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 },
197 { "httpsproxy", &opt.https_proxy, cmd_string },
198 { "httpuser", &opt.http_user, cmd_string },
199 { "ignorecase", &opt.ignore_case, cmd_boolean },
200 { "ignorelength", &opt.ignore_length, cmd_boolean },
201 { "ignoretags", &opt.ignore_tags, cmd_vector },
202 { "includedirectories", &opt.includes, cmd_directory_vector },
204 { "inet4only", &opt.ipv4_only, cmd_boolean },
205 { "inet6only", &opt.ipv6_only, cmd_boolean },
207 { "input", &opt.input_filename, cmd_file },
208 { "iri", &opt.enable_iri, cmd_boolean },
209 { "keepsessioncookies", &opt.keep_session_cookies, cmd_boolean },
210 { "limitrate", &opt.limit_rate, cmd_bytes },
211 { "loadcookies", &opt.cookies_input, cmd_file },
212 { "localencoding", &opt.locale, cmd_string },
213 { "logfile", &opt.lfilename, cmd_file },
214 { "login", &opt.ftp_user, cmd_string },/* deprecated*/
215 { "maxredirect", &opt.max_redirect, cmd_number },
216 { "method", &opt.method, cmd_string_uppercase },
217 { "mirror", NULL, cmd_spec_mirror },
218 { "netrc", &opt.netrc, cmd_boolean },
219 { "noclobber", &opt.noclobber, cmd_boolean },
220 { "noparent", &opt.no_parent, cmd_boolean },
221 { "noproxy", &opt.no_proxy, cmd_vector },
222 { "numtries", &opt.ntry, cmd_number_inf },/* deprecated*/
223 { "outputdocument", &opt.output_document, cmd_file },
224 { "pagerequisites", &opt.page_requisites, cmd_boolean },
225 { "passiveftp", &opt.ftp_pasv, cmd_boolean },
226 { "passwd", &opt.ftp_passwd, cmd_string },/* deprecated*/
227 { "password", &opt.passwd, cmd_string },
228 { "postdata", &opt.post_data, cmd_string },
229 { "postfile", &opt.post_file_name, cmd_file },
230 { "preferfamily", NULL, cmd_spec_prefer_family },
231 { "preservepermissions", &opt.preserve_perm, cmd_boolean },
233 { "privatekey", &opt.private_key, cmd_file },
234 { "privatekeytype", &opt.private_key_type, cmd_cert_type },
236 { "progress", &opt.progress_type, cmd_spec_progress },
237 { "protocoldirectories", &opt.protocol_directories, cmd_boolean },
238 { "proxypasswd", &opt.proxy_passwd, cmd_string }, /* deprecated */
239 { "proxypassword", &opt.proxy_passwd, cmd_string },
240 { "proxyuser", &opt.proxy_user, cmd_string },
241 { "quiet", &opt.quiet, cmd_boolean },
242 { "quota", &opt.quota, cmd_bytes_sum },
244 { "randomfile", &opt.random_file, cmd_file },
246 { "randomwait", &opt.random_wait, cmd_boolean },
247 { "readtimeout", &opt.read_timeout, cmd_time },
248 { "reclevel", &opt.reclevel, cmd_number_inf },
249 { "recursive", NULL, cmd_spec_recursive },
250 { "referer", &opt.referer, cmd_string },
251 { "regextype", &opt.regex_type, cmd_spec_regex_type },
252 { "reject", &opt.rejects, cmd_vector },
253 { "rejectregex", &opt.rejectregex_s, cmd_string },
254 { "relativeonly", &opt.relative_only, cmd_boolean },
255 { "remoteencoding", &opt.encoding_remote, cmd_string },
256 { "removelisting", &opt.remove_listing, cmd_boolean },
257 { "reportspeed", &opt.report_bps, cmd_spec_report_speed},
258 { "restrictfilenames", NULL, cmd_spec_restrict_file_names },
259 { "retrsymlinks", &opt.retr_symlinks, cmd_boolean },
260 { "retryconnrefused", &opt.retry_connrefused, cmd_boolean },
261 { "robots", &opt.use_robots, cmd_boolean },
262 { "savecookies", &opt.cookies_output, cmd_file },
263 { "saveheaders", &opt.save_headers, cmd_boolean },
265 { "secureprotocol", &opt.secure_protocol, cmd_spec_secure_protocol },
267 { "serverresponse", &opt.server_response, cmd_boolean },
268 { "showalldnsentries", &opt.show_all_dns_entries, cmd_boolean },
269 { "spanhosts", &opt.spanhost, cmd_boolean },
270 { "spider", &opt.spider, cmd_boolean },
271 { "strictcomments", &opt.strict_comments, cmd_boolean },
272 { "timeout", NULL, cmd_spec_timeout },
273 { "timestamping", &opt.timestamping, cmd_boolean },
274 { "tries", &opt.ntry, cmd_number_inf },
275 { "trustservernames", &opt.trustservernames, cmd_boolean },
276 { "unlink", &opt.unlink, cmd_boolean },
277 { "useproxy", &opt.use_proxy, cmd_boolean },
278 { "user", &opt.user, cmd_string },
279 { "useragent", NULL, cmd_spec_useragent },
280 { "useservertimestamps", &opt.useservertimestamps, cmd_boolean },
281 { "verbose", NULL, cmd_spec_verbose },
282 { "wait", &opt.wait, cmd_time },
283 { "waitretry", &opt.waitretry, cmd_time },
284 { "warccdx", &opt.warc_cdx_enabled, cmd_boolean },
