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 */
71 #include "spider.h" /* for spider_cleanup */
79 #define CMD_DECLARE(func) static bool func (const char *, const char *, void *)
81 CMD_DECLARE (cmd_boolean);
82 CMD_DECLARE (cmd_bytes);
83 CMD_DECLARE (cmd_bytes_sum);
85 CMD_DECLARE (cmd_cert_type);
87 CMD_DECLARE (cmd_directory_vector);
88 CMD_DECLARE (cmd_number);
89 CMD_DECLARE (cmd_number_inf);
90 CMD_DECLARE (cmd_string);
91 CMD_DECLARE (cmd_string_uppercase);
92 CMD_DECLARE (cmd_file);
93 CMD_DECLARE (cmd_directory);
94 CMD_DECLARE (cmd_time);
95 CMD_DECLARE (cmd_vector);
97 CMD_DECLARE (cmd_spec_dirstruct);
98 CMD_DECLARE (cmd_spec_header);
99 CMD_DECLARE (cmd_spec_warc_header);
100 CMD_DECLARE (cmd_spec_htmlify);
101 CMD_DECLARE (cmd_spec_mirror);
102 CMD_DECLARE (cmd_spec_prefer_family);
103 CMD_DECLARE (cmd_spec_progress);
104 CMD_DECLARE (cmd_spec_recursive);
105 CMD_DECLARE (cmd_spec_regex_type);
106 CMD_DECLARE (cmd_spec_restrict_file_names);
107 CMD_DECLARE (cmd_spec_report_speed);
109 CMD_DECLARE (cmd_spec_secure_protocol);
111 CMD_DECLARE (cmd_spec_timeout);
112 CMD_DECLARE (cmd_spec_useragent);
113 CMD_DECLARE (cmd_spec_verbose);
115 /* List of recognized commands, each consisting of name, place and
116 function. When adding a new command, simply add it to the list,
117 but be sure to keep the list sorted alphabetically, as
118 command_by_name's binary search depends on it. Also, be sure to
119 add any entries that allocate memory (e.g. cmd_string and
120 cmd_vector) to the cleanup() function below. */
122 static const struct {
125 bool (*action) (const char *, const char *, void *);
127 /* KEEP THIS LIST ALPHABETICALLY SORTED */
128 { "accept", &opt.accepts, cmd_vector },
129 { "acceptregex", &opt.acceptregex_s, cmd_string },
130 { "addhostdir", &opt.add_hostdir, cmd_boolean },
131 { "adjustextension", &opt.adjust_extension, cmd_boolean },
132 { "alwaysrest", &opt.always_rest, cmd_boolean }, /* deprecated */
133 { "askpassword", &opt.ask_passwd, cmd_boolean },
134 { "authnochallenge", &opt.auth_without_challenge,
136 { "background", &opt.background, cmd_boolean },
137 { "backupconverted", &opt.backup_converted, cmd_boolean },
138 { "backups", &opt.backups, cmd_number },
139 { "base", &opt.base_href, cmd_string },
140 { "bindaddress", &opt.bind_address, cmd_string },
141 { "bodydata", &opt.body_data, cmd_string },
142 { "bodyfile", &opt.body_file, cmd_string },
144 { "cacertificate", &opt.ca_cert, cmd_file },
146 { "cache", &opt.allow_cache, cmd_boolean },
148 { "cadirectory", &opt.ca_directory, cmd_directory },
149 { "certificate", &opt.cert_file, cmd_file },
150 { "certificatetype", &opt.cert_type, cmd_cert_type },
151 { "checkcertificate", &opt.check_cert, cmd_boolean },
153 { "chooseconfig", &opt.choose_config, cmd_file },
154 { "connecttimeout", &opt.connect_timeout, cmd_time },
155 { "contentdisposition", &opt.content_disposition, cmd_boolean },
156 { "contentonerror", &opt.content_on_error, cmd_boolean },
157 { "continue", &opt.always_rest, cmd_boolean },
158 { "convertlinks", &opt.convert_links, cmd_boolean },
159 { "cookies", &opt.cookies, cmd_boolean },
160 { "cutdirs", &opt.cut_dirs, cmd_number },
161 { "debug", &opt.debug, cmd_boolean },
162 { "defaultpage", &opt.default_page, cmd_string },
163 { "deleteafter", &opt.delete_after, cmd_boolean },
164 { "dirprefix", &opt.dir_prefix, cmd_directory },
165 { "dirstruct", NULL, cmd_spec_dirstruct },
166 { "dnscache", &opt.dns_cache, cmd_boolean },
167 { "dnstimeout", &opt.dns_timeout, cmd_time },
168 { "domains", &opt.domains, cmd_vector },
169 { "dotbytes", &opt.dot_bytes, cmd_bytes },
170 { "dotsinline", &opt.dots_in_line, cmd_number },
171 { "dotspacing", &opt.dot_spacing, cmd_number },
172 { "dotstyle", &opt.dot_style, cmd_string }, /* deprecated */
174 { "egdfile", &opt.egd_file, cmd_file },
176 { "excludedirectories", &opt.excludes, cmd_directory_vector },
177 { "excludedomains", &opt.exclude_domains, cmd_vector },
178 { "followftp", &opt.follow_ftp, cmd_boolean },
