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 },
160 { "debug", &opt.debug, cmd_boolean },
161 { "defaultpage", &opt.default_page, cmd_string },
162 { "deleteafter", &opt.delete_after, cmd_boolean },
163 { "dirprefix", &opt.dir_prefix, cmd_directory },
164 { "dirstruct", NULL, cmd_spec_dirstruct },
165 { "dnscache", &opt.dns_cache, cmd_boolean },
166 { "dnstimeout", &opt.dns_timeout, cmd_time },
167 { "domains", &opt.domains, cmd_vector },
168 { "dotbytes", &opt.dot_bytes, cmd_bytes },
169 { "dotsinline", &opt.dots_in_line, cmd_number },
170 { "dotspacing", &opt.dot_spacing, cmd_number },
171 { "dotstyle", &opt.dot_style, cmd_string }, /* deprecated */
173 { "egdfile", &opt.egd_file, cmd_file },
175 { "excludedirectories", &opt.excludes, cmd_directory_vector },
176 { "excludedomains", &opt.exclude_domains, cmd_vector },
177 { "followftp", &opt.follow_ftp, cmd_boolean },
178 { "followtags", &opt.follow_tags, cmd_vector },
179 { "forcehtml", &opt.force_html, cmd_boolean },
180 { "ftppasswd", &opt.ftp_passwd, cmd_string }, /* deprecated */
181 { "ftppassword", &opt.ftp_passwd, cmd_string },
182 { "ftpproxy", &opt.ftp_proxy, cmd_string },
184 { "ftpstmlf", &opt.ftp_stmlf, cmd_boolean },
185 #endif /* def __VMS */
186 { "ftpuser", &opt.ftp_user, cmd_string },
187 { "glob", &opt.ftp_glob, cmd_boolean },
188 { "header", NULL, cmd_spec_header },
189 { "htmlextension", &opt.adjust_extension, cmd_boolean }, /* deprecated */
190 { "htmlify", NULL, cmd_spec_htmlify },
191 { "httpkeepalive", &opt.http_keep_alive, cmd_boolean },
192 { "httppasswd", &opt.http_passwd, cmd_string }, /* deprecated */
193 { "httppassword", &opt.http_passwd, cmd_string },
194 { "httpproxy", &opt.http_proxy, cmd_string },
196 { "httpsonly", &opt.https_only, cmd_boolean },
198 { "httpsproxy", &opt.https_proxy, cmd_string },
199 { "httpuser", &opt.http_user, cmd_string },
200 { "ignorecase", &opt.ignore_case, cmd_boolean },
201 { "ignorelength", &opt.ignore_length, cmd_boolean },
202 { "ignoretags", &opt.ignore_tags, cmd_vector },
203 { "includedirectories", &opt.includes, cmd_directory_vector },
205 { "inet4only", &opt.ipv4_only, cmd_boolean },
206 { "inet6only", &opt.ipv6_only, cmd_boolean },
208 { "input", &opt.input_filename, cmd_file },
209 { "iri", &opt.enable_iri, cmd_boolean },
210 { "keepsessioncookies", &opt.keep_session_cookies, cmd_boolean },
211 { "limitrate", &opt.limit_rate, cmd_bytes },
212 { "loadcookies", &opt.cookies_input, cmd_file },
213 { "localencoding", &opt.locale, cmd_string },
214 { "logfile", &opt.lfilename, cmd_file },
215 { "login", &opt.ftp_user, cmd_string },/* deprecated*/
216 { "maxredirect", &opt.max_redirect, cmd_number },
217 { "method", &opt.method, cmd_string_uppercase },
218 { "mirror", NULL, cmd_spec_mirror },
219 { "netrc", &opt.netrc, cmd_boolean },
220 { "noclobber", &opt.noclobber, cmd_boolean },
221 { "noconfig", &opt.noconfig, cmd_boolean },
222 { "noparent", &opt.no_parent, cmd_boolean },
223 { "noproxy", &opt.no_proxy, cmd_vector },
224 { "numtries", &opt.ntry, cmd_number_inf },/* deprecated*/
225 { "outputdocument", &opt.output_document, cmd_file },
226 { "pagerequisites", &opt.page_requisites, cmd_boolean },
227 { "passiveftp", &opt.ftp_pasv, cmd_boolean },
228 { "passwd", &opt.ftp_passwd, cmd_string },/* deprecated*/
229 { "password", &opt.passwd, cmd_string },
230 { "postdata", &opt.post_data, cmd_string },
231 { "postfile", &opt.post_file_name, cmd_file },
232 { "preferfamily", NULL, cmd_spec_prefer_family },
233 { "preservepermissions", &opt.preserve_perm, cmd_boolean },
235 { "privatekey", &opt.private_key, cmd_file },
236 { "privatekeytype", &opt.private_key_type, cmd_cert_type },
238 { "progress", &opt.progress_type, cmd_spec_progress },
239 { "protocoldirectories", &opt.protocol_directories, cmd_boolean },
240 { "proxypasswd", &opt.proxy_passwd, cmd_string }, /* deprecated */
241 { "proxypassword", &opt.proxy_passwd, cmd_string },
242 { "proxyuser", &opt.proxy_user, cmd_string },
243 { "quiet", &opt.quiet, cmd_boolean },
244 { "quota", &opt.quota, cmd_bytes_sum },
246 { "randomfile", &opt.random_file, cmd_file },
248 { "randomwait", &opt.random_wait, cmd_boolean },
249 { "readtimeout", &opt.read_timeout, cmd_time },
250 { "reclevel", &opt.reclevel, cmd_number_inf },
251 { "recursive", NULL, cmd_spec_recursive },
252 { "referer", &opt.referer, cmd_string },
253 { "regextype", &opt.regex_type, cmd_spec_regex_type },
254 { "reject", &opt.rejects, cmd_vector },
255 { "rejectregex", &opt.rejectregex_s, cmd_string },
256 { "relativeonly", &opt.relative_only, cmd_boolean },
257 { "remoteencoding", &opt.encoding_remote, cmd_string },
258 { "removelisting", &opt.remove_listing, cmd_boolean },
259 { "reportspeed", &opt.report_bps, cmd_spec_report_speed},
260 { "restrictfilenames", NULL, cmd_spec_restrict_file_names },
