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. */
41 /* not all systems provide PATH_MAX in limits.h */
43 # include <sys/param.h>
45 # define PATH_MAX MAXPATHLEN
64 #include "recur.h" /* for INFINITE_RECURSION */
65 #include "convert.h" /* for convert_cleanup */
66 #include "res.h" /* for res_cleanup */
67 #include "http.h" /* for http_cleanup */
68 #include "retr.h" /* for output_stream */
76 #define CMD_DECLARE(func) static bool func (const char *, const char *, void *)
78 CMD_DECLARE (cmd_boolean);
79 CMD_DECLARE (cmd_bytes);
80 CMD_DECLARE (cmd_bytes_sum);
82 CMD_DECLARE (cmd_cert_type);
84 CMD_DECLARE (cmd_directory_vector);
85 CMD_DECLARE (cmd_number);
86 CMD_DECLARE (cmd_number_inf);
87 CMD_DECLARE (cmd_string);
88 CMD_DECLARE (cmd_file);
89 CMD_DECLARE (cmd_directory);
90 CMD_DECLARE (cmd_time);
91 CMD_DECLARE (cmd_vector);
93 CMD_DECLARE (cmd_spec_dirstruct);
94 CMD_DECLARE (cmd_spec_header);
95 CMD_DECLARE (cmd_spec_warc_header);
96 CMD_DECLARE (cmd_spec_htmlify);
97 CMD_DECLARE (cmd_spec_mirror);
98 CMD_DECLARE (cmd_spec_prefer_family);
99 CMD_DECLARE (cmd_spec_progress);
100 CMD_DECLARE (cmd_spec_recursive);
101 CMD_DECLARE (cmd_spec_regex_type);
102 CMD_DECLARE (cmd_spec_restrict_file_names);
103 CMD_DECLARE (cmd_spec_report_speed);
105 CMD_DECLARE (cmd_spec_secure_protocol);
107 CMD_DECLARE (cmd_spec_timeout);
108 CMD_DECLARE (cmd_spec_useragent);
109 CMD_DECLARE (cmd_spec_verbose);
111 /* List of recognized commands, each consisting of name, place and
112 function. When adding a new command, simply add it to the list,
113 but be sure to keep the list sorted alphabetically, as
114 command_by_name's binary search depends on it. Also, be sure to
115 add any entries that allocate memory (e.g. cmd_string and
116 cmd_vector) to the cleanup() function below. */
118 static const struct {
121 bool (*action) (const char *, const char *, void *);
123 /* KEEP THIS LIST ALPHABETICALLY SORTED */
124 { "accept", &opt.accepts, cmd_vector },
125 { "acceptregex", &opt.acceptregex_s, cmd_string },
126 { "addhostdir", &opt.add_hostdir, cmd_boolean },
127 { "adjustextension", &opt.adjust_extension, cmd_boolean },
128 { "alwaysrest", &opt.always_rest, cmd_boolean }, /* deprecated */
129 { "askpassword", &opt.ask_passwd, cmd_boolean },
130 { "authnochallenge", &opt.auth_without_challenge,
132 { "background", &opt.background, cmd_boolean },
133 { "backupconverted", &opt.backup_converted, cmd_boolean },
134 { "backups", &opt.backups, cmd_number },
135 { "base", &opt.base_href, cmd_string },
136 { "bindaddress", &opt.bind_address, cmd_string },
138 { "cacertificate", &opt.ca_cert, cmd_file },
140 { "cache", &opt.allow_cache, cmd_boolean },
142 { "cadirectory", &opt.ca_directory, cmd_directory },
143 { "certificate", &opt.cert_file, cmd_file },
144 { "certificatetype", &opt.cert_type, cmd_cert_type },
145 { "checkcertificate", &opt.check_cert, cmd_boolean },
147 { "chooseconfig", &opt.choose_config, cmd_file },
148 { "connecttimeout", &opt.connect_timeout, cmd_time },
149 { "contentdisposition", &opt.content_disposition, cmd_boolean },
150 { "contentonerror", &opt.content_on_error, cmd_boolean },
151 { "continue", &opt.always_rest, cmd_boolean },
152 { "convertlinks", &opt.convert_links, cmd_boolean },
153 { "cookies", &opt.cookies, cmd_boolean },
154 { "cutdirs", &opt.cut_dirs, cmd_number },
156 { "debug", &opt.debug, cmd_boolean },
158 { "defaultpage", &opt.default_page, cmd_string},