179 { "followtags", &opt.follow_tags, cmd_vector },
180 { "forcehtml", &opt.force_html, cmd_boolean },
181 { "ftppasswd", &opt.ftp_passwd, cmd_string }, /* deprecated */
182 { "ftppassword", &opt.ftp_passwd, cmd_string },
183 { "ftpproxy", &opt.ftp_proxy, cmd_string },
185 { "ftpstmlf", &opt.ftp_stmlf, cmd_boolean },
186 #endif /* def __VMS */
187 { "ftpuser", &opt.ftp_user, cmd_string },
188 { "glob", &opt.ftp_glob, cmd_boolean },
189 { "header", NULL, cmd_spec_header },
190 { "htmlextension", &opt.adjust_extension, cmd_boolean }, /* deprecated */
191 { "htmlify", NULL, cmd_spec_htmlify },
192 { "httpkeepalive", &opt.http_keep_alive, cmd_boolean },
193 { "httppasswd", &opt.http_passwd, cmd_string }, /* deprecated */
194 { "httppassword", &opt.http_passwd, cmd_string },
195 { "httpproxy", &opt.http_proxy, cmd_string },
197 { "httpsonly", &opt.https_only, cmd_boolean },
199 { "httpsproxy", &opt.https_proxy, cmd_string },
200 { "httpuser", &opt.http_user, cmd_string },
201 { "ignorecase", &opt.ignore_case, cmd_boolean },
202 { "ignorelength", &opt.ignore_length, cmd_boolean },
203 { "ignoretags", &opt.ignore_tags, cmd_vector },
204 { "includedirectories", &opt.includes, cmd_directory_vector },
206 { "inet4only", &opt.ipv4_only, cmd_boolean },
207 { "inet6only", &opt.ipv6_only, cmd_boolean },
209 { "input", &opt.input_filename, cmd_file },
210 { "iri", &opt.enable_iri, cmd_boolean },
211 { "keepsessioncookies", &opt.keep_session_cookies, cmd_boolean },
212 { "limitrate", &opt.limit_rate, cmd_bytes },
213 { "loadcookies", &opt.cookies_input, cmd_file },
214 { "localencoding", &opt.locale, cmd_string },
215 { "logfile", &opt.lfilename, cmd_file },
216 { "login", &opt.ftp_user, cmd_string },/* deprecated*/
217 { "maxredirect", &opt.max_redirect, cmd_number },
218 { "method", &opt.method, cmd_string_uppercase },
219 { "mirror", NULL, cmd_spec_mirror },
220 { "netrc", &opt.netrc, cmd_boolean },
221 { "noclobber", &opt.noclobber, cmd_boolean },
222 { "noconfig", &opt.noconfig, 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 { "showprogress", &opt.show_progress, cmd_boolean },
273 { "spanhosts", &opt.spanhost, cmd_boolean },
274 { "spider", &opt.spider, cmd_boolean },
275 { "startpos", &opt.start_pos, cmd_bytes },
276 { "strictcomments", &opt.strict_comments, cmd_boolean },
277 { "timeout", NULL, cmd_spec_timeout },
278 { "timestamping", &opt.timestamping, cmd_boolean },
279 { "tries", &opt.ntry, cmd_number_inf },
280 { "trustservernames", &opt.trustservernames, cmd_boolean },
281 { "unlink", &opt.unlink, cmd_boolean },
282 { "useproxy", &opt.use_proxy, cmd_boolean },
283 { "user", &opt.user, cmd_string },
284 { "useragent", NULL, cmd_spec_useragent },
285 { "useservertimestamps", &opt.useservertimestamps, cmd_boolean },
286 { "verbose", NULL, cmd_spec_verbose },
287 { "wait", &opt.wait, cmd_time },
288 { "waitretry", &opt.waitretry, cmd_time },
289 { "warccdx", &opt.warc_cdx_enabled, cmd_boolean },
290 { "warccdxdedup", &opt.warc_cdx_dedup_filename, cmd_file },
292 { "warccompression", &opt.warc_compression_enabled, cmd_boolean },
294 { "warcdigests", &opt.warc_digests_enabled, cmd_boolean },
295 { "warcfile", &opt.warc_filename, cmd_file },
296 { "warcheader", NULL, cmd_spec_warc_header },
297 { "warckeeplog", &opt.warc_keep_log, cmd_boolean },
298 { "warcmaxsize", &opt.warc_maxsize, cmd_bytes },
299 { "warctempdir", &opt.warc_tempdir, cmd_directory },
301 { "wdebug", &opt.wdebug, cmd_boolean },
305 /* Look up CMDNAME in the commands[] and return its position in the
306 array. If CMDNAME is not found, return -1. */
309 command_by_name (const char *cmdname)
311 /* Use binary search for speed. Wget has ~100 commands, which
312 guarantees a worst case performance of 7 string comparisons. */
313 int lo = 0, hi = countof (commands) - 1;
317 int mid = (lo + hi) >> 1;
318 int cmp = strcasecmp (cmdname, commands[mid].name);
329 /* Reset the variables to default values. */
335 /* Most of the default values are 0 (and 0.0, NULL, and false).