285 { "warccdxdedup", &opt.warc_cdx_dedup_filename, cmd_file },
287 { "warccompression", &opt.warc_compression_enabled, cmd_boolean },
289 { "warcdigests", &opt.warc_digests_enabled, cmd_boolean },
290 { "warcfile", &opt.warc_filename, cmd_file },
291 { "warcheader", NULL, cmd_spec_warc_header },
292 { "warckeeplog", &opt.warc_keep_log, cmd_boolean },
293 { "warcmaxsize", &opt.warc_maxsize, cmd_bytes },
294 { "warctempdir", &opt.warc_tempdir, cmd_directory },
296 { "wdebug", &opt.wdebug, cmd_boolean },
300 /* Look up CMDNAME in the commands[] and return its position in the
301 array. If CMDNAME is not found, return -1. */
304 command_by_name (const char *cmdname)
306 /* Use binary search for speed. Wget has ~100 commands, which
307 guarantees a worst case performance of 7 string comparisons. */
308 int lo = 0, hi = countof (commands) - 1;
312 int mid = (lo + hi) >> 1;
313 int cmp = strcasecmp (cmdname, commands[mid].name);
324 /* Reset the variables to default values. */
330 /* Most of the default values are 0 (and 0.0, NULL, and false).
331 Just reset everything, and fill in the non-zero values. Note
332 that initializing pointers to NULL this way is technically
333 illegal, but porting Wget to a machine where NULL is not all-zero
334 bit pattern will be the least of the implementors' worries. */
341 opt.add_hostdir = true;
345 opt.http_keep_alive = true;
346 opt.use_proxy = true;
347 tmp = getenv ("no_proxy");
349 opt.no_proxy = sepstring (tmp);
350 opt.prefer_family = prefer_none;
351 opt.allow_cache = true;
353 opt.read_timeout = 900;
354 opt.use_robots = true;
356 opt.remove_listing = true;
358 opt.dot_bytes = 1024;
359 opt.dot_spacing = 10;
360 opt.dots_in_line = 50;
362 opt.dns_cache = true;
366 opt.check_cert = true;
369 /* The default for file name restriction defaults to the OS type. */
370 #if defined(WINDOWS) || defined(MSDOS) || defined(__CYGWIN__)
371 opt.restrict_files_os = restrict_windows;
373 opt.restrict_files_os = restrict_unix;
375 opt.restrict_files_ctrl = true;
376 opt.restrict_files_nonascii = false;
377 opt.restrict_files_case = restrict_no_case_restriction;
379 opt.regex_type = regex_type_posix;
381 opt.max_redirect = 20;
386 opt.enable_iri = true;
388 opt.enable_iri = false;
391 opt.encoding_remote = NULL;
393 opt.useservertimestamps = true;
394 opt.show_all_dns_entries = false;
396 opt.warc_maxsize = 0; /* 1024 * 1024 * 1024; */
398 opt.warc_compression_enabled = true;
400 opt.warc_compression_enabled = false;
402 opt.warc_digests_enabled = true;
403 opt.warc_cdx_enabled = false;
404 opt.warc_cdx_dedup_filename = NULL;
405 opt.warc_tempdir = NULL;
406 opt.warc_keep_log = true;
409 /* Return the user's home directory (strdup-ed), or NULL if none is
414 static char *buf = NULL;
415 static char *home, *ret;
419 home = getenv ("HOME");
425 /* Under MSDOS, if $HOME isn't defined, use the directory where
426 `wget.exe' resides. */
427 const char *_w32_get_argv0 (void); /* in libwatt.a/pcconfig.c */
430 buff = _w32_get_argv0 ();
432 p = strrchr (buf, '/'); /* djgpp */
434 p = strrchr (buf, '\\'); /* others */
438 buff = malloc (len + 1);
442 strncpy (buff, _w32_get_argv0 (), len);
446 #elif !defined(WINDOWS)
447 /* If HOME is not defined, try getting it from the password
449 struct passwd *pwd = getpwuid (getuid ());
450 if (!pwd || !pwd->pw_dir)
454 /* Under Windows, if $HOME isn't defined, use the directory where
455 `wget.exe' resides. */
461 ret = home ? xstrdup (home) : NULL;
468 /* Check the 'WGETRC' environment variable and return the file name
469 if 'WGETRC' is set and is a valid file.
470 If the `WGETRC' variable exists but the file does not exist, the
471 function will exit(). */
473 wgetrc_env_file_name (void)
475 char *env = getenv ("WGETRC");
478 if (!file_exists_p (env))
480 fprintf (stderr, _("%s: WGETRC points to %s, which doesn't exist.\n"),
484 return xstrdup (env);
489 /* Check for the existance of '$HOME/.wgetrc' and return its path
490 if it exists and is set. */
492 wgetrc_user_file_name (void)
496 /* If that failed, try $HOME/.wgetrc (or equivalent). */
499 file = "SYS$LOGIN:.wgetrc";
500 #else /* def __VMS */
503 file = aprintf ("%s/.wgetrc", home);
505 #endif /* def __VMS [else] */
509 if (!file_exists_p (file))
517 /* Return the path to the user's .wgetrc. This is either the value of
518 `WGETRC' environment variable, or `$HOME/.wgetrc'.