261 { "retrsymlinks", &opt.retr_symlinks, cmd_boolean },
262 { "retryconnrefused", &opt.retry_connrefused, cmd_boolean },
263 { "robots", &opt.use_robots, cmd_boolean },
264 { "savecookies", &opt.cookies_output, cmd_file },
265 { "saveheaders", &opt.save_headers, cmd_boolean },
267 { "secureprotocol", &opt.secure_protocol, cmd_spec_secure_protocol },
269 { "serverresponse", &opt.server_response, cmd_boolean },
270 { "showalldnsentries", &opt.show_all_dns_entries, cmd_boolean },
271 { "spanhosts", &opt.spanhost, cmd_boolean },
272 { "spider", &opt.spider, cmd_boolean },
273 { "strictcomments", &opt.strict_comments, cmd_boolean },
274 { "timeout", NULL, cmd_spec_timeout },
275 { "timestamping", &opt.timestamping, cmd_boolean },
276 { "tries", &opt.ntry, cmd_number_inf },
277 { "trustservernames", &opt.trustservernames, cmd_boolean },
278 { "unlink", &opt.unlink, cmd_boolean },
279 { "useproxy", &opt.use_proxy, cmd_boolean },
280 { "user", &opt.user, cmd_string },
281 { "useragent", NULL, cmd_spec_useragent },
282 { "useservertimestamps", &opt.useservertimestamps, cmd_boolean },
283 { "verbose", NULL, cmd_spec_verbose },
284 { "wait", &opt.wait, cmd_time },
285 { "waitretry", &opt.waitretry, cmd_time },
286 { "warccdx", &opt.warc_cdx_enabled, cmd_boolean },
287 { "warccdxdedup", &opt.warc_cdx_dedup_filename, cmd_file },
289 { "warccompression", &opt.warc_compression_enabled, cmd_boolean },
291 { "warcdigests", &opt.warc_digests_enabled, cmd_boolean },
292 { "warcfile", &opt.warc_filename, cmd_file },
293 { "warcheader", NULL, cmd_spec_warc_header },
294 { "warckeeplog", &opt.warc_keep_log, cmd_boolean },
295 { "warcmaxsize", &opt.warc_maxsize, cmd_bytes },
296 { "warctempdir", &opt.warc_tempdir, cmd_directory },
298 { "wdebug", &opt.wdebug, cmd_boolean },
302 /* Look up CMDNAME in the commands[] and return its position in the
303 array. If CMDNAME is not found, return -1. */
306 command_by_name (const char *cmdname)
308 /* Use binary search for speed. Wget has ~100 commands, which
309 guarantees a worst case performance of 7 string comparisons. */
310 int lo = 0, hi = countof (commands) - 1;
314 int mid = (lo + hi) >> 1;
315 int cmp = strcasecmp (cmdname, commands[mid].name);
326 /* Reset the variables to default values. */
332 /* Most of the default values are 0 (and 0.0, NULL, and false).
333 Just reset everything, and fill in the non-zero values. Note
334 that initializing pointers to NULL this way is technically
335 illegal, but porting Wget to a machine where NULL is not all-zero
336 bit pattern will be the least of the implementors' worries. */
343 opt.add_hostdir = true;
347 opt.http_keep_alive = true;
348 opt.use_proxy = true;
349 tmp = getenv ("no_proxy");
351 opt.no_proxy = sepstring (tmp);
352 opt.prefer_family = prefer_none;
353 opt.allow_cache = true;
355 opt.read_timeout = 900;
356 opt.use_robots = true;
358 opt.remove_listing = true;
360 opt.dot_bytes = 1024;
361 opt.dot_spacing = 10;
362 opt.dots_in_line = 50;
364 opt.dns_cache = true;
368 opt.check_cert = true;
371 /* The default for file name restriction defaults to the OS type. */
372 #if defined(WINDOWS) || defined(MSDOS) || defined(__CYGWIN__)
373 opt.restrict_files_os = restrict_windows;
375 opt.restrict_files_os = restrict_unix;
377 opt.restrict_files_ctrl = true;
378 opt.restrict_files_nonascii = false;
379 opt.restrict_files_case = restrict_no_case_restriction;
381 opt.regex_type = regex_type_posix;
383 opt.max_redirect = 20;
388 opt.enable_iri = true;
390 opt.enable_iri = false;
393 opt.encoding_remote = NULL;
395 opt.useservertimestamps = true;
396 opt.show_all_dns_entries = false;
398 opt.warc_maxsize = 0; /* 1024 * 1024 * 1024; */
400 opt.warc_compression_enabled = true;
402 opt.warc_compression_enabled = false;
404 opt.warc_digests_enabled = true;
405 opt.warc_cdx_enabled = false;
406 opt.warc_cdx_dedup_filename = NULL;
407 opt.warc_tempdir = NULL;
408 opt.warc_keep_log = true;
411 /* Return the user's home directory (strdup-ed), or NULL if none is
416 static char *buf = NULL;
417 static char *home, *ret;
421 home = getenv ("HOME");
427 /* Under MSDOS, if $HOME isn't defined, use the directory where
428 `wget.exe' resides. */
429 const char *_w32_get_argv0 (void); /* in libwatt.a/pcconfig.c */
432 buff = _w32_get_argv0 ();
434 p = strrchr (buf, '/'); /* djgpp */
436 p = strrchr (buf, '\\'); /* others */
440 buff = malloc (len + 1);
444 strncpy (buff, _w32_get_argv0 (), len);
448 #elif !defined(WINDOWS)
449 /* If HOME is not defined, try getting it from the password
451 struct passwd *pwd = getpwuid (getuid ());
452 if (!pwd || !pwd->pw_dir)
456 /* Under Windows, if $HOME isn't defined, use the directory where
457 `wget.exe' resides. */
463 ret = home ? xstrdup (home) : NULL;
469 /* Check the 'WGETRC' environment variable and return the file name
470 if 'WGETRC' is set and is a valid file.