159 { "deleteafter", &opt.delete_after, cmd_boolean },
160 { "dirprefix", &opt.dir_prefix, cmd_directory },
161 { "dirstruct", NULL, cmd_spec_dirstruct },
162 { "dnscache", &opt.dns_cache, cmd_boolean },
163 { "dnstimeout", &opt.dns_timeout, cmd_time },
164 { "domains", &opt.domains, cmd_vector },
165 { "dotbytes", &opt.dot_bytes, cmd_bytes },
166 { "dotsinline", &opt.dots_in_line, cmd_number },
167 { "dotspacing", &opt.dot_spacing, cmd_number },
168 { "dotstyle", &opt.dot_style, cmd_string }, /* deprecated */
170 { "egdfile", &opt.egd_file, cmd_file },
172 { "excludedirectories", &opt.excludes, cmd_directory_vector },
173 { "excludedomains", &opt.exclude_domains, cmd_vector },
174 { "followftp", &opt.follow_ftp, cmd_boolean },
175 { "followtags", &opt.follow_tags, cmd_vector },
176 { "forcehtml", &opt.force_html, cmd_boolean },
177 { "ftppasswd", &opt.ftp_passwd, cmd_string }, /* deprecated */
178 { "ftppassword", &opt.ftp_passwd, cmd_string },
179 { "ftpproxy", &opt.ftp_proxy, cmd_string },
181 { "ftpstmlf", &opt.ftp_stmlf, cmd_boolean },
182 #endif /* def __VMS */
183 { "ftpuser", &opt.ftp_user, cmd_string },
184 { "glob", &opt.ftp_glob, cmd_boolean },
185 { "header", NULL, cmd_spec_header },
186 { "htmlextension", &opt.adjust_extension, cmd_boolean }, /* deprecated */
187 { "htmlify", NULL, cmd_spec_htmlify },
188 { "httpkeepalive", &opt.http_keep_alive, cmd_boolean },
189 { "httppasswd", &opt.http_passwd, cmd_string }, /* deprecated */
190 { "httppassword", &opt.http_passwd, cmd_string },
191 { "httpproxy", &opt.http_proxy, cmd_string },
192 { "httpsproxy", &opt.https_proxy, cmd_string },
193 { "httpuser", &opt.http_user, cmd_string },
194 { "ignorecase", &opt.ignore_case, cmd_boolean },
195 { "ignorelength", &opt.ignore_length, cmd_boolean },
196 { "ignoretags", &opt.ignore_tags, cmd_vector },
197 { "includedirectories", &opt.includes, cmd_directory_vector },
199 { "inet4only", &opt.ipv4_only, cmd_boolean },
200 { "inet6only", &opt.ipv6_only, cmd_boolean },
202 { "input", &opt.input_filename, cmd_file },
203 { "iri", &opt.enable_iri, cmd_boolean },
204 { "keepsessioncookies", &opt.keep_session_cookies, cmd_boolean },
205 { "limitrate", &opt.limit_rate, cmd_bytes },
206 { "loadcookies", &opt.cookies_input, cmd_file },
207 { "localencoding", &opt.locale, cmd_string },
208 { "logfile", &opt.lfilename, cmd_file },
209 { "login", &opt.ftp_user, cmd_string },/* deprecated*/
210 { "maxredirect", &opt.max_redirect, cmd_number },
211 { "mirror", NULL, cmd_spec_mirror },
212 { "netrc", &opt.netrc, cmd_boolean },
213 { "noclobber", &opt.noclobber, cmd_boolean },
214 { "noparent", &opt.no_parent, cmd_boolean },
215 { "noproxy", &opt.no_proxy, cmd_vector },
216 { "numtries", &opt.ntry, cmd_number_inf },/* deprecated*/
217 { "outputdocument", &opt.output_document, cmd_file },
218 { "pagerequisites", &opt.page_requisites, cmd_boolean },
219 { "passiveftp", &opt.ftp_pasv, cmd_boolean },
220 { "passwd", &opt.ftp_passwd, cmd_string },/* deprecated*/
221 { "password", &opt.passwd, cmd_string },
222 { "postdata", &opt.post_data, cmd_string },
223 { "postfile", &opt.post_file_name, cmd_file },
224 { "preferfamily", NULL, cmd_spec_prefer_family },
225 { "preservepermissions", &opt.preserve_perm, cmd_boolean },
227 { "privatekey", &opt.private_key, cmd_file },
228 { "privatekeytype", &opt.private_key_type, cmd_cert_type },
230 { "progress", &opt.progress_type, cmd_spec_progress },
