, _("%s: %s: Invalid WARC header %s.\n"),
1314 exec_name, com, quote (val));
1317 opt.warc_user_headers = vec_append (opt.warc_user_headers, val);
1322 cmd_spec_htmlify (const char *com, const char *val, void *place_ignored)
1324 int flag = cmd_boolean (com, val, &opt.htmlify);
1325 if (flag && !opt.htmlify)
1326 opt.remove_listing = false;
1330 /* Set the "mirror" mode. It means: recursive download, timestamping,
1331 no limit on max. recursion depth, and don't remove listings. */
1334 cmd_spec_mirror (const char *com, const char *val, void *place_ignored)
1338 if (!cmd_boolean (com, val, &mirror))
1342 opt.recursive = true;
1343 if (!opt.no_dirstruct)
1344 opt.dirstruct = true;
1345 opt.timestamping = true;
1346 opt.reclevel = INFINITE_RECURSION;
1347 opt.remove_listing = false;
1352 /* Validate --prefer-family and set the choice. Allowed values are
1353 "IPv4", "IPv6", and "none". */
1356 cmd_spec_prefer_family (const char *com, const char *val, void *place_ignored)
1358 static const struct decode_item choices[] = {
1359 { "IPv4", prefer_ipv4 },
1360 { "IPv6", prefer_ipv6 },
1361 { "none", prefer_none },
1363 int prefer_family = prefer_none;
1364 int ok = decode_string (val, choices, countof (choices), &prefer_family);
1366 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1367 opt.prefer_family = prefer_family;
1371 /* Set progress.type to VAL, but verify that it's a valid progress
1372 implementation before that. */
1375 cmd_spec_progress (const char *com, const char *val, void *place_ignored)
1377 if (!valid_progress_implementation_p (val))
1379 fprintf (stderr, _("%s: %s: Invalid progress type %s.\n"),
1380 exec_name, com, quote (val));
1383 xfree_null (opt.progress_type);
1385 /* Don't call set_progress_implementation here. It will be called
1386 in main() when it becomes clear what the log output is. */
1387 opt.progress_type = xstrdup (val);
1391 /* Set opt.recursive to VAL as with cmd_boolean. If opt.recursive is
1392 set to true, also set opt.dirstruct to true, unless opt.no_dirstruct
1396 cmd_spec_recursive (const char *com, const char *val, void *place_ignored)
1398 if (!cmd_boolean (com, val, &opt.recursive))
1402 if (opt.recursive && !opt.no_dirstruct)
1403 opt.dirstruct = true;
1408 /* Validate --regex-type and set the choice. */
1411 cmd_spec_regex_type (const char *com, const char *val, void *place_ignored)
1413 static const struct decode_item choices[] = {
1414 { "posix", regex_type_posix },
1416 { "pcre", regex_type_pcre },
1419 int regex_type = regex_type_posix;
1420 int ok = decode_string (val, choices, countof (choices), ®ex_type);
1422 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1423 opt.regex_type = regex_type;
1428 cmd_spec_restrict_file_names (const char *com, const char *val, void *place_ignored)
1430 int restrict_os = opt.restrict_files_os;
1431 int restrict_ctrl = opt.restrict_files_ctrl;
1432 int restrict_case = opt.restrict_files_case;
1433 int restrict_nonascii = opt.restrict_files_nonascii;
1437 #define VAL_IS(string_literal) BOUNDED_EQUAL (val, end, string_literal)
1441 end = strchr (val, ',');
1443 end = val + strlen (val);
1445 if (VAL_IS ("unix"))
1446 restrict_os = restrict_unix;
1447 else if (VAL_IS ("windows"))
1448 restrict_os = restrict_windows;
1449 else if (VAL_IS ("lowercase"))
1450 restrict_case = restrict_lowercase;
1451 else if (VAL_IS ("uppercase"))
1452 restrict_case = restrict_uppercase;
1453 else if (VAL_IS ("nocontrol"))
1454 restrict_ctrl = false;
1455 else if (VAL_IS ("ascii"))
1456 restrict_nonascii = true;
1459 fprintf (stderr, _("\
1460 %s: %s: Invalid restriction %s,\n\
1461 use [unix|windows],[lowercase|uppercase],[nocontrol],[ascii].\n"),
1462 exec_name, com, quote (val));
1469 while (*val && *end);
1473 opt.restrict_files_os = restrict_os;
1474 opt.restrict_files_ctrl = restrict_ctrl;
1475 opt.restrict_files_case = restrict_case;