471 If the `WGETRC' variable exists but the file does not exist, the
472 function will exit(). */
474 wgetrc_env_file_name (void)
476 char *env = getenv ("WGETRC");
479 if (!file_exists_p (env))
481 fprintf (stderr, _("%s: WGETRC points to %s, which doesn't exist.\n"),
485 return xstrdup (env);
490 /* Check for the existance of '$HOME/.wgetrc' and return its path
491 if it exists and is set. */
493 wgetrc_user_file_name (void)
497 /* If that failed, try $HOME/.wgetrc (or equivalent). */
500 file = "SYS$LOGIN:.wgetrc";
501 #else /* def __VMS */
504 file = aprintf ("%s/.wgetrc", home);
506 #endif /* def __VMS [else] */
510 if (!file_exists_p (file))
518 /* Return the path to the user's .wgetrc. This is either the value of
519 `WGETRC' environment variable, or `$HOME/.wgetrc'.
521 Additionally, for windows, look in the directory where wget.exe
524 wgetrc_file_name (void)
526 char *file = wgetrc_env_file_name ();
530 file = wgetrc_user_file_name ();
533 /* Under Windows, if we still haven't found .wgetrc, look for the file
534 `wget.ini' in the directory where `wget.exe' resides; we do this for
535 backward compatibility with previous versions of Wget.
536 SYSTEM_WGETRC should not be defined under WINDOWS. */
539 char *home = home_dir ();
545 file = aprintf ("%s/wget.ini", home);
546 if (!file_exists_p (file))
559 /* Return values of parse_line. */
567 static enum parse_line parse_line (const char *, char **, char **, int *);
568 static bool setval_internal (int, const char *, const char *);
569 static bool setval_internal_tilde (int, const char *, const char *);
571 /* Initialize variables from a wgetrc file. Returns zero (failure) if
572 there were errors in the file. */
575 run_wgetrc (const char *file)
583 fp = fopen (file, "r");
586 fprintf (stderr, _("%s: Cannot read %s (%s).\n"), exec_name,
587 file, strerror (errno));
588 return true; /* not a fatal error */
591 while (getline (&line, &bufsize, fp) > 0)
593 char *com = NULL, *val = NULL;
596 /* Parse the line. */
597 switch (parse_line (line, &com, &val, &comind))
600 /* If everything is OK, set the value. */
601 if (!setval_internal_tilde (comind, com, val))
603 fprintf (stderr, _("%s: Error in %s at line %d.\n"),
604 exec_name, file, ln);
608 case line_syntax_error:
609 fprintf (stderr, _("%s: Syntax error in %s at line %d.\n"),
610 exec_name, file, ln);
613 case line_unknown_command:
614 fprintf (stderr, _("%s: Unknown command %s in %s at line %d.\n"),
615 exec_name, quote (com), file, ln);
633 /* Initialize the defaults and run the system wgetrc and user's own
638 char *file, *env_sysrc;
641 /* Run a non-standard system rc file when the according environment
642 variable has been set. For internal testing purposes only! */
643 env_sysrc = getenv ("SYSTEM_WGETRC");
644 if (env_sysrc && file_exists_p (env_sysrc))
646 ok &= run_wgetrc (env_sysrc);
647 /* If there are any problems parsing the system wgetrc file, tell
651 fprintf (stderr, _("\
652 Parsing system wgetrc file (env SYSTEM_WGETRC) failed. Please check\n\
654 or specify a different file using --config.\n"), env_sysrc);
658 /* Otherwise, if SYSTEM_WGETRC is defined, use it. */
660 else if (file_exists_p (SYSTEM_WGETRC))
661 ok &= run_wgetrc (SYSTEM_WGETRC);
662 /* If there are any problems parsing the system wgetrc file, tell
666 fprintf (stderr, _("\
667 Parsing system wgetrc file failed. Please check\n\
669 or specify a different file using --config.\n"), SYSTEM_WGETRC);
673 /* Override it with your own, if one exists. */
674 file = wgetrc_file_name ();
677 /* #### We should canonicalize `file' and SYSTEM_WGETRC with
678 something like realpath() before comparing them with `strcmp' */
680 if (!strcmp (file, SYSTEM_WGETRC))
682 fprintf (stderr, _("\
683 %s: Warning: Both system and user wgetrc point to %s.\n"),
684 exec_name, quote (file));
688 ok &= run_wgetrc (file);
690 /* If there were errors processing either `.wgetrc', abort. */
698 /* Remove dashes and underscores from S, modifying S in the
704 char *t = s; /* t - tortoise */
705 char *h = s; /* h - hare */
707 if (*h == '_' || *h == '-')
714 /* Parse the line pointed by line, with the syntax:
715 <sp>* command <sp>* = <sp>* value <sp>*
716 Uses malloc to allocate space for command and value.
718 Returns one of line_ok, line_empty, line_syntax_error, or
719 line_unknown_command.