231 { "protocoldirectories", &opt.protocol_directories, cmd_boolean },
232 { "proxypasswd", &opt.proxy_passwd, cmd_string }, /* deprecated */
233 { "proxypassword", &opt.proxy_passwd, cmd_string },
234 { "proxyuser", &opt.proxy_user, cmd_string },
235 { "quiet", &opt.quiet, cmd_boolean },
236 { "quota", &opt.quota, cmd_bytes_sum },
238 { "randomfile", &opt.random_file, cmd_file },
240 { "randomwait", &opt.random_wait, cmd_boolean },
241 { "readtimeout", &opt.read_timeout, cmd_time },
242 { "reclevel", &opt.reclevel, cmd_number_inf },
243 { "recursive", NULL, cmd_spec_recursive },
244 { "referer", &opt.referer, cmd_string },
245 { "regextype", &opt.regex_type, cmd_spec_regex_type },
246 { "reject", &opt.rejects, cmd_vector },
247 { "rejectregex", &opt.rejectregex_s, cmd_string },
248 { "relativeonly", &opt.relative_only, cmd_boolean },
249 { "remoteencoding", &opt.encoding_remote, cmd_string },
250 { "removelisting", &opt.remove_listing, cmd_boolean },
251 { "reportspeed", &opt.report_bps, cmd_spec_report_speed},
252 { "restrictfilenames", NULL, cmd_spec_restrict_file_names },
253 { "retrsymlinks", &opt.retr_symlinks, cmd_boolean },
254 { "retryconnrefused", &opt.retry_connrefused, cmd_boolean },
255 { "robots", &opt.use_robots, cmd_boolean },
256 { "savecookies", &opt.cookies_output, cmd_file },
257 { "saveheaders", &opt.save_headers, cmd_boolean },
259 { "secureprotocol", &opt.secure_protocol, cmd_spec_secure_protocol },
261 { "serverresponse", &opt.server_response, cmd_boolean },
262 { "showalldnsentries", &opt.show_all_dns_entries, cmd_boolean },
263 { "spanhosts", &opt.spanhost, cmd_boolean },
264 { "spider", &opt.spider, cmd_boolean },
265 { "strictcomments", &opt.strict_comments, cmd_boolean },
266 { "timeout", NULL, cmd_spec_timeout },
267 { "timestamping", &opt.timestamping, cmd_boolean },
268 { "tries", &opt.ntry, cmd_number_inf },
269 { "trustservernames", &opt.trustservernames, cmd_boolean },
270 { "unlink", &opt.unlink, cmd_boolean },
271 { "useproxy", &opt.use_proxy, cmd_boolean },
272 { "user", &opt.user, cmd_string },
273 { "useragent", NULL, cmd_spec_useragent },
274 { "useservertimestamps", &opt.useservertimestamps, cmd_boolean },
275 { "verbose", NULL, cmd_spec_verbose },
276 { "wait", &opt.wait, cmd_time },
277 { "waitretry", &opt.waitretry, cmd_time },
278 { "warccdx", &opt.warc_cdx_enabled, cmd_boolean },
279 { "warccdxdedup", &opt.warc_cdx_dedup_filename, cmd_file },
281 { "warccompression", &opt.warc_compression_enabled, cmd_boolean },
283 { "warcdigests", &opt.warc_digests_enabled, cmd_boolean },
284 { "warcfile", &opt.warc_filename, cmd_file },
285 { "warcheader", NULL, cmd_spec_warc_header },
286 { "warckeeplog", &opt.warc_keep_log, cmd_boolean },
287 { "warcmaxsize", &opt.warc_maxsize, cmd_bytes },
288 { "warctempdir", &opt.warc_tempdir, cmd_directory },
290 { "wdebug", &opt.wdebug, cmd_boolean },
294 /* Look up CMDNAME in the commands[] and return its position in the
295 array. If CMDNAME is not found, return -1. */
298 command_by_name (const char *cmdname)
300 /* Use binary search for speed. Wget has ~100 commands, which
301 guarantees a worst case performance of 7 string comparisons. */
302 int lo = 0, hi = countof (commands) - 1;
306 int mid = (lo + hi) >> 1;
307 int cmp = strcasecmp (cmdname, commands[mid].name);
318 /* Reset the variables to default values. */
324 /* Most of the default values are 0 (and 0.0, NULL, and false).