336 Just reset everything, and fill in the non-zero values. Note
337 that initializing pointers to NULL this way is technically
338 illegal, but porting Wget to a machine where NULL is not all-zero
339 bit pattern will be the least of the implementors' worries. */
346 opt.add_hostdir = true;
350 opt.http_keep_alive = true;
351 opt.use_proxy = true;
352 tmp = getenv ("no_proxy");
354 opt.no_proxy = sepstring (tmp);
355 opt.prefer_family = prefer_none;
356 opt.allow_cache = true;
358 opt.read_timeout = 900;
359 opt.use_robots = true;
361 opt.remove_listing = true;
363 opt.dot_bytes = 1024;
364 opt.dot_spacing = 10;
365 opt.dots_in_line = 50;
367 opt.dns_cache = true;
371 opt.check_cert = true;
374 /* The default for file name restriction defaults to the OS type. */
375 #if defined(WINDOWS) || defined(MSDOS) || defined(__CYGWIN__)
376 opt.restrict_files_os = restrict_windows;
378 opt.restrict_files_os = restrict_unix;
380 opt.restrict_files_ctrl = true;
381 opt.restrict_files_nonascii = false;
382 opt.restrict_files_case = restrict_no_case_restriction;
384 opt.regex_type = regex_type_posix;
386 opt.max_redirect = 20;
391 opt.enable_iri = true;
393 opt.enable_iri = false;
396 opt.encoding_remote = NULL;
398 opt.useservertimestamps = true;
399 opt.show_all_dns_entries = false;
401 opt.warc_maxsize = 0; /* 1024 * 1024 * 1024; */
403 opt.warc_compression_enabled = true;
405 opt.warc_compression_enabled = false;
407 opt.warc_digests_enabled = true;
408 opt.warc_cdx_enabled = false;
409 opt.warc_cdx_dedup_filename = NULL;
410 opt.warc_tempdir = NULL;
411 opt.warc_keep_log = true;
413 /* Use a negative value to mark the absence of --start-pos option */
415 opt.show_progress = false;
416 opt.noscroll = false;
419 /* Return the user's home directory (strdup-ed), or NULL if none is
424 static char *buf = NULL;
425 static char *home, *ret;
429 home = getenv ("HOME");
435 /* Under MSDOS, if $HOME isn't defined, use the directory where
436 `wget.exe' resides. */
437 const char *_w32_get_argv0 (void); /* in libwatt.a/pcconfig.c */
440 buff = _w32_get_argv0 ();
442 p = strrchr (buf, '/'); /* djgpp */
444 p = strrchr (buf, '\\'); /* others */
448 buff = malloc (len + 1);
452 strncpy (buff, _w32_get_argv0 (), len);
456 #elif !defined(WINDOWS)
457 /* If HOME is not defined, try getting it from the password
459 struct passwd *pwd = getpwuid (getuid ());
460 if (!pwd || !pwd->pw_dir)
464 /* Under Windows, if $HOME isn't defined, use the directory where
465 `wget.exe' resides. */
471 ret = home ? xstrdup (home) : NULL;
477 /* Check the 'WGETRC' environment variable and return the file name
478 if 'WGETRC' is set and is a valid file.
479 If the `WGETRC' variable exists but the file does not exist, the
480 function will exit(). */
482 wgetrc_env_file_name (void)
484 char *env = getenv ("WGETRC");
487 if (!file_exists_p (env))
489 fprintf (stderr, _("%s: WGETRC points to %s, which doesn't exist.\n"),
493 return xstrdup (env);
498 /* Check for the existance of '$HOME/.wgetrc' and return its path
499 if it exists and is set. */
501 wgetrc_user_file_name (void)
505 /* If that failed, try $HOME/.wgetrc (or equivalent). */
508 file = "SYS$LOGIN:.wgetrc";
509 #else /* def __VMS */
512 file = aprintf ("%s/.wgetrc", home);
514 #endif /* def __VMS [else] */
518 if (!file_exists_p (file))
526 /* Return the path to the user's .wgetrc. This is either the value of
527 `WGETRC' environment variable, or `$HOME/.wgetrc'.
529 Additionally, for windows, look in the directory where wget.exe
532 wgetrc_file_name (void)
534 char *file = wgetrc_env_file_name ();
538 file = wgetrc_user_file_name ();
541 /* Under Windows, if we still haven't found .wgetrc, look for the file
542 `wget.ini' in the directory where `wget.exe' resides; we do this for
543 backward compatibility with previous versions of Wget.