721 In case of line_ok, *COM and *VAL point to freshly allocated
722 strings, and *COMIND points to com's index. In case of error or
723 empty line, their values are unmodified. */
725 static enum parse_line
726 parse_line (const char *line, char **com, char **val, int *comind)
729 const char *end = line + strlen (line);
730 const char *cmdstart, *cmdend;
731 const char *valstart, *valend;
736 /* Skip leading and trailing whitespace. */
737 while (*line && c_isspace (*line))
739 while (end > line && c_isspace (end[-1]))
742 /* Skip empty lines and comments. */
743 if (!*line || *line == '#')
749 while (p < end && (c_isalnum (*p) || *p == '_' || *p == '-'))
753 /* Skip '=', as well as any space before or after it. */
754 while (p < end && c_isspace (*p))
756 if (p == end || *p != '=')
757 return line_syntax_error;
759 while (p < end && c_isspace (*p))
765 /* The syntax is valid (even though the command might not be). Fill
766 in the command name and value. */
767 *com = strdupdelim (cmdstart, cmdend);
768 *val = strdupdelim (valstart, valend);
770 /* The line now known to be syntactically correct. Check whether
771 the command is valid. */
772 BOUNDED_TO_ALLOCA (cmdstart, cmdend, cmdcopy);
774 ind = command_by_name (cmdcopy);
776 return line_unknown_command;
778 /* Report success to the caller. */
783 #if defined(WINDOWS) || defined(MSDOS)
784 # define ISSEP(c) ((c) == '/' || (c) == '\\')
786 # define ISSEP(c) ((c) == '/')
789 /* Run commands[comind].action. */
792 setval_internal (int comind, const char *com, const char *val)
794 assert (0 <= comind && ((size_t) comind) < countof (commands));
795 DEBUGP (("Setting %s (%s) to %s\n", com, commands[comind].name, val));
796 return commands[comind].action (com, val, commands[comind].place);
800 setval_internal_tilde (int comind, const char *com, const char *val)
806 ret = setval_internal (comind, com, val);
808 /* We make tilde expansion for cmd_file and cmd_directory */
809 if (((commands[comind].action == cmd_file) ||
810 (commands[comind].action == cmd_directory))
811 && ret && (*val == '~' && ISSEP (val[1])))
813 pstring = commands[comind].place;
817 homelen = strlen (home);
818 while (homelen && ISSEP (home[homelen - 1]))
819 home[--homelen] = '\0';
821 /* Skip the leading "~/". */
822 for (++val; ISSEP (*val); val++)
824 *pstring = concat_strings (home, "/", val, (char *)0);
830 /* Run command COM with value VAL. If running the command produces an
831 error, report the error and exit.
833 This is intended to be called from main() to modify Wget's behavior
834 through command-line switches. Since COM is hard-coded in main(),
835 it is not canonicalized, and this aborts when COM is not found.
837 If COMIND's are exported to init.h, this function will be changed
838 to accept COMIND directly. */
841 setoptval (const char *com, const char *val, const char *optname)
843 /* Prepend "--" to OPTNAME. */
844 char *dd_optname = (char *) alloca (2 + strlen (optname) + 1);
847 strcpy (dd_optname + 2, optname);
849 assert (val != NULL);
850 if (!setval_internal (command_by_name (com), dd_optname, val))
854 /* Parse OPT into command and value and run it. For example,
855 run_command("foo=bar") is equivalent to setoptval("foo", "bar").