325 Just reset everything, and fill in the non-zero values. Note
326 that initializing pointers to NULL this way is technically
327 illegal, but porting Wget to a machine where NULL is not all-zero
328 bit pattern will be the least of the implementors' worries. */
335 opt.add_hostdir = true;
339 opt.http_keep_alive = true;
340 opt.use_proxy = true;
341 tmp = getenv ("no_proxy");
343 opt.no_proxy = sepstring (tmp);
344 opt.prefer_family = prefer_none;
345 opt.allow_cache = true;
347 opt.read_timeout = 900;
348 opt.use_robots = true;
350 opt.remove_listing = true;
352 opt.dot_bytes = 1024;
353 opt.dot_spacing = 10;
354 opt.dots_in_line = 50;
356 opt.dns_cache = true;
360 opt.check_cert = true;
363 /* The default for file name restriction defaults to the OS type. */
364 #if defined(WINDOWS) || defined(MSDOS) || defined(__CYGWIN__)
365 opt.restrict_files_os = restrict_windows;
367 opt.restrict_files_os = restrict_unix;
369 opt.restrict_files_ctrl = true;
370 opt.restrict_files_nonascii = false;
371 opt.restrict_files_case = restrict_no_case_restriction;
373 opt.regex_type = regex_type_posix;
375 opt.max_redirect = 20;
380 opt.enable_iri = true;
382 opt.enable_iri = false;
385 opt.encoding_remote = NULL;
387 opt.useservertimestamps = true;
388 opt.show_all_dns_entries = false;
390 opt.warc_maxsize = 0; /* 1024 * 1024 * 1024; */
392 opt.warc_compression_enabled = true;
394 opt.warc_compression_enabled = false;
396 opt.warc_digests_enabled = true;
397 opt.warc_cdx_enabled = false;
398 opt.warc_cdx_dedup_filename = NULL;
399 opt.warc_tempdir = NULL;
400 opt.warc_keep_log = true;
403 /* Return the user's home directory (strdup-ed), or NULL if none is
408 static char *buf = NULL;
409 static char *home, *ret;
413 home = getenv ("HOME");
419 /* Under MSDOS, if $HOME isn't defined, use the directory where
420 `wget.exe' resides. */
421 const char *_w32_get_argv0 (void); /* in libwatt.a/pcconfig.c */
424 buff = _w32_get_argv0 ();
426 p = strrchr (buf, '/'); /* djgpp */
428 p = strrchr (buf, '\\'); /* others */
432 buff = malloc (len + 1);
436 strncpy (buff, _w32_get_argv0 (), len);
440 #elif !defined(WINDOWS)
441 /* If HOME is not defined, try getting it from the password
443 struct passwd *pwd = getpwuid (getuid ());
444 if (!pwd || !pwd->pw_dir)
448 /* Under Windows, if $HOME isn't defined, use the directory where
449 `wget.exe' resides. */
455 ret = home ? xstrdup (home) : NULL;
462 /* Check the 'WGETRC' environment variable and return the file name
463 if 'WGETRC' is set and is a valid file.
464 If the `WGETRC' variable exists but the file does not exist, the
465 function will exit(). */
467 wgetrc_env_file_name (void)
469 char *env = getenv ("WGETRC");
472 if (!file_exists_p (env))
474 fprintf (stderr, _("%s: WGETRC points to %s, which doesn't exist.\n"),
478 return xstrdup (env);
483 /* Check for the existance of '$HOME/.wgetrc' and return its path
484 if it exists and is set. */
486 wgetrc_user_file_name (void)
490 /* If that failed, try $HOME/.wgetrc (or equivalent). */
493 file = "SYS$LOGIN:.wgetrc";
494 #else /* def __VMS */
497 file = aprintf ("%s/.wgetrc", home);
499 #endif /* def __VMS [else] */
503 if (!file_exists_p (file))
511 /* Return the path to the user's .wgetrc. This is either the value of
512 `WGETRC' environment variable, or `$HOME/.wgetrc'.
514 Additionally, for windows, look in the directory where wget.exe
517 wgetrc_file_name (void)
519 char *file = wgetrc_env_file_name ();
523 file = wgetrc_user_file_name ();
526 /* Under Windows, if we still haven't found .wgetrc, look for the file
527 `wget.ini' in the directory where `wget.exe' resides; we do this for
528 backward compatibility with previous versions of Wget.