544 SYSTEM_WGETRC should not be defined under WINDOWS. */
547 char *home = home_dir ();
553 file = aprintf ("%s/wget.ini", home);
554 if (!file_exists_p (file))
567 /* Return values of parse_line. */
575 static enum parse_line parse_line (const char *, char **, char **, int *);
576 static bool setval_internal (int, const char *, const char *);
577 static bool setval_internal_tilde (int, const char *, const char *);
579 /* Initialize variables from a wgetrc file. Returns zero (failure) if
580 there were errors in the file. */
583 run_wgetrc (const char *file)
591 fp = fopen (file, "r");
594 fprintf (stderr, _("%s: Cannot read %s (%s).\n"), exec_name,
595 file, strerror (errno));
596 return true; /* not a fatal error */
599 while (getline (&line, &bufsize, fp) > 0)
601 char *com = NULL, *val = NULL;
604 /* Parse the line. */
605 switch (parse_line (line, &com, &val, &comind))
608 /* If everything is OK, set the value. */
609 if (!setval_internal_tilde (comind, com, val))
611 fprintf (stderr, _("%s: Error in %s at line %d.\n"),
612 exec_name, file, ln);
616 case line_syntax_error:
617 fprintf (stderr, _("%s: Syntax error in %s at line %d.\n"),
618 exec_name, file, ln);
621 case line_unknown_command:
622 fprintf (stderr, _("%s: Unknown command %s in %s at line %d.\n"),
623 exec_name, quote (com), file, ln);
641 /* Initialize the defaults and run the system wgetrc and user's own
646 char *file, *env_sysrc;
649 /* Run a non-standard system rc file when the according environment
650 variable has been set. For internal testing purposes only! */
651 env_sysrc = getenv ("SYSTEM_WGETRC");
652 if (env_sysrc && file_exists_p (env_sysrc))
654 ok &= run_wgetrc (env_sysrc);
655 /* If there are any problems parsing the system wgetrc file, tell
659 fprintf (stderr, _("\
660 Parsing system wgetrc file (env SYSTEM_WGETRC) failed. Please check\n\
662 or specify a different file using --config.\n"), env_sysrc);
666 /* Otherwise, if SYSTEM_WGETRC is defined, use it. */
668 else if (file_exists_p (SYSTEM_WGETRC))
669 ok &= run_wgetrc (SYSTEM_WGETRC);
670 /* If there are any problems parsing the system wgetrc file, tell
674 fprintf (stderr, _("\
675 Parsing system wgetrc file failed. Please check\n\
677 or specify a different file using --config.\n"), SYSTEM_WGETRC);
681 /* Override it with your own, if one exists. */
682 file = wgetrc_file_name ();
685 /* #### We should canonicalize `file' and SYSTEM_WGETRC with
686 something like realpath() before comparing them with `strcmp' */
688 if (!strcmp (file, SYSTEM_WGETRC))
690 fprintf (stderr, _("\
691 %s: Warning: Both system and user wgetrc point to %s.\n"),
692 exec_name, quote (file));
696 ok &= run_wgetrc (file);
698 /* If there were errors processing either `.wgetrc', abort. */
706 /* Remove dashes and underscores from S, modifying S in the
712 char *t = s; /* t - tortoise */
713 char *h = s; /* h - hare */
715 if (*h == '_' || *h == '-')
722 /* Parse the line pointed by line, with the syntax:
723 <sp>* command <sp>* = <sp>* value <sp>*
724 Uses malloc to allocate space for command and value.
726 Returns one of line_ok, line_empty, line_syntax_error, or
727 line_unknown_command.
729 In case of line_ok, *COM and *VAL point to freshly allocated
730 strings, and *COMIND points to com's index. In case of error or
731 empty line, their values are unmodified. */
733 static enum parse_line
734 parse_line (const char *line, char **com, char **val, int *comind)
737 const char *end = line + strlen (line);
738 const char *cmdstart, *cmdend;
739 const char *valstart, *valend;
744 /* Skip leading and trailing whitespace. */
745 while (*line && c_isspace (*line))
747 while (end > line && c_isspace (end[-1]))
750 /* Skip empty lines and comments. */
751 if (!*line || *line == '#')
757 while (p < end && (c_isalnum (*p) || *p == '_' || *p == '-'))
761 /* Skip '=', as well as any space before or after it. */
762 while (p < end && c_isspace (*p))
764 if (p == end || *p != '=')
765 return line_syntax_error;
767 while (p < end && c_isspace (*p))
773 /* The syntax is valid (even though the command might not be). Fill
774 in the command name and value. */
775 *com = strdupdelim (cmdstart, cmdend);
776 *val = strdupdelim (valstart, valend);
778 /* The line now known to be syntactically correct. Check whether
779 the command is valid. */
780 BOUNDED_TO_ALLOCA (cmdstart, cmdend, cmdcopy);
782 ind = command_by_name (cmdcopy);
784 return line_unknown_command;
786 /* Report success to the caller. */
791 #if defined(WINDOWS) || defined(MSDOS)
792 # define ISSEP(c) ((c) == '/' || (c) == '\\')
794 # define ISSEP(c) ((c) == '/')
797 /* Run commands[comind].action. */
800 setval_internal (int comind, const char *com, const char *val)
802 assert (0 <= comind && ((size_t) comind) < countof (commands));
803 DEBUGP (("Setting %s (%s) to %s\n", com, commands[comind].name, val));
804 return commands[comind].action (com, val, commands[comind].place);
808 setval_internal_tilde (int comind, const char *com, const char *val)
814 ret = setval_internal (comind, com, val);
816 /* We make tilde expansion for cmd_file and cmd_directory */
817 if (((commands[comind].action == cmd_file) ||
818 (commands[comind].action == cmd_directory))
819 && ret && (*val == '~' && ISSEP (val[1])))
821 pstring = commands[comind].place;
825 homelen = strlen (home);
826 while (homelen && ISSEP (home[homelen - 1]))
827 home[--homelen] = '\0';
829 /* Skip the leading "~/". */
830 for (++val; ISSEP (*val); val++)
832 *pstring = concat_strings (home, "/", val, (char *)0);
838 /* Run command COM with value VAL. If running the command produces an
839 error, report the error and exit.
841 This is intended to be called from main() to modify Wget's behavior
842 through command-line switches. Since COM is hard-coded in main(),
843 it is not canonicalized, and this aborts when COM is not found.