856 This is used by the `--execute' flag in main.c. */
859 run_command (const char *opt)
863 switch (parse_line (opt, &com, &val, &comind))
866 if (!setval_internal (comind, com, val))
872 fprintf (stderr, _("%s: Invalid --execute command %s\n"),
873 exec_name, quote (opt));
878 /* Generic helper functions, for use with `commands'. */
880 /* Forward declarations: */
885 static bool decode_string (const char *, const struct decode_item *, int, int *);
886 static bool simple_atoi (const char *, const char *, int *);
887 static bool simple_atof (const char *, const char *, double *);
889 #define CMP1(p, c0) (c_tolower((p)[0]) == (c0) && (p)[1] == '\0')
891 #define CMP2(p, c0, c1) (c_tolower((p)[0]) == (c0) \
892 && c_tolower((p)[1]) == (c1) \
895 #define CMP3(p, c0, c1, c2) (c_tolower((p)[0]) == (c0) \
896 && c_tolower((p)[1]) == (c1) \
897 && c_tolower((p)[2]) == (c2) \
901 /* Store the boolean value from VAL to PLACE. COM is ignored,
902 except for error messages. */
904 cmd_boolean (const char *com, const char *val, void *place)
908 if (CMP2 (val, 'o', 'n') || CMP3 (val, 'y', 'e', 's') || CMP1 (val, '1'))
909 /* "on", "yes" and "1" mean true. */
911 else if (CMP3 (val, 'o', 'f', 'f') || CMP2 (val, 'n', 'o') || CMP1 (val, '0'))
912 /* "off", "no" and "0" mean false. */
917 _("%s: %s: Invalid boolean %s; use `on' or `off'.\n"),
918 exec_name, com, quote (val));
922 *(bool *) place = value;
926 /* Set the non-negative integer value from VAL to PLACE. With
927 incorrect specification, the number remains unchanged. */
929 cmd_number (const char *com, const char *val, void *place)
931 if (!simple_atoi (val, val + strlen (val), place)
932 || *(int *) place < 0)
934 fprintf (stderr, _("%s: %s: Invalid number %s.\n"),
935 exec_name, com, quote (val));
941 /* Similar to cmd_number(), only accepts `inf' as a synonym for 0. */
943 cmd_number_inf (const char *com, const char *val, void *place)
945 if (!strcasecmp (val, "inf"))
950 return cmd_number (com, val, place);
953 /* Copy (strdup) the string at COM to a new location and place a
954 pointer to *PLACE. */
956 cmd_string (const char *com, const char *val, void *place)
958 char **pstring = (char **)place;
960 xfree_null (*pstring);
961 *pstring = xstrdup (val);
965 /* Like cmd_string but ensure the string is upper case. */
967 cmd_string_uppercase (const char *com, const char *val, void *place)
970 pstring = (char **)place;
971 xfree_null (*pstring);
973 *pstring = xmalloc (strlen (val) + 1);
975 for (q = *pstring; *val; val++, q++)
976 *q = c_toupper (*val);
983 /* Like cmd_string, but handles tilde-expansion when reading a user's
984 `.wgetrc'. In that case, and if VAL begins with `~', the tilde
985 gets expanded to the user's home directory. */
987 cmd_file (const char *com, const char *val, void *place)
989 char **pstring = (char **)place;
991 xfree_null (*pstring);
993 /* #### If VAL is empty, perhaps should set *PLACE to NULL. */
995 *pstring = xstrdup (val);
997 #if defined(WINDOWS) || defined(MSDOS)
998 /* Convert "\" to "/". */
1001 for (s = *pstring; *s; s++)
1009 /* Like cmd_file, but strips trailing '/' characters. */
1011 cmd_directory (const char *com, const char *val, void *place)
1015 /* Call cmd_file() for tilde expansion and separator
1016 canonicalization (backslash -> slash under Windows). These
1017 things should perhaps be in a separate function. */
1018 if (!cmd_file (com, val, place))
1021 s = *(char **)place;
1023 while (t > s && *--t == '/')
1029 /* Split VAL by space to a vector of values, and append those values
1030 to vector pointed to by the PLACE argument. If VAL is empty, the
1031 PLACE vector is cleared instead. */
1034 cmd_vector (const char *com, const char *val, void *place)
1036 char ***pvec = (char ***)place;
1039 *pvec = merge_vecs (*pvec, sepstring (val));
1049 cmd_directory_vector (const char *com, const char *val, void *place)
1051 char ***pvec = (char ***)place;
1055 /* Strip the trailing slashes from directories. */
1058 seps = sepstring (val);
1059 for (t = seps; t && *t; t++)
1061 int len = strlen (*t);
1062 /* Skip degenerate case of root directory. */
1065 if ((*t)[len - 1] == '/')
1066 (*t)[len - 1] = '\0';
1069 *pvec = merge_vecs (*pvec, seps);
1079 /* Engine for cmd_bytes and cmd_bytes_sum: converts a string such as
1080 "100k" or "2.5G" to a floating point number. */
1083 parse_bytes_helper (const char *val, double *result)
1085 double number, mult;
1086 const char *end = val + strlen (val);
1088 /* Check for "inf". */
1089 if (0 == strcmp (val, "inf"))
1095 /* Strip trailing whitespace. */
1096 while (val < end && c_isspace (end[-1]))
1101 switch (c_tolower (end[-1]))
1104 --end, mult = 1024.0;
1107 --end, mult = 1048576.0;
1110 --end, mult = 1073741824.0;
1113 --end, mult = 1099511627776.0;
1116 /* Not a recognized suffix: assume it's a digit. (If not,
1117 simple_atof will raise an error.) */
1121 /* Skip leading and trailing whitespace. */
1122 while (val < end && c_isspace (*val))
1124 while (val < end && c_isspace (end[-1]))
1129 if (!simple_atof (val, end, &number) || number < 0)
1132 *result = number * mult;
1136 /* Parse VAL as a number and set its value to PLACE (which should
1139 By default, the value is assumed to be in bytes. If "K", "M", or
1140 "G" are appended, the value is multiplied with 1<<10, 1<<20, or
1141 1<<30, respectively. Floating point values are allowed and are
1142 cast to integer before use. The idea is to be able to use things
1143 like 1.5k instead of "1536".