529 SYSTEM_WGETRC should not be defined under WINDOWS. */
532 char *home = home_dir ();
538 file = aprintf ("%s/wget.ini", home);
539 if (!file_exists_p (file))
552 /* Return values of parse_line. */
560 static enum parse_line parse_line (const char *, char **, char **, int *);
561 static bool setval_internal (int, const char *, const char *);
562 static bool setval_internal_tilde (int, const char *, const char *);
564 /* Initialize variables from a wgetrc file. Returns zero (failure) if
565 there were errors in the file. */
568 run_wgetrc (const char *file)
575 fp = fopen (file, "r");
578 fprintf (stderr, _("%s: Cannot read %s (%s).\n"), exec_name,
579 file, strerror (errno));
580 return true; /* not a fatal error */
583 while ((line = read_whole_line (fp)) != NULL)
585 char *com = NULL, *val = NULL;
588 /* Parse the line. */
589 switch (parse_line (line, &com, &val, &comind))
592 /* If everything is OK, set the value. */
593 if (!setval_internal_tilde (comind, com, val))
595 fprintf (stderr, _("%s: Error in %s at line %d.\n"),
596 exec_name, file, ln);
600 case line_syntax_error:
601 fprintf (stderr, _("%s: Syntax error in %s at line %d.\n"),
602 exec_name, file, ln);
605 case line_unknown_command:
606 fprintf (stderr, _("%s: Unknown command %s in %s at line %d.\n"),
607 exec_name, quote (com), file, ln);
625 /* Initialize the defaults and run the system wgetrc and user's own
630 char *file, *env_sysrc;
633 /* Run a non-standard system rc file when the according environment
634 variable has been set. For internal testing purposes only! */
635 env_sysrc = getenv ("SYSTEM_WGETRC");
636 if (env_sysrc && file_exists_p (env_sysrc))
638 ok &= run_wgetrc (env_sysrc);
639 /* If there are any problems parsing the system wgetrc file, tell
643 fprintf (stderr, _("\
644 Parsing system wgetrc file (env SYSTEM_WGETRC) failed. Please check\n\
646 or specify a different file using --config.\n"), env_sysrc);
650 /* Otherwise, if SYSTEM_WGETRC is defined, use it. */
652 else if (file_exists_p (SYSTEM_WGETRC))
653 ok &= run_wgetrc (SYSTEM_WGETRC);
654 /* If there are any problems parsing the system wgetrc file, tell
658 fprintf (stderr, _("\
659 Parsing system wgetrc file failed. Please check\n\
661 or specify a different file using --config.\n"), SYSTEM_WGETRC);
665 /* Override it with your own, if one exists. */
666 file = wgetrc_file_name ();
669 /* #### We should canonicalize `file' and SYSTEM_WGETRC with
670 something like realpath() before comparing them with `strcmp' */
672 if (!strcmp (file, SYSTEM_WGETRC))
674 fprintf (stderr, _("\
675 %s: Warning: Both system and user wgetrc point to %s.\n"),
676 exec_name, quote (file));
680 ok &= run_wgetrc (file);
682 /* If there were errors processing either `.wgetrc', abort. */
690 /* Remove dashes and underscores from S, modifying S in the
696 char *t = s; /* t - tortoise */
697 char *h = s; /* h - hare */
699 if (*h == '_' || *h == '-')
706 /* Parse the line pointed by line, with the syntax:
707 <sp>* command <sp>* = <sp>* value <sp>*
708 Uses malloc to allocate space for command and value.
710 Returns one of line_ok, line_empty, line_syntax_error, or
711 line_unknown_command.
713 In case of line_ok, *COM and *VAL point to freshly allocated
714 strings, and *COMIND points to com's index. In case of error or
715 empty line, their values are unmodified. */
717 static enum parse_line
718 parse_line (const char *line, char **com, char **val, int *comind)
721 const char *end = line + strlen (line);
722 const char *cmdstart, *cmdend;
723 const char *valstart, *valend;
728 /* Skip leading and trailing whitespace. */
729 while (*line && c_isspace (*line))
731 while (end > line && c_isspace (end[-1]))
734 /* Skip empty lines and comments. */
735 if (!*line || *line == '#')
741 while (p < end && (c_isalnum (*p) || *p == '_' || *p == '-'))
745 /* Skip '=', as well as any space before or after it. */
746 while (p < end && c_isspace (*p))
748 if (p == end || *p != '=')
749 return line_syntax_error;
751 while (p < end && c_isspace (*p))
757 /* The syntax is valid (even though the command might not be). Fill
758 in the command name and value. */
759 *com = strdupdelim (cmdstart, cmdend);
760 *val = strdupdelim (valstart, valend);
762 /* The line now known to be syntactically correct. Check whether
763 the command is valid. */
764 BOUNDED_TO_ALLOCA (cmdstart, cmdend, cmdcopy);
766 ind = command_by_name (cmdcopy);
768 return line_unknown_command;
770 /* Report success to the caller. */
775 #if defined(WINDOWS) || defined(MSDOS)
776 # define ISSEP(c) ((c) == '/' || (c) == '\\')
778 # define ISSEP(c) ((c) == '/')
781 /* Run commands[comind].action. */
784 setval_internal (int comind, const char *com, const char *val)
786 assert (0 <= comind && ((size_t) comind) < countof (commands));
787 DEBUGP (("Setting %s (%s) to %s\n", com, commands[comind].name, val));
788 return commands[comind].action (com, val, commands[comind].place);
792 setval_internal_tilde (int comind, const char *com, const char *val)
798 ret = setval_internal (comind, com, val);
800 /* We make tilde expansion for cmd_file and cmd_directory */
801 if (((commands[comind].action == cmd_file) ||
802 (commands[comind].action == cmd_directory))
803 && ret && (*val == '~' && ISSEP (val[1])))
805 pstring = commands[comind].place;
809 homelen = strlen (home);
810 while (homelen && ISSEP (home[homelen - 1]))
811 home[--homelen] = '\0';
813 /* Skip the leading "~/". */
814 for (++val; ISSEP (*val); val++)
816 *pstring = concat_strings (home, "/", val, (char *)0);
822 /* Run command COM with value VAL. If running the command produces an
823 error, report the error and exit.