845 If COMIND's are exported to init.h, this function will be changed
846 to accept COMIND directly. */
849 setoptval (const char *com, const char *val, const char *optname)
851 /* Prepend "--" to OPTNAME. */
852 char *dd_optname = (char *) alloca (2 + strlen (optname) + 1);
855 strcpy (dd_optname + 2, optname);
857 assert (val != NULL);
858 if (!setval_internal (command_by_name (com), dd_optname, val))
862 /* Parse OPT into command and value and run it. For example,
863 run_command("foo=bar") is equivalent to setoptval("foo", "bar").
864 This is used by the `--execute' flag in main.c. */
867 run_command (const char *cmdopt)
871 switch (parse_line (cmdopt, &com, &val, &comind))
874 if (!setval_internal (comind, com, val))
880 fprintf (stderr, _("%s: Invalid --execute command %s\n"),
881 exec_name, quote (cmdopt));
886 /* Generic helper functions, for use with `commands'. */
888 /* Forward declarations: */
893 static bool decode_string (const char *, const struct decode_item *, int, int *);
894 static bool simple_atoi (const char *, const char *, int *);
895 static bool simple_atof (const char *, const char *, double *);
897 #define CMP1(p, c0) (c_tolower((p)[0]) == (c0) && (p)[1] == '\0')
899 #define CMP2(p, c0, c1) (c_tolower((p)[0]) == (c0) \
900 && c_tolower((p)[1]) == (c1) \
903 #define CMP3(p, c0, c1, c2) (c_tolower((p)[0]) == (c0) \
904 && c_tolower((p)[1]) == (c1) \
905 && c_tolower((p)[2]) == (c2) \
909 /* Store the boolean value from VAL to PLACE. COM is ignored,
910 except for error messages. */
912 cmd_boolean (const char *com, const char *val, void *place)
916 if (CMP2 (val, 'o', 'n') || CMP3 (val, 'y', 'e', 's') || CMP1 (val, '1'))
917 /* "on", "yes" and "1" mean true. */
919 else if (CMP3 (val, 'o', 'f', 'f') || CMP2 (val, 'n', 'o') || CMP1 (val, '0'))
920 /* "off", "no" and "0" mean false. */
925 _("%s: %s: Invalid boolean %s; use `on' or `off'.\n"),
926 exec_name, com, quote (val));
930 *(bool *) place = value;
934 /* Set the non-negative integer value from VAL to PLACE. With
935 incorrect specification, the number remains unchanged. */
937 cmd_number (const char *com, const char *val, void *place)
939 if (!simple_atoi (val, val + strlen (val), place)
940 || *(int *) place < 0)
942 fprintf (stderr, _("%s: %s: Invalid number %s.\n"),
943 exec_name, com, quote (val));
949 /* Similar to cmd_number(), only accepts `inf' as a synonym for 0. */
951 cmd_number_inf (const char *com, const char *val, void *place)
953 if (!strcasecmp (val, "inf"))
958 return cmd_number (com, val, place);
961 /* Copy (strdup) the string at COM to a new location and place a
962 pointer to *PLACE. */
964 cmd_string (const char *com _GL_UNUSED, const char *val, void *place)
966 char **pstring = (char **)place;
968 xfree_null (*pstring);
969 *pstring = xstrdup (val);
973 /* Like cmd_string but ensure the string is upper case. */
975 cmd_string_uppercase (const char *com _GL_UNUSED, const char *val, void *place)
978 pstring = (char **)place;
979 xfree_null (*pstring);
981 *pstring = xmalloc (strlen (val) + 1);
983 for (q = *pstring; *val; val++, q++)
984 *q = c_toupper (*val);
991 /* Like cmd_string, but handles tilde-expansion when reading a user's
992 `.wgetrc'. In that case, and if VAL begins with `~', the tilde
993 gets expanded to the user's home directory. */
995 cmd_file (const char *com _GL_UNUSED, const char *val, void *place)
997 char **pstring = (char **)place;
999 xfree_null (*pstring);
1001 /* #### If VAL is empty, perhaps should set *PLACE to NULL. */
1003 *pstring = xstrdup (val);
1005 #if defined(WINDOWS) || defined(MSDOS)
1006 /* Convert "\" to "/". */
1009 for (s = *pstring; *s; s++)
1017 /* Like cmd_file, but strips trailing '/' characters. */
1019 cmd_directory (const char *com, const char *val, void *place)
1023 /* Call cmd_file() for tilde expansion and separator
1024 canonicalization (backslash -> slash under Windows). These
1025 things should perhaps be in a separate function. */
1026 if (!cmd_file (com, val, place))
1029 s = *(char **)place;
1031 while (t > s && *--t == '/')
1037 /* Split VAL by space to a vector of values, and append those values
1038 to vector pointed to by the PLACE argument. If VAL is empty, the
1039 PLACE vector is cleared instead. */
1042 cmd_vector (const char *com _GL_UNUSED, const char *val, void *place)
1044 char ***pvec = (char ***)place;
1047 *pvec = merge_vecs (*pvec, sepstring (val));
1057 cmd_directory_vector (const char *com _GL_UNUSED, const char *val, void *place)
1059 char ***pvec = (char ***)place;
1063 /* Strip the trailing slashes from directories. */
1066 seps = sepstring (val);
1067 for (t = seps; t && *t; t++)
1069 int len = strlen (*t);
1070 /* Skip degenerate case of root directory. */
1073 if ((*t)[len - 1] == '/')
1074 (*t)[len - 1] = '\0';
1077 *pvec = merge_vecs (*pvec, seps);
1087 /* Engine for cmd_bytes and cmd_bytes_sum: converts a string such as
1088 "100k" or "2.5G" to a floating point number. */
1091 parse_bytes_helper (const char *val, double *result)
1093 double number, mult;
1094 const char *end = val + strlen (val);
1096 /* Check for "inf". */
1097 if (0 == strcmp (val, "inf"))
1103 /* Strip trailing whitespace. */
1104 while (val < end && c_isspace (end[-1]))
1109 switch (c_tolower (end[-1]))
1112 --end, mult = 1024.0;
1115 --end, mult = 1048576.0;
1118 --end, mult = 1073741824.0;
1121 --end, mult = 1099511627776.0;
1124 /* Not a recognized suffix: assume it's a digit. (If not,
1125 simple_atof will raise an error.) */
1129 /* Skip leading and trailing whitespace. */
1130 while (val < end && c_isspace (*val))
1132 while (val < end && c_isspace (end[-1]))
1137 if (!simple_atof (val, end, &number) || number < 0)
1140 *result = number * mult;
1144 /* Parse VAL as a number and set its value to PLACE (which should
1147 By default, the value is assumed to be in bytes. If "K", "M", or
1148 "G" are appended, the value is multiplied with 1<<10, 1<<20, or
1149 1<<30, respectively. Floating point values are allowed and are
1150 cast to integer before use. The idea is to be able to use things
1151 like 1.5k instead of "1536".