1145 The string "inf" is returned as 0.
1147 In case of error, false is returned and memory pointed to by PLACE
1148 remains unmodified. */
1151 cmd_bytes (const char *com, const char *val, void *place)
1154 if (!parse_bytes_helper (val, &byte_value))
1156 fprintf (stderr, _("%s: %s: Invalid byte value %s\n"),
1157 exec_name, com, quote (val));
1160 *(wgint *)place = (wgint)byte_value;
1164 /* Like cmd_bytes, but PLACE is interpreted as a pointer to
1165 SIZE_SUM. It works by converting the string to double, therefore
1166 working with values up to 2^53-1 without loss of precision. This
1167 value (8192 TB) is large enough to serve for a while. */
1170 cmd_bytes_sum (const char *com, const char *val, void *place)
1173 if (!parse_bytes_helper (val, &byte_value))
1175 fprintf (stderr, _("%s: %s: Invalid byte value %s\n"),
1176 exec_name, com, quote (val));
1179 *(SUM_SIZE_INT *) place = (SUM_SIZE_INT) byte_value;
1183 /* Store the value of VAL to *OUT. The value is a time period, by
1184 default expressed in seconds, but also accepting suffixes "m", "h",
1185 "d", and "w" for minutes, hours, days, and weeks respectively. */
1188 cmd_time (const char *com, const char *val, void *place)
1190 double number, mult;
1191 const char *end = val + strlen (val);
1193 /* Strip trailing whitespace. */
1194 while (val < end && c_isspace (end[-1]))
1200 fprintf (stderr, _("%s: %s: Invalid time period %s\n"),
1201 exec_name, com, quote (val));
1205 switch (c_tolower (end[-1]))
1208 --end, mult = 1; /* seconds */
1211 --end, mult = 60; /* minutes */
1214 --end, mult = 3600; /* hours */
1217 --end, mult = 86400.0; /* days */
1220 --end, mult = 604800.0; /* weeks */
1223 /* Not a recognized suffix: assume it belongs to the number.
1224 (If not, simple_atof will raise an error.) */
1228 /* Skip leading and trailing whitespace. */
1229 while (val < end && c_isspace (*val))
1231 while (val < end && c_isspace (end[-1]))
1236 if (!simple_atof (val, end, &number))
1239 *(double *)place = number * mult;
1245 cmd_cert_type (const char *com, const char *val, void *place)
1247 static const struct decode_item choices[] = {
1248 { "pem", keyfile_pem },
1249 { "der", keyfile_asn1 },
1250 { "asn1", keyfile_asn1 },
1252 int ok = decode_string (val, choices, countof (choices), place);
1254 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1259 /* Specialized helper functions, used by `commands' to handle some
1260 options specially. */
1262 static bool check_user_specified_header (const char *);
1265 cmd_spec_dirstruct (const char *com, const char *val, void *place_ignored)
1267 if (!cmd_boolean (com, val, &opt.dirstruct))
1269 /* Since dirstruct behaviour is explicitly changed, no_dirstruct
1270 must be affected inversely. */
1272 opt.no_dirstruct = false;
1274 opt.no_dirstruct = true;
1279 cmd_spec_header (const char *com, const char *val, void *place_ignored)
1281 /* Empty value means reset the list of headers. */
1284 free_vec (opt.user_headers);
1285 opt.user_headers = NULL;
1289 if (!check_user_specified_header (val))
1291 fprintf (stderr, _("%s: %s: Invalid header %s.\n"),
1292 exec_name, com, quote (val));
1295 opt.user_headers = vec_append (opt.user_headers, val);
1300 cmd_spec_warc_header (const char *com, const char *val, void *place_ignored)
1302 /* Empty value means reset the list of headers. */
1305 free_vec (opt.warc_user_headers);
1306 opt.warc_user_headers = NULL;
1310 if (!check_user_specified_header (val))
1312 fprintf (stderr, _("%s: %s: Invalid WARC header %s.\n"),
1313 exec_name, com, quote (val));
1316 opt.warc_user_headers = vec_append (opt.warc_user_headers, val);
1321 cmd_spec_htmlify (const char *com, const char *val, void *place_ignored)
1323 int flag = cmd_boolean (com, val, &opt.htmlify);
1324 if (flag && !opt.htmlify)
1325 opt.remove_listing = false;
1329 /* Set the "mirror" mode. It means: recursive download, timestamping,
1330 no limit on max. recursion depth, and don't remove listings. */
1333 cmd_spec_mirror (const char *com, const char *val, void *place_ignored)
1337 if (!cmd_boolean (com, val, &mirror))
1341 opt.recursive = true;
1342 if (!opt.no_dirstruct)
1343 opt.dirstruct = true;
1344 opt.timestamping = true;
1345 opt.reclevel = INFINITE_RECURSION;
1346 opt.remove_listing = false;
1351 /* Validate --prefer-family and set the choice. Allowed values are
1352 "IPv4", "IPv6", and "none". */
1355 cmd_spec_prefer_family (const char *com, const char *val, void *place_ignored)
1357 static const struct decode_item choices[] = {
1358 { "IPv4", prefer_ipv4 },
1359 { "IPv6", prefer_ipv6 },
1360 { "none", prefer_none },
1362 int prefer_family = prefer_none;
1363 int ok = decode_string (val, choices, countof (choices), &prefer_family);
1365 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1366 opt.prefer_family = prefer_family;
1370 /* Set progress.type to VAL, but verify that it's a valid progress
1371 implementation before that. */
1374 cmd_spec_progress (const char *com, const char *val, void *place_ignored)