825 This is intended to be called from main() to modify Wget's behavior
826 through command-line switches. Since COM is hard-coded in main(),
827 it is not canonicalized, and this aborts when COM is not found.
829 If COMIND's are exported to init.h, this function will be changed
830 to accept COMIND directly. */
833 setoptval (const char *com, const char *val, const char *optname)
835 /* Prepend "--" to OPTNAME. */
836 char *dd_optname = (char *) alloca (2 + strlen (optname) + 1);
839 strcpy (dd_optname + 2, optname);
841 assert (val != NULL);
842 if (!setval_internal (command_by_name (com), dd_optname, val))
846 /* Parse OPT into command and value and run it. For example,
847 run_command("foo=bar") is equivalent to setoptval("foo", "bar").
848 This is used by the `--execute' flag in main.c. */
851 run_command (const char *opt)
855 switch (parse_line (opt, &com, &val, &comind))
858 if (!setval_internal (comind, com, val))
864 fprintf (stderr, _("%s: Invalid --execute command %s\n"),
865 exec_name, quote (opt));
870 /* Generic helper functions, for use with `commands'. */
872 /* Forward declarations: */
877 static bool decode_string (const char *, const struct decode_item *, int, int *);
878 static bool simple_atoi (const char *, const char *, int *);
879 static bool simple_atof (const char *, const char *, double *);
881 #define CMP1(p, c0) (c_tolower((p)[0]) == (c0) && (p)[1] == '\0')
883 #define CMP2(p, c0, c1) (c_tolower((p)[0]) == (c0) \
884 && c_tolower((p)[1]) == (c1) \
887 #define CMP3(p, c0, c1, c2) (c_tolower((p)[0]) == (c0) \
888 && c_tolower((p)[1]) == (c1) \
889 && c_tolower((p)[2]) == (c2) \
893 /* Store the boolean value from VAL to PLACE. COM is ignored,
894 except for error messages. */
896 cmd_boolean (const char *com, const char *val, void *place)
900 if (CMP2 (val, 'o', 'n') || CMP3 (val, 'y', 'e', 's') || CMP1 (val, '1'))
901 /* "on", "yes" and "1" mean true. */
903 else if (CMP3 (val, 'o', 'f', 'f') || CMP2 (val, 'n', 'o') || CMP1 (val, '0'))
904 /* "off", "no" and "0" mean false. */
909 _("%s: %s: Invalid boolean %s; use `on' or `off'.\n"),
910 exec_name, com, quote (val));
914 *(bool *) place = value;
918 /* Set the non-negative integer value from VAL to PLACE. With
919 incorrect specification, the number remains unchanged. */
921 cmd_number (const char *com, const char *val, void *place)
923 if (!simple_atoi (val, val + strlen (val), place)
924 || *(int *) place < 0)
926 fprintf (stderr, _("%s: %s: Invalid number %s.\n"),
927 exec_name, com, quote (val));
933 /* Similar to cmd_number(), only accepts `inf' as a synonym for 0. */
935 cmd_number_inf (const char *com, const char *val, void *place)
937 if (!strcasecmp (val, "inf"))
942 return cmd_number (com, val, place);
945 /* Copy (strdup) the string at COM to a new location and place a
946 pointer to *PLACE. */
948 cmd_string (const char *com, const char *val, void *place)
950 char **pstring = (char **)place;
952 xfree_null (*pstring);
953 *pstring = xstrdup (val);
958 /* Like the above, but handles tilde-expansion when reading a user's
959 `.wgetrc'. In that case, and if VAL begins with `~', the tilde
960 gets expanded to the user's home directory. */
962 cmd_file (const char *com, const char *val, void *place)
964 char **pstring = (char **)place;
966 xfree_null (*pstring);
968 /* #### If VAL is empty, perhaps should set *PLACE to NULL. */
970 *pstring = xstrdup (val);
972 #if defined(WINDOWS) || defined(MSDOS)
973 /* Convert "\" to "/". */
976 for (s = *pstring; *s; s++)
984 /* Like cmd_file, but strips trailing '/' characters. */
986 cmd_directory (const char *com, const char *val, void *place)
990 /* Call cmd_file() for tilde expansion and separator
991 canonicalization (backslash -> slash under Windows). These
992 things should perhaps be in a separate function. */