520 Additionally, for windows, look in the directory where wget.exe
523 wgetrc_file_name (void)
525 char *file = wgetrc_env_file_name ();
529 file = wgetrc_user_file_name ();
532 /* Under Windows, if we still haven't found .wgetrc, look for the file
533 `wget.ini' in the directory where `wget.exe' resides; we do this for
534 backward compatibility with previous versions of Wget.
535 SYSTEM_WGETRC should not be defined under WINDOWS. */
538 char *home = home_dir ();
544 file = aprintf ("%s/wget.ini", home);
545 if (!file_exists_p (file))
558 /* Return values of parse_line. */
566 static enum parse_line parse_line (const char *, char **, char **, int *);
567 static bool setval_internal (int, const char *, const char *);
568 static bool setval_internal_tilde (int, const char *, const char *);
570 /* Initialize variables from a wgetrc file. Returns zero (failure) if
571 there were errors in the file. */
574 run_wgetrc (const char *file)
582 fp = fopen (file, "r");
585 fprintf (stderr, _("%s: Cannot read %s (%s).\n"), exec_name,
586 file, strerror (errno));
587 return true; /* not a fatal error */
590 while (getline (&line, &bufsize, fp) > 0)
592 char *com = NULL, *val = NULL;
595 /* Parse the line. */
596 switch (parse_line (line, &com, &val, &comind))
599 /* If everything is OK, set the value. */
600 if (!setval_internal_tilde (comind, com, val))
602 fprintf (stderr, _("%s: Error in %s at line %d.\n"),
603 exec_name, file, ln);
607 case line_syntax_error:
608 fprintf (stderr, _("%s: Syntax error in %s at line %d.\n"),
609 exec_name, file, ln);
612 case line_unknown_command:
613 fprintf (stderr, _("%s: Unknown command %s in %s at line %d.\n"),
614 exec_name, quote (com), file, ln);
632 /* Initialize the defaults and run the system wgetrc and user's own
637 char *file, *env_sysrc;
640 /* Run a non-standard system rc file when the according environment
641 variable has been set. For internal testing purposes only! */
642 env_sysrc = getenv ("SYSTEM_WGETRC");
643 if (env_sysrc && file_exists_p (env_sysrc))
645 ok &= run_wgetrc (env_sysrc);
646 /* If there are any problems parsing the system wgetrc file, tell
650 fprintf (stderr, _("\
651 Parsing system wgetrc file (env SYSTEM_WGETRC) failed. Please check\n\
653 or specify a different file using --config.\n"), env_sysrc);
657 /* Otherwise, if SYSTEM_WGETRC is defined, use it. */
659 else if (file_exists_p (SYSTEM_WGETRC))
660 ok &= run_wgetrc (SYSTEM_WGETRC);
661 /* If there are any problems parsing the system wgetrc file, tell
665 fprintf (stderr, _("\
666 Parsing system wgetrc file failed. Please check\n\
668 or specify a different file using --config.\n"), SYSTEM_WGETRC);
672 /* Override it with your own, if one exists. */
673 file = wgetrc_file_name ();
676 /* #### We should canonicalize `file' and SYSTEM_WGETRC with
677 something like realpath() before comparing them with `strcmp' */
679 if (!strcmp (file, SYSTEM_WGETRC))
681 fprintf (stderr, _("\
682 %s: Warning: Both system and user wgetrc point to %s.\n"),
683 exec_name, quote (file));
687 ok &= run_wgetrc (file);
689 /* If there were errors processing either `.wgetrc', abort. */
697 /* Remove dashes and underscores from S, modifying S in the
703 char *t = s; /* t - tortoise */
704 char *h = s; /* h - hare */
706 if (*h == '_' || *h == '-')
713 /* Parse the line pointed by line, with the syntax:
714 <sp>* command <sp>* = <sp>* value <sp>*
715 Uses malloc to allocate space for command and value.
717 Returns one of line_ok, line_empty, line_syntax_error, or
718 line_unknown_command.