1476 opt.restrict_files_nonascii = restrict_nonascii;
1482 cmd_spec_report_speed (const char *com, const char *val, void *place_ignored)
1484 opt.report_bps = strcasecmp (val, "bits") == 0;
1485 if (!opt.report_bps)
1486 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1487 return opt.report_bps;
1492 cmd_spec_secure_protocol (const char *com, const char *val, void *place)
1494 static const struct decode_item choices[] = {
1495 { "auto", secure_protocol_auto },
1496 { "sslv2", secure_protocol_sslv2 },
1497 { "sslv3", secure_protocol_sslv3 },
1498 { "tlsv1", secure_protocol_tlsv1 },
1499 { "pfs", secure_protocol_pfs },
1501 int ok = decode_string (val, choices, countof (choices), place);
1503 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1508 /* Set all three timeout values. */
1511 cmd_spec_timeout (const char *com, const char *val, void *place_ignored)
1514 if (!cmd_time (com, val, &value))
1516 opt.read_timeout = value;
1517 opt.connect_timeout = value;
1518 opt.dns_timeout = value;
1523 cmd_spec_useragent (const char *com, const char *val, void *place_ignored)
1525 /* Disallow embedded newlines. */
1526 if (strchr (val, '\n'))
1528 fprintf (stderr, _("%s: %s: Invalid value %s.\n"),
1529 exec_name, com, quote (val));
1532 xfree_null (opt.useragent);
1533 opt.useragent = xstrdup (val);
1537 /* The "verbose" option cannot be cmd_boolean because the variable is
1538 not bool -- it's of type int (-1 means uninitialized because of
1539 some random hackery for disallowing -q -v). */
1542 cmd_spec_verbose (const char *com, const char *val, void *place_ignored)
1545 if (cmd_boolean (com, val, &flag))
1553 /* Miscellaneous useful routines. */
1555 /* A very simple atoi clone, more useful than atoi because it works on
1556 delimited strings, and has error reportage. Returns true on success,
1557 false on failure. If successful, stores result to *DEST. */
1560 simple_atoi (const char *beg, const char *end, int *dest)
1563 bool negative = false;
1564 const char *p = beg;
1566 while (p < end && c_isspace (*p))
1568 if (p < end && (*p == '-' || *p == '+'))
1570 negative = (*p == '-');
1576 /* Read negative numbers in a separate loop because the most
1577 negative integer cannot be represented as a positive number. */
1580 for (; p < end && c_isdigit (*p); p++)
1582 int next = (10 * result) + (*p - '0');
1584 return false; /* overflow */
1588 for (; p < end && c_isdigit (*p); p++)
1590 int next = (10 * result) - (*p - '0');
1592 return false; /* underflow */
1603 /* Trivial atof, with error reporting. Handles "<digits>[.<digits>]",
1604 doesn't handle exponential notation. Returns true on success,
1605 false on failure. In case of success, stores its result to
1609 simple_atof (const char *beg, const char *end, double *dest)
1613 bool negative = false;
1614 bool seen_dot = false;
1615 bool seen_digit = false;
1618 const char *p = beg;
1620 while (p < end && c_isspace (*p))
1622 if (p < end && (*p == '-' || *p == '+'))
1624 negative = (*p == '-');
1628 for (; p < end; p++)
1634 result = (10 * result) + (ch - '0');
1636 result += (ch - '0') / (divider *= 10);
1658 /* Verify that the user-specified header in S is valid. It must
1659 contain a colon preceded by non-white-space characters and must not
1660 contain newlines. */
1663 check_user_specified_header (const char *s)
1667 for (p = s; *p && *p != ':' && !c_isspace (*p); p++)
1669 /* The header MUST contain `:' preceded by at least one
1670 non-whitespace character. */
1671 if (*p != ':' || p == s)
1673 /* The header MUST NOT contain newlines. */
1674 if (strchr (s, '\n'))
1679 /* Decode VAL into a number, according to ITEMS. */
1682 decode_string (const char *val, const struct decode_item *items, int itemcount,
1686 for (i = 0; i < itemcount; i++)
1687 if (0 == strcasecmp (val, items[i].name))
1689 *place = items[i].code;