993 if (!cmd_file (com, val, place))
998 while (t > s && *--t == '/')
1004 /* Split VAL by space to a vector of values, and append those values
1005 to vector pointed to by the PLACE argument. If VAL is empty, the
1006 PLACE vector is cleared instead. */
1009 cmd_vector (const char *com, const char *val, void *place)
1011 char ***pvec = (char ***)place;
1014 *pvec = merge_vecs (*pvec, sepstring (val));
1024 cmd_directory_vector (const char *com, const char *val, void *place)
1026 char ***pvec = (char ***)place;
1030 /* Strip the trailing slashes from directories. */
1033 seps = sepstring (val);
1034 for (t = seps; t && *t; t++)
1036 int len = strlen (*t);
1037 /* Skip degenerate case of root directory. */
1040 if ((*t)[len - 1] == '/')
1041 (*t)[len - 1] = '\0';
1044 *pvec = merge_vecs (*pvec, seps);
1054 /* Engine for cmd_bytes and cmd_bytes_sum: converts a string such as
1055 "100k" or "2.5G" to a floating point number. */
1058 parse_bytes_helper (const char *val, double *result)
1060 double number, mult;
1061 const char *end = val + strlen (val);
1063 /* Check for "inf". */
1064 if (0 == strcmp (val, "inf"))
1070 /* Strip trailing whitespace. */
1071 while (val < end && c_isspace (end[-1]))
1076 switch (c_tolower (end[-1]))
1079 --end, mult = 1024.0;
1082 --end, mult = 1048576.0;
1085 --end, mult = 1073741824.0;
1088 --end, mult = 1099511627776.0;
1091 /* Not a recognized suffix: assume it's a digit. (If not,
1092 simple_atof will raise an error.) */
1096 /* Skip leading and trailing whitespace. */
1097 while (val < end && c_isspace (*val))
1099 while (val < end && c_isspace (end[-1]))
1104 if (!simple_atof (val, end, &number) || number < 0)
1107 *result = number * mult;
1111 /* Parse VAL as a number and set its value to PLACE (which should
1114 By default, the value is assumed to be in bytes. If "K", "M", or
1115 "G" are appended, the value is multiplied with 1<<10, 1<<20, or
1116 1<<30, respectively. Floating point values are allowed and are
1117 cast to integer before use. The idea is to be able to use things
1118 like 1.5k instead of "1536".
1120 The string "inf" is returned as 0.
1122 In case of error, false is returned and memory pointed to by PLACE
1123 remains unmodified. */
1126 cmd_bytes (const char *com, const char *val, void *place)
1129 if (!parse_bytes_helper (val, &byte_value))
1131 fprintf (stderr, _("%s: %s: Invalid byte value %s\n"),
1132 exec_name, com, quote (val));
1135 *(wgint *)place = (wgint)byte_value;
1139 /* Like cmd_bytes, but PLACE is interpreted as a pointer to
1140 SIZE_SUM. It works by converting the string to double, therefore
1141 working with values up to 2^53-1 without loss of precision. This
1142 value (8192 TB) is large enough to serve for a while. */
1145 cmd_bytes_sum (const char *com, const char *val, void *place)
1148 if (!parse_bytes_helper (val, &byte_value))
1150 fprintf (stderr, _("%s: %s: Invalid byte value %s\n"),
1151 exec_name, com, quote (val));
1154 *(SUM_SIZE_INT *) place = (SUM_SIZE_INT) byte_value;
1158 /* Store the value of VAL to *OUT. The value is a time period, by
1159 default expressed in seconds, but also accepting suffixes "m", "h",
1160 "d", and "w" for minutes, hours, days, and weeks respectively. */
1163 cmd_time (const char *com, const char *val, void *place)
1165 double number, mult;
1166 const char *end = val + strlen (val);
1168 /* Strip trailing whitespace. */
1169 while (val < end && c_isspace (end[-1]))
1175 fprintf (stderr, _("%s: %s: Invalid time period %s\n"),
1176 exec_name, com, quote (val));
1180 switch (c_tolower (end[-1]))
1183 --end, mult = 1; /* seconds */
1186 --end, mult = 60; /* minutes */
1189 --end, mult = 3600; /* hours */
1192 --end, mult = 86400.0; /* days */
1195 --end, mult = 604800.0; /* weeks */
1198 /* Not a recognized suffix: assume it belongs to the number.