1376 if (!valid_progress_implementation_p (val))
1378 fprintf (stderr, _("%s: %s: Invalid progress type %s.\n"),
1379 exec_name, com, quote (val));
1382 xfree_null (opt.progress_type);
1384 /* Don't call set_progress_implementation here. It will be called
1385 in main() when it becomes clear what the log output is. */
1386 opt.progress_type = xstrdup (val);
1390 /* Set opt.recursive to VAL as with cmd_boolean. If opt.recursive is
1391 set to true, also set opt.dirstruct to true, unless opt.no_dirstruct
1395 cmd_spec_recursive (const char *com, const char *val, void *place_ignored)
1397 if (!cmd_boolean (com, val, &opt.recursive))
1401 if (opt.recursive && !opt.no_dirstruct)
1402 opt.dirstruct = true;
1407 /* Validate --regex-type and set the choice. */
1410 cmd_spec_regex_type (const char *com, const char *val, void *place_ignored)
1412 static const struct decode_item choices[] = {
1413 { "posix", regex_type_posix },
1415 { "pcre", regex_type_pcre },
1418 int regex_type = regex_type_posix;
1419 int ok = decode_string (val, choices, countof (choices), ®ex_type);
1421 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1422 opt.regex_type = regex_type;
1427 cmd_spec_restrict_file_names (const char *com, const char *val, void *place_ignored)
1429 int restrict_os = opt.restrict_files_os;
1430 int restrict_ctrl = opt.restrict_files_ctrl;
1431 int restrict_case = opt.restrict_files_case;
1432 int restrict_nonascii = opt.restrict_files_nonascii;
1436 #define VAL_IS(string_literal) BOUNDED_EQUAL (val, end, string_literal)
1440 end = strchr (val, ',');
1442 end = val + strlen (val);
1444 if (VAL_IS ("unix"))
1445 restrict_os = restrict_unix;
1446 else if (VAL_IS ("windows"))
1447 restrict_os = restrict_windows;
1448 else if (VAL_IS ("lowercase"))
1449 restrict_case = restrict_lowercase;
1450 else if (VAL_IS ("uppercase"))
1451 restrict_case = restrict_uppercase;
1452 else if (VAL_IS ("nocontrol"))
1453 restrict_ctrl = false;
1454 else if (VAL_IS ("ascii"))
1455 restrict_nonascii = true;
1458 fprintf (stderr, _("\
1459 %s: %s: Invalid restriction %s,\n\
1460 use [unix|windows],[lowercase|uppercase],[nocontrol],[ascii].\n"),
1461 exec_name, com, quote (val));
1468 while (*val && *end);
1472 opt.restrict_files_os = restrict_os;
1473 opt.restrict_files_ctrl = restrict_ctrl;
1474 opt.restrict_files_case = restrict_case;
1475 opt.restrict_files_nonascii = restrict_nonascii;
1481 cmd_spec_report_speed (const char *com, const char *val, void *place_ignored)
1483 opt.report_bps = strcasecmp (val, "bits") == 0;
1484 if (!opt.report_bps)
1485 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1486 return opt.report_bps;
1491 cmd_spec_secure_protocol (const char *com, const char *val, void *place)
1493 static const struct decode_item choices[] = {
1494 { "auto", secure_protocol_auto },
1495 { "sslv2", secure_protocol_sslv2 },
1496 { "sslv3", secure_protocol_sslv3 },
1497 { "tlsv1", secure_protocol_tlsv1 },
1498 { "pfs", secure_protocol_pfs },
1500 int ok = decode_string (val, choices, countof (choices), place);
1502 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1507 /* Set all three timeout values. */
1510 cmd_spec_timeout (const char *com, const char *val, void *place_ignored)
1513 if (!cmd_time (com, val, &value))
1515 opt.read_timeout = value;
1516 opt.connect_timeout = value;
1517 opt.dns_timeout = value;
1522 cmd_spec_useragent (const char *com, const char *val, void *place_ignored)
1524 /* Disallow embedded newlines. */
1525 if (strchr (val, '\n'))
1527 fprintf (stderr, _("%s: %s: Invalid value %s.\n"),
1528 exec_name, com, quote (val));
1531 xfree_null (opt.useragent);
1532 opt.useragent = xstrdup (val);
1536 /* The "verbose" option cannot be cmd_boolean because the variable is
1537 not bool -- it's of type int (-1 means uninitialized because of
1538 some random hackery for disallowing -q -v). */
1541 cmd_spec_verbose (const char *com, const char *val, void *place_ignored)
1544 if (cmd_boolean (com, val, &flag))
1552 /* Miscellaneous useful routines. */
1554 /* A very simple atoi clone, more useful than atoi because it works on
1555 delimited strings, and has error reportage. Returns true on success,
1556 false on failure. If successful, stores result to *DEST. */
1559 simple_atoi (const char *beg, const char *end, int *dest)
1562 bool negative = false;
1563 const char *p = beg;
1565 while (p < end && c_isspace (*p))
1567 if (p < end && (*p == '-' || *p == '+'))
1569 negative = (*p == '-');
1575 /* Read negative numbers in a separate loop because the most
1576 negative integer cannot be represented as a positive number. */
1579 for (; p < end && c_isdigit (*p); p++)
1581 int next = (10 * result) + (*p - '0');
1583 return false; /* overflow */
1587 for (; p < end && c_isdigit (*p); p++)
1589 int next = (10 * result) - (*p - '0');
1591 return false; /* underflow */
1602 /* Trivial atof, with error reporting. Handles "<digits>[.<digits>]",
1603 doesn't handle exponential notation. Returns true on success,