720 In case of line_ok, *COM and *VAL point to freshly allocated
721 strings, and *COMIND points to com's index. In case of error or
722 empty line, their values are unmodified. */
724 static enum parse_line
725 parse_line (const char *line, char **com, char **val, int *comind)
728 const char *end = line + strlen (line);
729 const char *cmdstart, *cmdend;
730 const char *valstart, *valend;
735 /* Skip leading and trailing whitespace. */
736 while (*line && c_isspace (*line))
738 while (end > line && c_isspace (end[-1]))
741 /* Skip empty lines and comments. */
742 if (!*line || *line == '#')
748 while (p < end && (c_isalnum (*p) || *p == '_' || *p == '-'))
752 /* Skip '=', as well as any space before or after it. */
753 while (p < end && c_isspace (*p))
755 if (p == end || *p != '=')
756 return line_syntax_error;
758 while (p < end && c_isspace (*p))
764 /* The syntax is valid (even though the command might not be). Fill
765 in the command name and value. */
766 *com = strdupdelim (cmdstart, cmdend);
767 *val = strdupdelim (valstart, valend);
769 /* The line now known to be syntactically correct. Check whether
770 the command is valid. */
771 BOUNDED_TO_ALLOCA (cmdstart, cmdend, cmdcopy);
773 ind = command_by_name (cmdcopy);
775 return line_unknown_command;
777 /* Report success to the caller. */
782 #if defined(WINDOWS) || defined(MSDOS)
783 # define ISSEP(c) ((c) == '/' || (c) == '\\')
785 # define ISSEP(c) ((c) == '/')
788 /* Run commands[comind].action. */
791 setval_internal (int comind, const char *com, const char *val)
793 assert (0 <= comind && ((size_t) comind) < countof (commands));
794 DEBUGP (("Setting %s (%s) to %s\n", com, commands[comind].name, val));
795 return commands[comind].action (com, val, commands[comind].place);
799 setval_internal_tilde (int comind, const char *com, const char *val)
805 ret = setval_internal (comind, com, val);
807 /* We make tilde expansion for cmd_file and cmd_directory */
808 if (((commands[comind].action == cmd_file) ||
809 (commands[comind].action == cmd_directory))
810 && ret && (*val == '~' && ISSEP (val[1])))
812 pstring = commands[comind].place;
816 homelen = strlen (home);
817 while (homelen && ISSEP (home[homelen - 1]))
818 home[--homelen] = '\0';
820 /* Skip the leading "~/". */
821 for (++val; ISSEP (*val); val++)
823 *pstring = concat_strings (home, "/", val, (char *)0);
829 /* Run command COM with value VAL. If running the command produces an
830 error, report the error and exit.
832 This is intended to be called from main() to modify Wget's behavior
833 through command-line switches. Since COM is hard-coded in main(),
834 it is not canonicalized, and this aborts when COM is not found.
836 If COMIND's are exported to init.h, this function will be changed
837 to accept COMIND directly. */
840 setoptval (const char *com, const char *val, const char *optname)
842 /* Prepend "--" to OPTNAME. */
843 char *dd_optname = (char *) alloca (2 + strlen (optname) + 1);
846 strcpy (dd_optname + 2, optname);
848 assert (val != NULL);
849 if (!setval_internal (command_by_name (com), dd_optname, val))
853 /* Parse OPT into command and value and run it. For example,
854 run_command("foo=bar") is equivalent to setoptval("foo", "bar").
855 This is used by the `--execute' flag in main.c. */
858 run_command (const char *opt)
862 switch (parse_line (opt, &com, &val, &comind))
865 if (!setval_internal (comind, com, val))
871 fprintf (stderr, _("%s: Invalid --execute command %s\n"),
872 exec_name, quote (opt));
877 /* Generic helper functions, for use with `commands'. */
879 /* Forward declarations: */
884 static bool decode_string (const char *, const struct decode_item *, int, int *);
885 static bool simple_atoi (const char *, const char *, int *);
886 static bool simple_atof (const char *, const char *, double *);
888 #define CMP1(p, c0) (c_tolower((p)[0]) == (c0) && (p)[1] == '\0')
890 #define CMP2(p, c0, c1) (c_tolower((p)[0]) == (c0) \
891 && c_tolower((p)[1]) == (c1) \
894 #define CMP3(p, c0, c1, c2) (c_tolower((p)[0]) == (c0) \
895 && c_tolower((p)[1]) == (c1) \
896 && c_tolower((p)[2]) == (c2) \
900 /* Store the boolean value from VAL to PLACE. COM is ignored,
901 except for error messages. */
903 cmd_boolean (const char *com, const char *val, void *place)
907 if (CMP2 (val, 'o', 'n') || CMP3 (val, 'y', 'e', 's') || CMP1 (val, '1'))
908 /* "on", "yes" and "1" mean true. */
910 else if (CMP3 (val, 'o', 'f', 'f') || CMP2 (val, 'n', 'o') || CMP1 (val, '0'))
911 /* "off", "no" and "0" mean false. */
916 _("%s: %s: Invalid boolean %s; use `on' or `off'.\n"),
917 exec_name, com, quote (val));
921 *(bool *) place = value;
925 /* Set the non-negative integer value from VAL to PLACE. With
926 incorrect specification, the number remains unchanged. */
928 cmd_number (const char *com, const char *val, void *place)
930 if (!simple_atoi (val, val + strlen (val), place)
931 || *(int *) place < 0)
933 fprintf (stderr, _("%s: %s: Invalid number %s.\n"),
934 exec_name, com, quote (val));
940 /* Similar to cmd_number(), only accepts `inf' as a synonym for 0. */
942 cmd_number_inf (const char *com, const char *val, void *place)
944 if (!strcasecmp (val, "inf"))
949 return cmd_number (com, val, place);