1153 The string "inf" is returned as 0.
1155 In case of error, false is returned and memory pointed to by PLACE
1156 remains unmodified. */
1159 cmd_bytes (const char *com, const char *val, void *place)
1162 if (!parse_bytes_helper (val, &byte_value))
1164 fprintf (stderr, _("%s: %s: Invalid byte value %s\n"),
1165 exec_name, com, quote (val));
1168 *(wgint *)place = (wgint)byte_value;
1172 /* Like cmd_bytes, but PLACE is interpreted as a pointer to
1173 SIZE_SUM. It works by converting the string to double, therefore
1174 working with values up to 2^53-1 without loss of precision. This
1175 value (8192 TB) is large enough to serve for a while. */
1178 cmd_bytes_sum (const char *com, const char *val, void *place)
1181 if (!parse_bytes_helper (val, &byte_value))
1183 fprintf (stderr, _("%s: %s: Invalid byte value %s\n"),
1184 exec_name, com, quote (val));
1187 *(SUM_SIZE_INT *) place = (SUM_SIZE_INT) byte_value;
1191 /* Store the value of VAL to *OUT. The value is a time period, by
1192 default expressed in seconds, but also accepting suffixes "m", "h",
1193 "d", and "w" for minutes, hours, days, and weeks respectively. */
1196 cmd_time (const char *com, const char *val, void *place)
1198 double number, mult;
1199 const char *end = val + strlen (val);
1201 /* Strip trailing whitespace. */
1202 while (val < end && c_isspace (end[-1]))
1208 fprintf (stderr, _("%s: %s: Invalid time period %s\n"),
1209 exec_name, com, quote (val));
1213 switch (c_tolower (end[-1]))
1216 --end, mult = 1; /* seconds */
1219 --end, mult = 60; /* minutes */
1222 --end, mult = 3600; /* hours */
1225 --end, mult = 86400.0; /* days */
1228 --end, mult = 604800.0; /* weeks */
1231 /* Not a recognized suffix: assume it belongs to the number.
1232 (If not, simple_atof will raise an error.) */
1236 /* Skip leading and trailing whitespace. */
1237 while (val < end && c_isspace (*val))
1239 while (val < end && c_isspace (end[-1]))
1244 if (!simple_atof (val, end, &number))
1247 *(double *)place = number * mult;
1253 cmd_cert_type (const char *com, const char *val, void *place)
1255 static const struct decode_item choices[] = {
1256 { "pem", keyfile_pem },
1257 { "der", keyfile_asn1 },
1258 { "asn1", keyfile_asn1 },
1260 int ok = decode_string (val, choices, countof (choices), place);
1262 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1267 /* Specialized helper functions, used by `commands' to handle some
1268 options specially. */
1270 static bool check_user_specified_header (const char *);
1273 cmd_spec_dirstruct (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1275 if (!cmd_boolean (com, val, &opt.dirstruct))
1277 /* Since dirstruct behaviour is explicitly changed, no_dirstruct
1278 must be affected inversely. */
1280 opt.no_dirstruct = false;
1282 opt.no_dirstruct = true;
1287 cmd_spec_header (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1289 /* Empty value means reset the list of headers. */
1292 free_vec (opt.user_headers);
1293 opt.user_headers = NULL;
1297 if (!check_user_specified_header (val))
1299 fprintf (stderr, _("%s: %s: Invalid header %s.\n"),
1300 exec_name, com, quote (val));
1303 opt.user_headers = vec_append (opt.user_headers, val);
1308 cmd_spec_warc_header (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1310 /* Empty value means reset the list of headers. */
1313 free_vec (opt.warc_user_headers);
1314 opt.warc_user_headers = NULL;
1318 if (!check_user_specified_header (val))
1320 fprintf (stderr, _("%s: %s: Invalid WARC header %s.\n"),
1321 exec_name, com, quote (val));
1324 opt.warc_user_headers = vec_append (opt.warc_user_headers, val);
1329 cmd_spec_htmlify (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1331 int flag = cmd_boolean (com, val, &opt.htmlify);
1332 if (flag && !opt.htmlify)
1333 opt.remove_listing = false;
1337 /* Set the "mirror" mode. It means: recursive download, timestamping,
1338 no limit on max. recursion depth, and don't remove listings. */
1341 cmd_spec_mirror (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1345 if (!cmd_boolean (com, val, &mirror))
1349 opt.recursive = true;
1350 if (!opt.no_dirstruct)
1351 opt.dirstruct = true;
1352 opt.timestamping = true;
1353 opt.reclevel = INFINITE_RECURSION;
1354 opt.remove_listing = false;
1359 /* Validate --prefer-family and set the choice. Allowed values are
1360 "IPv4", "IPv6", and "none". */
1363 cmd_spec_prefer_family (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1365 static const struct decode_item choices[] = {
1366 { "IPv4", prefer_ipv4 },