1604 false on failure. In case of success, stores its result to
1608 simple_atof (const char *beg, const char *end, double *dest)
1612 bool negative = false;
1613 bool seen_dot = false;
1614 bool seen_digit = false;
1617 const char *p = beg;
1619 while (p < end && c_isspace (*p))
1621 if (p < end && (*p == '-' || *p == '+'))
1623 negative = (*p == '-');
1627 for (; p < end; p++)
1633 result = (10 * result) + (ch - '0');
1635 result += (ch - '0') / (divider *= 10);
1657 /* Verify that the user-specified header in S is valid. It must
1658 contain a colon preceded by non-white-space characters and must not
1659 contain newlines. */
1662 check_user_specified_header (const char *s)
1666 for (p = s; *p && *p != ':' && !c_isspace (*p); p++)
1668 /* The header MUST contain `:' preceded by at least one
1669 non-whitespace character. */
1670 if (*p != ':' || p == s)
1672 /* The header MUST NOT contain newlines. */
1673 if (strchr (s, '\n'))
1678 /* Decode VAL into a number, according to ITEMS. */
1681 decode_string (const char *val, const struct decode_item *items, int itemcount,
1685 for (i = 0; i < itemcount; i++)
1686 if (0 == strcasecmp (val, items[i].name))
1688 *place = items[i].code;
1695 void cleanup_html_url (void);
1696 void spider_cleanup (void);
1699 /* Free the memory allocated by global variables. */
1703 /* Free external resources, close files, etc. */
1705 /* Close WARC file. */
1706 if (opt.warc_filename != 0)
1712 if (fclose (output_stream) == EOF)
1713 inform_exit_status (CLOSEFAILED);
1715 /* No need to check for error because Wget flushes its output (and
1716 checks for errors) after any data arrives. */
1718 /* We're exiting anyway so there's no real need to call free()
1719 hundreds of times. Skipping the frees will make Wget exit
1722 However, when detecting leaks, it's crucial to free() everything
1723 because then you can find the real leaks, i.e. the allocated
1724 memory which grows with the size of the program. */
1730 cleanup_html_url ();
1735 for (i = 0; i < nurl; i++)
1739 extern acc_t *netrc_list;
1740 free_netrc (netrc_list);
1742 xfree_null (opt.choose_config);
1743 xfree_null (opt.lfilename);
1744 xfree_null (opt.dir_prefix);
1745 xfree_null (opt.input_filename);
1746 xfree_null (opt.output_document);
1747 free_vec (opt.accepts);
1748 free_vec (opt.rejects);
1749 free_vec (opt.excludes);
1750 free_vec (opt.includes);
1751 free_vec (opt.domains);
1752 free_vec (opt.follow_tags);
1753 free_vec (opt.ignore_tags);
1754 xfree_null (opt.progress_type);
1755 xfree_null (opt.ftp_user);
1756 xfree_null (opt.ftp_passwd);
1757 xfree_null (opt.ftp_proxy);
1758 xfree_null (opt.https_proxy);
1759 xfree_null (opt.http_proxy);
1760 free_vec (opt.no_proxy);
1761 xfree_null (opt.useragent);
1762 xfree_null (opt.referer);
1763 xfree_null (opt.http_user);
1764 xfree_null (opt.http_passwd);
1765 free_vec (opt.user_headers);
1766 free_vec (opt.warc_user_headers);
1768 xfree_null (opt.cert_file);
1769 xfree_null (opt.private_key);
1770 xfree_null (opt.ca_directory);
1771 xfree_null (opt.ca_cert);
1772 xfree_null (opt.random_file);
1773 xfree_null (opt.egd_file);
1775 xfree_null (opt.bind_address);
1776 xfree_null (opt.cookies_input);
1777 xfree_null (opt.cookies_output);
1778 xfree_null (opt.user);
1779 xfree_null (opt.passwd);
1780 xfree_null (opt.base_href);
1781 xfree_null (opt.method);
1783 #endif /* DEBUG_MALLOC */
1786 /* Unit testing routines. */
1791 test_commands_sorted()
1793 int prev_idx = 0, next_idx = 1;
1794 int command_count = countof (commands) - 1;
1796 while (next_idx <= command_count)
1798 cmp = strcasecmp (commands[prev_idx].name, commands[next_idx].name);
1801 mu_assert ("FAILED", false);
1814 test_cmd_spec_restrict_file_names()
1819 int expected_restrict_files_os;
1820 int expected_restrict_files_ctrl;
1821 int expected_restrict_files_case;
1824 { "windows", restrict_windows, true, restrict_no_case_restriction, true },
1825 { "windows,", restrict_windows, true, restrict_no_case_restriction, true },
1826 { "windows,lowercase", restrict_windows, true, restrict_lowercase, true },
1827 { "unix,nocontrol,lowercase,", restrict_unix, false, restrict_lowercase, true },
1830 for (i = 0; i < sizeof(test_array)/sizeof(test_array[0]); ++i)
1835 res = cmd_spec_restrict_file_names ("dummy", test_array[i].val, NULL);
1838 fprintf (stderr, "test_cmd_spec_restrict_file_names: TEST %d\n", i); fflush (stderr);
1839 fprintf (stderr, "opt.restrict_files_os: %d\n", opt.restrict_files_os); fflush (stderr);
1840 fprintf (stderr, "opt.restrict_files_ctrl: %d\n", opt.restrict_files_ctrl); fflush (stderr);
1841 fprintf (stderr, "opt.restrict_files_case: %d\n", opt.restrict_files_case); fflush (stderr);
1843 mu_assert ("test_cmd_spec_restrict_file_names: wrong result",
1844 res == test_array[i].result
1845 && opt.restrict_files_os == test_array[i].expected_restrict_files_os
1846 && opt.restrict_files_ctrl == test_array[i].expected_restrict_files_ctrl
1847 && opt.restrict_files_case == test_array[i].expected_restrict_files_case);
1853 #endif /* TESTING */