952 /* Copy (strdup) the string at COM to a new location and place a
953 pointer to *PLACE. */
955 cmd_string (const char *com, const char *val, void *place)
957 char **pstring = (char **)place;
959 xfree_null (*pstring);
960 *pstring = xstrdup (val);
964 /* Like cmd_string but ensure the string is upper case. */
966 cmd_string_uppercase (const char *com, const char *val, void *place)
969 pstring = (char **)place;
970 xfree_null (*pstring);
972 *pstring = xmalloc (strlen (val) + 1);
974 for (q = *pstring; *val; val++, q++)
975 *q = c_toupper (*val);
982 /* Like cmd_string, but handles tilde-expansion when reading a user's
983 `.wgetrc'. In that case, and if VAL begins with `~', the tilde
984 gets expanded to the user's home directory. */
986 cmd_file (const char *com, const char *val, void *place)
988 char **pstring = (char **)place;
990 xfree_null (*pstring);
992 /* #### If VAL is empty, perhaps should set *PLACE to NULL. */
994 *pstring = xstrdup (val);
996 #if defined(WINDOWS) || defined(MSDOS)
997 /* Convert "\" to "/". */
1000 for (s = *pstring; *s; s++)
1008 /* Like cmd_file, but strips trailing '/' characters. */
1010 cmd_directory (const char *com, const char *val, void *place)
1014 /* Call cmd_file() for tilde expansion and separator
1015 canonicalization (backslash -> slash under Windows). These
1016 things should perhaps be in a separate function. */
1017 if (!cmd_file (com, val, place))
1020 s = *(char **)place;
1022 while (t > s && *--t == '/')
1028 /* Split VAL by space to a vector of values, and append those values
1029 to vector pointed to by the PLACE argument. If VAL is empty, the
1030 PLACE vector is cleared instead. */
1033 cmd_vector (const char *com, const char *val, void *place)
1035 char ***pvec = (char ***)place;
1038 *pvec = merge_vecs (*pvec, sepstring (val));
1048 cmd_directory_vector (const char *com, const char *val, void *place)
1050 char ***pvec = (char ***)place;
1054 /* Strip the trailing slashes from directories. */
1057 seps = sepstring (val);
1058 for (t = seps; t && *t; t++)
1060 int len = strlen (*t);
1061 /* Skip degenerate case of root directory. */
1064 if ((*t)[len - 1] == '/')
1065 (*t)[len - 1] = '\0';
1068 *pvec = merge_vecs (*pvec, seps);
1078 /* Engine for cmd_bytes and cmd_bytes_sum: converts a string such as
1079 "100k" or "2.5G" to a floating point number. */
1082 parse_bytes_helper (const char *val, double *result)
1084 double number, mult;
1085 const char *end = val + strlen (val);
1087 /* Check for "inf". */
1088 if (0 == strcmp (val, "inf"))
1094 /* Strip trailing whitespace. */
1095 while (val < end && c_isspace (end[-1]))
1100 switch (c_tolower (end[-1]))
1103 --end, mult = 1024.0;
1106 --end, mult = 1048576.0;
1109 --end, mult = 1073741824.0;
1112 --end, mult = 1099511627776.0;
1115 /* Not a recognized suffix: assume it's a digit. (If not,
1116 simple_atof will raise an error.) */
1120 /* Skip leading and trailing whitespace. */
1121 while (val < end && c_isspace (*val))
1123 while (val < end && c_isspace (end[-1]))
1128 if (!simple_atof (val, end, &number) || number < 0)
1131 *result = number * mult;
1135 /* Parse VAL as a number and set its value to PLACE (which should
1138 By default, the value is assumed to be in bytes. If "K", "M", or
1139 "G" are appended, the value is multiplied with 1<<10, 1<<20, or
1140 1<<30, respectively. Floating point values are allowed and are
1141 cast to integer before use. The idea is to be able to use things
1142 like 1.5k instead of "1536".
1144 The string "inf" is returned as 0.
1146 In case of error, false is returned and memory pointed to by PLACE
1147 remains unmodified. */
1150 cmd_bytes (const char *com, const char *val, void *place)
1153 if (!parse_bytes_helper (val, &byte_value))
1155 fprintf (stderr, _("%s: %s: Invalid byte value %s\n"),
1156 exec_name, com, quote (val));
1159 *(wgint *)place = (wgint)byte_value;
1163 /* Like cmd_bytes, but PLACE is interpreted as a pointer to
1164 SIZE_SUM. It works by converting the string to double, therefore
1165 working with values up to 2^53-1 without loss of precision. This
1166 value (8192 TB) is large enough to serve for a while. */
1169 cmd_bytes_sum (const char *com, const char *val, void *place)
1172 if (!parse_bytes_helper (val, &byte_value))
1174 fprintf (stderr, _("%s: %s: Invalid byte value %s\n"),
1175 exec_name, com, quote (val));
1178 *(SUM_SIZE_INT *) place = (SUM_SIZE_INT) byte_value;
1182 /* Store the value of VAL to *OUT. The value is a time period, by
1183 default expressed in seconds, but also accepting suffixes "m", "h",
1184 "d", and "w" for minutes, hours, days, and weeks respectively. */
1187 cmd_time (const char *com, const char *val, void *place)
1189 double number, mult;
1190 const char *end = val + strlen (val);
1192 /* Strip trailing whitespace. */
1193 while (val < end && c_isspace (end[-1]))
1199 fprintf (stderr, _("%s: %s: Invalid time period %s\n"),
1200 exec_name, com, quote (val));
1204 switch (c_tolower (end[-1]))
1207 --end, mult = 1; /* seconds */
1210 --end, mult = 60; /* minutes */
1213 --end, mult = 3600; /* hours */
1216 --end, mult = 86400.0; /* days */
1219 --end, mult = 604800.0; /* weeks */
1222 /* Not a recognized suffix: assume it belongs to the number.