1367 { "IPv6", prefer_ipv6 },
1368 { "none", prefer_none },
1370 int prefer_family = prefer_none;
1371 int ok = decode_string (val, choices, countof (choices), &prefer_family);
1373 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1374 opt.prefer_family = prefer_family;
1378 /* Set progress.type to VAL, but verify that it's a valid progress
1379 implementation before that. */
1382 cmd_spec_progress (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1384 if (!valid_progress_implementation_p (val))
1386 fprintf (stderr, _("%s: %s: Invalid progress type %s.\n"),
1387 exec_name, com, quote (val));
1390 xfree_null (opt.progress_type);
1392 /* Don't call set_progress_implementation here. It will be called
1393 in main() when it becomes clear what the log output is. */
1394 opt.progress_type = xstrdup (val);
1398 /* Set opt.recursive to VAL as with cmd_boolean. If opt.recursive is
1399 set to true, also set opt.dirstruct to true, unless opt.no_dirstruct
1403 cmd_spec_recursive (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1405 if (!cmd_boolean (com, val, &opt.recursive))
1409 if (opt.recursive && !opt.no_dirstruct)
1410 opt.dirstruct = true;
1415 /* Validate --regex-type and set the choice. */
1418 cmd_spec_regex_type (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1420 static const struct decode_item choices[] = {
1421 { "posix", regex_type_posix },
1423 { "pcre", regex_type_pcre },
1426 int regex_type = regex_type_posix;
1427 int ok = decode_string (val, choices, countof (choices), ®ex_type);
1429 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1430 opt.regex_type = regex_type;
1435 cmd_spec_restrict_file_names (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1437 int restrict_os = opt.restrict_files_os;
1438 int restrict_ctrl = opt.restrict_files_ctrl;
1439 int restrict_case = opt.restrict_files_case;
1440 int restrict_nonascii = opt.restrict_files_nonascii;
1444 #define VAL_IS(string_literal) BOUNDED_EQUAL (val, end, string_literal)
1448 end = strchr (val, ',');
1450 end = val + strlen (val);
1452 if (VAL_IS ("unix"))
1453 restrict_os = restrict_unix;
1454 else if (VAL_IS ("windows"))
1455 restrict_os = restrict_windows;
1456 else if (VAL_IS ("lowercase"))
1457 restrict_case = restrict_lowercase;
1458 else if (VAL_IS ("uppercase"))
1459 restrict_case = restrict_uppercase;
1460 else if (VAL_IS ("nocontrol"))
1461 restrict_ctrl = false;
1462 else if (VAL_IS ("ascii"))
1463 restrict_nonascii = true;
1466 fprintf (stderr, _("\
1467 %s: %s: Invalid restriction %s,\n\
1468 use [unix|windows],[lowercase|uppercase],[nocontrol],[ascii].\n"),
1469 exec_name, com, quote (val));
1476 while (*val && *end);
1480 opt.restrict_files_os = restrict_os;
1481 opt.restrict_files_ctrl = restrict_ctrl;
1482 opt.restrict_files_case = restrict_case;
1483 opt.restrict_files_nonascii = restrict_nonascii;
1489 cmd_spec_report_speed (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1491 opt.report_bps = strcasecmp (val, "bits") == 0;
1492 if (!opt.report_bps)
1493 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1494 return opt.report_bps;
1499 cmd_spec_secure_protocol (const char *com, const char *val, void *place)
1501 static const struct decode_item choices[] = {
1502 { "auto", secure_protocol_auto },
1503 { "sslv2", secure_protocol_sslv2 },
1504 { "sslv3", secure_protocol_sslv3 },
1505 { "tlsv1", secure_protocol_tlsv1 },
1506 { "pfs", secure_protocol_pfs },
1508 int ok = decode_string (val, choices, countof (choices), place);
1510 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1515 /* Set all three timeout values. */
1518 cmd_spec_timeout (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1521 if (!cmd_time (com, val, &value))
1523 opt.read_timeout = value;
1524 opt.connect_timeout = value;
1525 opt.dns_timeout = value;
1530 cmd_spec_useragent (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1532 /* Disallow embedded newlines. */
1533 if (strchr (val, '\n'))
1535 fprintf (stderr, _("%s: %s: Invalid value %s.\n"),
1536 exec_name, com, quote (val));
1539 xfree_null (opt.useragent);
1540 opt.useragent = xstrdup (val);
1544 /* The "verbose" option cannot be cmd_boolean because the variable is
1545 not bool -- it's of type int (-1 means uninitialized because of
1546 some random hackery for disallowing -q -v). */
1549 cmd_spec_verbose (const char *com, const char *val, void *place_ignored _GL_UNUSED)
1552 if (cmd_boolean (com, val, &flag))
1560 /* Miscellaneous useful routines. */