1696 void cleanup_html_url (void);
1697 void spider_cleanup (void);
1700 /* Free the memory allocated by global variables. */
1704 /* Free external resources, close files, etc. */
1706 /* Close WARC file. */
1707 if (opt.warc_filename != 0)
1713 if (fclose (output_stream) == EOF)
1714 inform_exit_status (CLOSEFAILED);
1716 /* No need to check for error because Wget flushes its output (and
1717 checks for errors) after any data arrives. */
1719 /* We're exiting anyway so there's no real need to call free()
1720 hundreds of times. Skipping the frees will make Wget exit
1723 However, when detecting leaks, it's crucial to free() everything
1724 because then you can find the real leaks, i.e. the allocated
1725 memory which grows with the size of the program. */
1731 cleanup_html_url ();
1736 for (i = 0; i < nurl; i++)
1740 extern acc_t *netrc_list;
1741 free_netrc (netrc_list);
1743 xfree_null (opt.choose_config);
1744 xfree_null (opt.lfilename);
1745 xfree_null (opt.dir_prefix);
1746 xfree_null (opt.input_filename);
1747 xfree_null (opt.output_document);
1748 free_vec (opt.accepts);
1749 free_vec (opt.rejects);
1750 free_vec (opt.excludes);
1751 free_vec (opt.includes);
1752 free_vec (opt.domains);
1753 free_vec (opt.follow_tags);
1754 free_vec (opt.ignore_tags);
1755 xfree_null (opt.progress_type);
1756 xfree_null (opt.ftp_user);
1757 xfree_null (opt.ftp_passwd);
1758 xfree_null (opt.ftp_proxy);
1759 xfree_null (opt.https_proxy);
1760 xfree_null (opt.http_proxy);
1761 free_vec (opt.no_proxy);
1762 xfree_null (opt.useragent);
1763 xfree_null (opt.referer);
1764 xfree_null (opt.http_user);
1765 xfree_null (opt.http_passwd);
1766 free_vec (opt.user_headers);
1767 free_vec (opt.warc_user_headers);
1769 xfree_null (opt.cert_file);
1770 xfree_null (opt.private_key);
1771 xfree_null (opt.ca_directory);
1772 xfree_null (opt.ca_cert);
1773 xfree_null (opt.random_file);
1774 xfree_null (opt.egd_file);
1776 xfree_null (opt.bind_address);
1777 xfree_null (opt.cookies_input);
1778 xfree_null (opt.cookies_output);
1779 xfree_null (opt.user);
1780 xfree_null (opt.passwd);
1781 xfree_null (opt.base_href);
1782 xfree_null (opt.method);
1784 #endif /* DEBUG_MALLOC */
1787 /* Unit testing routines. */
1792 test_commands_sorted()
1794 int prev_idx = 0, next_idx = 1;
1795 int command_count = countof (commands) - 1;
1797 while (next_idx <= command_count)
1799 cmp = strcasecmp (commands[prev_idx].name, commands[next_idx].name);
1802 mu_assert ("FAILED", false);
1815 test_cmd_spec_restrict_file_names()
1820 int expected_restrict_files_os;
1821 int expected_restrict_files_ctrl;
1822 int expected_restrict_files_case;
1825 { "windows", restrict_windows, true, restrict_no_case_restriction, true },
1826 { "windows,", restrict_windows, true, restrict_no_case_restriction, true },
1827 { "windows,lowercase", restrict_windows, true, restrict_lowercase, true },
1828 { "unix,nocontrol,lowercase,", restrict_unix, false, restrict_lowercase, true },
1831 for (i = 0; i < sizeof(test_array)/sizeof(test_array[0]); ++i)
1836 res = cmd_spec_restrict_file_names ("dummy", test_array[i].val, NULL);
1839 fprintf (stderr, "test_cmd_spec_restrict_file_names: TEST %d\n", i); fflush (stderr);
1840 fprintf (stderr, "opt.restrict_files_os: %d\n", opt.restrict_files_os); fflush (stderr);
1841 fprintf (stderr, "opt.restrict_files_ctrl: %d\n", opt.restrict_files_ctrl); fflush (stderr);
1842 fprintf (stderr, "opt.restrict_files_case: %d\n", opt.restrict_files_case); fflush (stderr);
1844 mu_assert ("test_cmd_spec_restrict_file_names: wrong result",
1845 res == test_array[i].result
1846 && opt.restrict_files_os == test_array[i].expected_restrict_files_os
1847 && opt.restrict_files_ctrl == test_array[i].expected_restrict_files_ctrl
1848 && opt.restrict_files_case == test_array[i].expected_restrict_files_case);
1854 #endif /* TESTING */