1223 (If not, simple_atof will raise an error.) */
1227 /* Skip leading and trailing whitespace. */
1228 while (val < end && c_isspace (*val))
1230 while (val < end && c_isspace (end[-1]))
1235 if (!simple_atof (val, end, &number))
1238 *(double *)place = number * mult;
1244 cmd_cert_type (const char *com, const char *val, void *place)
1246 static const struct decode_item choices[] = {
1247 { "pem", keyfile_pem },
1248 { "der", keyfile_asn1 },
1249 { "asn1", keyfile_asn1 },
1251 int ok = decode_string (val, choices, countof (choices), place);
1253 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1258 /* Specialized helper functions, used by `commands' to handle some
1259 options specially. */
1261 static bool check_user_specified_header (const char *);
1264 cmd_spec_dirstruct (const char *com, const char *val, void *place_ignored)
1266 if (!cmd_boolean (com, val, &opt.dirstruct))
1268 /* Since dirstruct behaviour is explicitly changed, no_dirstruct
1269 must be affected inversely. */
1271 opt.no_dirstruct = false;
1273 opt.no_dirstruct = true;
1278 cmd_spec_header (const char *com, const char *val, void *place_ignored)
1280 /* Empty value means reset the list of headers. */
1283 free_vec (opt.user_headers);
1284 opt.user_headers = NULL;
1288 if (!check_user_specified_header (val))
1290 fprintf (stderr, _("%s: %s: Invalid header %s.\n"),
1291 exec_name, com, quote (val));
1294 opt.user_headers = vec_append (opt.user_headers, val);
1299 cmd_spec_warc_header (const char *com, const char *val, void *place_ignored)
1301 /* Empty value means reset the list of headers. */
1304 free_vec (opt.warc_user_headers);
1305 opt.warc_user_headers = NULL;
1309 if (!check_user_specified_header (val))
1311 fprintf (stderr, _("%s: %s: Invalid WARC header %s.\n"),
1312 exec_name, com, quote (val));
1315 opt.warc_user_headers = vec_append (opt.warc_user_headers, val);
1320 cmd_spec_htmlify (const char *com, const char *val, void *place_ignored)
1322 int flag = cmd_boolean (com, val, &opt.htmlify);
1323 if (flag && !opt.htmlify)
1324 opt.remove_listing = false;
1328 /* Set the "mirror" mode. It means: recursive download, timestamping,
1329 no limit on max. recursion depth, and don't remove listings. */
1332 cmd_spec_mirror (const char *com, const char *val, void *place_ignored)
1336 if (!cmd_boolean (com, val, &mirror))
1340 opt.recursive = true;
1341 if (!opt.no_dirstruct)
1342 opt.dirstruct = true;
1343 opt.timestamping = true;
1344 opt.reclevel = INFINITE_RECURSION;
1345 opt.remove_listing = false;
1350 /* Validate --prefer-family and set the choice. Allowed values are
1351 "IPv4", "IPv6", and "none". */
1354 cmd_spec_prefer_family (const char *com, const char *val, void *place_ignored)
1356 static const struct decode_item choices[] = {
1357 { "IPv4", prefer_ipv4 },
1358 { "IPv6", prefer_ipv6 },
1359 { "none", prefer_none },
1361 int prefer_family = prefer_none;
1362 int ok = decode_string (val, choices, countof (choices), &prefer_family);
1364 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1365 opt.prefer_family = prefer_family;
1369 /* Set progress.type to VAL, but verify that it's a valid progress
1370 implementation before that. */
1373 cmd_spec_progress (const char *com, const char *val, void *place_ignored)
1375 if (!valid_progress_implementation_p (val))
1377 fprintf (stderr, _("%s: %s: Invalid progress type %s.\n"),
1378 exec_name, com, quote (val));
1381 xfree_null (opt.progress_type);
1383 /* Don't call set_progress_implementation here. It will be called
1384 in main() when it becomes clear what the log output is. */
1385 opt.progress_type = xstrdup (val);
1389 /* Set opt.recursive to VAL as with cmd_boolean. If opt.recursive is
1390 set to true, also set opt.dirstruct to true, unless opt.no_dirstruct
1394 cmd_spec_recursive (const char *com, const char *val, void *place_ignored)
1396 if (!cmd_boolean (com, val, &opt.recursive))
1400 if (opt.recursive && !opt.no_dirstruct)
1401 opt.dirstruct = true;
1406 /* Validate --regex-type and set the choice. */
1409 cmd_spec_regex_type (const char *com, const char *val, void *place_ignored)
1411 static const struct decode_item choices[] = {
1412 { "posix", regex_type_posix },
1414 { "pcre", regex_type_pcre },
1417 int regex_type = regex_type_posix;
1418 int ok = decode_string (val, choices, countof (choices), ®ex_type);
1420 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1421 opt.regex_type = regex_type;