1562 /* A very simple atoi clone, more useful than atoi because it works on
1563 delimited strings, and has error reportage. Returns true on success,
1564 false on failure. If successful, stores result to *DEST. */
1567 simple_atoi (const char *beg, const char *end, int *dest)
1570 bool negative = false;
1571 const char *p = beg;
1573 while (p < end && c_isspace (*p))
1575 if (p < end && (*p == '-' || *p == '+'))
1577 negative = (*p == '-');
1583 /* Read negative numbers in a separate loop because the most
1584 negative integer cannot be represented as a positive number. */
1587 for (; p < end && c_isdigit (*p); p++)
1589 int next = (10 * result) + (*p - '0');
1591 return false; /* overflow */
1595 for (; p < end && c_isdigit (*p); p++)
1597 int next = (10 * result) - (*p - '0');
1599 return false; /* underflow */
1610 /* Trivial atof, with error reporting. Handles "<digits>[.<digits>]",
1611 doesn't handle exponential notation. Returns true on success,
1612 false on failure. In case of success, stores its result to
1616 simple_atof (const char *beg, const char *end, double *dest)
1620 bool negative = false;
1621 bool seen_dot = false;
1622 bool seen_digit = false;
1625 const char *p = beg;
1627 while (p < end && c_isspace (*p))
1629 if (p < end && (*p == '-' || *p == '+'))
1631 negative = (*p == '-');
1635 for (; p < end; p++)
1641 result = (10 * result) + (ch - '0');
1643 result += (ch - '0') / (divider *= 10);
1665 /* Verify that the user-specified header in S is valid. It must
1666 contain a colon preceded by non-white-space characters and must not
1667 contain newlines. */
1670 check_user_specified_header (const char *s)
1674 for (p = s; *p && *p != ':' && !c_isspace (*p); p++)
1676 /* The header MUST contain `:' preceded by at least one
1677 non-whitespace character. */
1678 if (*p != ':' || p == s)
1680 /* The header MUST NOT contain newlines. */
1681 if (strchr (s, '\n'))
1686 /* Decode VAL into a number, according to ITEMS. */
1689 decode_string (const char *val, const struct decode_item *items, int itemcount,
1693 for (i = 0; i < itemcount; i++)
1694 if (0 == strcasecmp (val, items[i].name))
1696 *place = items[i].code;
1702 /* Free the memory allocated by global variables. */
1706 /* Free external resources, close files, etc. */
1708 /* Close WARC file. */
1709 if (opt.warc_filename != 0)
1715 if (fclose (output_stream) == EOF)
1716 inform_exit_status (CLOSEFAILED);
1718 /* No need to check for error because Wget flushes its output (and
1719 checks for errors) after any data arrives. */
1721 /* We're exiting anyway so there's no real need to call free()
1722 hundreds of times. Skipping the frees will make Wget exit
1725 However, when detecting leaks, it's crucial to free() everything
1726 because then you can find the real leaks, i.e. the allocated
1727 memory which grows with the size of the program. */
1733 cleanup_html_url ();
1737 netrc_cleanup (netrc_list);
1739 for (i = 0; i < nurl; i++)
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(void)
1795 for (i = 1; i < countof(commands); ++i)
1797 if (strcasecmp (commands[i - 1].name, commands[i].name) > 0)
1799 mu_assert ("FAILED", false);
1807 test_cmd_spec_restrict_file_names(void)
1810 static const struct {
1812 int expected_restrict_files_os;
1813 int expected_restrict_files_ctrl;
1814 int expected_restrict_files_case;
1817 { "windows", restrict_windows, true, restrict_no_case_restriction, true },
1818 { "windows,", restrict_windows, true, restrict_no_case_restriction, true },
1819 { "windows,lowercase", restrict_windows, true, restrict_lowercase, true },
1820 { "unix,nocontrol,lowercase,", restrict_unix, false, restrict_lowercase, true },
1823 for (i = 0; i < countof(test_array); ++i)
1828 res = cmd_spec_restrict_file_names ("dummy", test_array[i].val, NULL);
1831 fprintf (stderr, "test_cmd_spec_restrict_file_names: TEST %d\n", i); fflush (stderr);
1832 fprintf (stderr, "opt.restrict_files_os: %d\n", opt.restrict_files_os); fflush (stderr);
1833 fprintf (stderr, "opt.restrict_files_ctrl: %d\n", opt.restrict_files_ctrl); fflush (stderr);
1834 fprintf (stderr, "opt.restrict_files_case: %d\n", opt.restrict_files_case); fflush (stderr);
1836 mu_assert ("test_cmd_spec_restrict_file_names: wrong result",
1837 res == test_array[i].result
1838 && opt.restrict_files_os == test_array[i].expected_restrict_files_os
1839 && opt.restrict_files_ctrl == test_array[i].expected_restrict_files_ctrl
1840 && opt.restrict_files_case == test_array[i].expected_restrict_files_case);
1846 #endif /* TESTING */