1426 cmd_spec_restrict_file_names (const char *com, const char *val, void *place_ignored)
1428 int restrict_os = opt.restrict_files_os;
1429 int restrict_ctrl = opt.restrict_files_ctrl;
1430 int restrict_case = opt.restrict_files_case;
1431 int restrict_nonascii = opt.restrict_files_nonascii;
1435 #define VAL_IS(string_literal) BOUNDED_EQUAL (val, end, string_literal)
1439 end = strchr (val, ',');
1441 end = val + strlen (val);
1443 if (VAL_IS ("unix"))
1444 restrict_os = restrict_unix;
1445 else if (VAL_IS ("windows"))
1446 restrict_os = restrict_windows;
1447 else if (VAL_IS ("lowercase"))
1448 restrict_case = restrict_lowercase;
1449 else if (VAL_IS ("uppercase"))
1450 restrict_case = restrict_uppercase;
1451 else if (VAL_IS ("nocontrol"))
1452 restrict_ctrl = false;
1453 else if (VAL_IS ("ascii"))
1454 restrict_nonascii = true;
1457 fprintf (stderr, _("\
1458 %s: %s: Invalid restriction %s,\n\
1459 use [unix|windows],[lowercase|uppercase],[nocontrol],[ascii].\n"),
1460 exec_name, com, quote (val));
1467 while (*val && *end);
1471 opt.restrict_files_os = restrict_os;
1472 opt.restrict_files_ctrl = restrict_ctrl;
1473 opt.restrict_files_case = restrict_case;
1474 opt.restrict_files_nonascii = restrict_nonascii;
1480 cmd_spec_report_speed (const char *com, const char *val, void *place_ignored)
1482 opt.report_bps = strcasecmp (val, "bits") == 0;
1483 if (!opt.report_bps)
1484 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1485 return opt.report_bps;
1490 cmd_spec_secure_protocol (const char *com, const char *val, void *place)
1492 static const struct decode_item choices[] = {
1493 { "auto", secure_protocol_auto },
1494 { "sslv2", secure_protocol_sslv2 },
1495 { "sslv3", secure_protocol_sslv3 },
1496 { "tlsv1", secure_protocol_tlsv1 },
1498 int ok = decode_string (val, choices, countof (choices), place);
1500 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1505 /* Set all three timeout values. */
1508 cmd_spec_timeout (const char *com, const char *val, void *place_ignored)
1511 if (!cmd_time (com, val, &value))
1513 opt.read_timeout = value;
1514 opt.connect_timeout = value;
1515 opt.dns_timeout = value;
1520 cmd_spec_useragent (const char *com, const char *val, void *place_ignored)
1522 /* Disallow embedded newlines. */
1523 if (strchr (val, '\n'))
1525 fprintf (stderr, _("%s: %s: Invalid value %s.\n"),
1526 exec_name, com, quote (val));
1529 xfree_null (opt.useragent);
1530 opt.useragent = xstrdup (val);
1534 /* The "verbose" option cannot be cmd_boolean because the variable is
1535 not bool -- it's of type int (-1 means uninitialized because of
1536 some random hackery for disallowing -q -v). */
1539 cmd_spec_verbose (const char *com, const char *val, void *place_ignored)
1542 if (cmd_boolean (com, val, &flag))
1550 /* Miscellaneous useful routines. */
1552 /* A very simple atoi clone, more useful than atoi because it works on
1553 delimited strings, and has error reportage. Returns true on success,
1554 false on failure. If successful, stores result to *DEST. */
1557 simple_atoi (const char *beg, const char *end, int *dest)
1560 bool negative = false;
1561 const char *p = beg;
1563 while (p < end && c_isspace (*p))
1565 if (p < end && (*p == '-' || *p == '+'))
1567 negative = (*p == '-');
1573 /* Read negative numbers in a separate loop because the most
1574 negative integer cannot be represented as a positive number. */
1577 for (; p < end && c_isdigit (*p); p++)
1579 int next = (10 * result) + (*p - '0');
1581 return false; /* overflow */
1585 for (; p < end && c_isdigit (*p); p++)
1587 int next = (10 * result) - (*p - '0');
1589 return false; /* underflow */
1600 /* Trivial atof, with error reporting. Handles "<digits>[.<digits>]",
1601 doesn't handle exponential notation. Returns true on success,
1602 false on failure. In case of success, stores its result to
1606 simple_atof (const char *beg, const char *end, double *dest)
1610 bool negative = false;
1611 bool seen_dot = false;
1612 bool seen_digit = false;
1615 const char *p = beg;
1617 while (p < end && c_isspace (*p))
1619 if (p < end && (*p == '-' || *p == '+'))
1621 negative = (*p == '-');
1625 for (; p < end; p++)
1631 result = (10 * result) + (ch - '0');
1633 result += (ch - '0') / (divider *= 10);
1655 /* Verify that the user-specified header in S is valid. It must
1656 contain a colon preceded by non-white-space characters and must not
1657 contain newlines. */
1660 check_user_specified_header (const char *s)
1664 for (p = s; *p && *p != ':' && !c_isspace (*p); p++)
1666 /* The header MUST contain `:' preceded by at least one
1667 non-whitespace character. */
1668 if (*p != ':' || p == s)
1670 /* The header MUST NOT contain newlines. */
1671 if (strchr (s, '\n'))
1676 /* Decode VAL into a number, according to ITEMS. */
1679 decode_string (const char *val, const struct decode_item *items, int itemcount,
1683 for (i = 0; i < itemcount; i++)
1684 if (0 == strcasecmp (val, items[i].name))
1686 *place = items[i].code;
1693 void cleanup_html_url (void);
1694 void spider_cleanup (void);
1697 /* Free the memory allocated by global variables. */
1701 /* Free external resources, close files, etc. */
1703 /* Close WARC file. */
1704 if (opt.warc_filename != 0)
1710 if (fclose (output_stream) == EOF)
1711 inform_exit_status (CLOSEFAILED);
1713 /* No need to check for error because Wget flushes its output (and
1714 checks for errors) after any data arrives. */
1716 /* We're exiting anyway so there's no real need to call free()
1717 hundreds of times. Skipping the frees will make Wget exit
1720 However, when detecting leaks, it's crucial to free() everything
1721 because then you can find the real leaks, i.e. the allocated
1722 memory which grows with the size of the program. */
1728 cleanup_html_url ();
1733 for (i = 0; i < nurl; i++)
1737 extern acc_t *netrc_list;
1738 free_netrc (netrc_list);
1740 xfree_null (opt.choose_config);
1741 xfree_null (opt.lfilename);
1742 xfree_null (opt.dir_prefix);
1743 xfree_null (opt.input_filename);
1744 xfree_null (opt.output_document);
1745 free_vec (opt.accepts);
1746 free_vec (opt.rejects);
1747 free_vec (opt.excludes);
1748 free_vec (opt.includes);
1749 free_vec (opt.domains);
1750 free_vec (opt.follow_tags);
1751 free_vec (opt.ignore_tags);
1752 xfree_null (opt.progress_type);
1753 xfree_null (opt.ftp_user);
1754 xfree_null (opt.ftp_passwd);
1755 xfree_null (opt.ftp_proxy);
1756 xfree_null (opt.https_proxy);
1757 xfree_null (opt.http_proxy);
1758 free_vec (opt.no_proxy);
1759 xfree_null (opt.useragent);
1760 xfree_null (opt.referer);
1761 xfree_null (opt.http_user);
1762 xfree_null (opt.http_passwd);
1763 free_vec (opt.user_headers);
1764 free_vec (opt.warc_user_headers);
1766 xfree_null (opt.cert_file);
1767 xfree_null (opt.private_key);
1768 xfree_null (opt.ca_directory);
1769 xfree_null (opt.ca_cert);
1770 xfree_null (opt.random_file);
1771 xfree_null (opt.egd_file);
1773 xfree_null (opt.bind_address);
1774 xfree_null (opt.cookies_input);
1775 xfree_null (opt.cookies_output);
1776 xfree_null (opt.user);
1777 xfree_null (opt.passwd);
1778 xfree_null (opt.base_href);
1779 xfree_null (opt.method);
1781 #endif /* DEBUG_MALLOC */
1784 /* Unit testing routines. */
1789 test_commands_sorted()
1791 int prev_idx = 0, next_idx = 1;
1792 int command_count = countof (commands) - 1;
1794 while (next_idx <= command_count)
1796 cmp = strcasecmp (commands[prev_idx].name, commands[next_idx].name);
1799 mu_assert ("FAILED", false);
1812 test_cmd_spec_restrict_file_names()
1817 int expected_restrict_files_os;
1818 int expected_restrict_files_ctrl;
1819 int expected_restrict_files_case;
1822 { "windows", restrict_windows, true, restrict_no_case_restriction, true },
1823 { "windows,", restrict_windows, true, restrict_no_case_restriction, true },
1824 { "windows,lowercase", restrict_windows, true, restrict_lowercase, true },
1825 { "unix,nocontrol,lowercase,", restrict_unix, false, restrict_lowercase, true },
1828 for (i = 0; i < sizeof(test_array)/sizeof(test_array[0]); ++i)
1833 res = cmd_spec_restrict_file_names ("dummy", test_array[i].val, NULL);
1836 fprintf (stderr, "test_cmd_spec_restrict_file_names: TEST %d\n", i); fflush (stderr);
1837 fprintf (stderr, "opt.restrict_files_os: %d\n", opt.restrict_files_os); fflush (stderr);
1838 fprintf (stderr, "opt.restrict_files_ctrl: %d\n", opt.restrict_files_ctrl); fflush (stderr);
1839 fprintf (stderr, "opt.restrict_files_case: %d\n", opt.restrict_files_case); fflush (stderr);
1841 mu_assert ("test_cmd_spec_restrict_file_names: wrong result",
1842 res == test_array[i].result
1843 && opt.restrict_files_os == test_array[i].expected_restrict_files_os
1844 && opt.restrict_files_ctrl == test_array[i].expected_restrict_files_ctrl
1845 && opt.restrict_files_case == test_array[i].expected_restrict_files_case);
1851 #endif /* TESTING */