1199 (If not, simple_atof will raise an error.) */
1203 /* Skip leading and trailing whitespace. */
1204 while (val < end && c_isspace (*val))
1206 while (val < end && c_isspace (end[-1]))
1211 if (!simple_atof (val, end, &number))
1214 *(double *)place = number * mult;
1220 cmd_cert_type (const char *com, const char *val, void *place)
1222 static const struct decode_item choices[] = {
1223 { "pem", keyfile_pem },
1224 { "der", keyfile_asn1 },
1225 { "asn1", keyfile_asn1 },
1227 int ok = decode_string (val, choices, countof (choices), place);
1229 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1234 /* Specialized helper functions, used by `commands' to handle some
1235 options specially. */
1237 static bool check_user_specified_header (const char *);
1240 cmd_spec_dirstruct (const char *com, const char *val, void *place_ignored)
1242 if (!cmd_boolean (com, val, &opt.dirstruct))
1244 /* Since dirstruct behaviour is explicitly changed, no_dirstruct
1245 must be affected inversely. */
1247 opt.no_dirstruct = false;
1249 opt.no_dirstruct = true;
1254 cmd_spec_header (const char *com, const char *val, void *place_ignored)
1256 /* Empty value means reset the list of headers. */
1259 free_vec (opt.user_headers);
1260 opt.user_headers = NULL;
1264 if (!check_user_specified_header (val))
1266 fprintf (stderr, _("%s: %s: Invalid header %s.\n"),
1267 exec_name, com, quote (val));
1270 opt.user_headers = vec_append (opt.user_headers, val);
1275 cmd_spec_warc_header (const char *com, const char *val, void *place_ignored)
1277 /* Empty value means reset the list of headers. */
1280 free_vec (opt.warc_user_headers);
1281 opt.warc_user_headers = NULL;
1285 if (!check_user_specified_header (val))
1287 fprintf (stderr, _("%s: %s: Invalid WARC header %s.\n"),
1288 exec_name, com, quote (val));
1291 opt.warc_user_headers = vec_append (opt.warc_user_headers, val);
1296 cmd_spec_htmlify (const char *com, const char *val, void *place_ignored)
1298 int flag = cmd_boolean (com, val, &opt.htmlify);
1299 if (flag && !opt.htmlify)
1300 opt.remove_listing = false;
1304 /* Set the "mirror" mode. It means: recursive download, timestamping,
1305 no limit on max. recursion depth, and don't remove listings. */
1308 cmd_spec_mirror (const char *com, const char *val, void *place_ignored)
1312 if (!cmd_boolean (com, val, &mirror))
1316 opt.recursive = true;
1317 if (!opt.no_dirstruct)
1318 opt.dirstruct = true;
1319 opt.timestamping = true;
1320 opt.reclevel = INFINITE_RECURSION;
1321 opt.remove_listing = false;
1326 /* Validate --prefer-family and set the choice. Allowed values are
1327 "IPv4", "IPv6", and "none". */
1330 cmd_spec_prefer_family (const char *com, const char *val, void *place_ignored)
1332 static const struct decode_item choices[] = {
1333 { "IPv4", prefer_ipv4 },
1334 { "IPv6", prefer_ipv6 },
1335 { "none", prefer_none },
1337 int prefer_family = prefer_none;
1338 int ok = decode_string (val, choices, countof (choices), &prefer_family);
1340 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1341 opt.prefer_family = prefer_family;
1345 /* Set progress.type to VAL, but verify that it's a valid progress
1346 implementation before that. */
1349 cmd_spec_progress (const char *com, const char *val, void *place_ignored)
1351 if (!valid_progress_implementation_p (val))
1353 fprintf (stderr, _("%s: %s: Invalid progress type %s.\n"),
1354 exec_name, com, quote (val));
1357 xfree_null (opt.progress_type);
1359 /* Don't call set_progress_implementation here. It will be called
1360 in main() when it becomes clear what the log output is. */
1361 opt.progress_type = xstrdup (val);
1365 /* Set opt.recursive to VAL as with cmd_boolean. If opt.recursive is
1366 set to true, also set opt.dirstruct to true, unless opt.no_dirstruct
1370 cmd_spec_recursive (const char *com, const char *val, void *place_ignored)
1372 if (!cmd_boolean (com, val, &opt.recursive))
1376 if (opt.recursive && !opt.no_dirstruct)
1377 opt.dirstruct = true;
1382 /* Validate --regex-type and set the choice. */
1385 cmd_spec_regex_type (const char *com, const char *val, void *place_ignored)
1387 static const struct decode_item choices[] = {
1388 { "posix", regex_type_posix },
1390 { "pcre", regex_type_pcre },
1393 int regex_type = regex_type_posix;
1394 int ok = decode_string (val, choices, countof (choices), ®ex_type);
1396 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1397 opt.regex_type = regex_type;
1402 cmd_spec_restrict_file_names (const char *com, const char *val, void *place_ignored)
1404 int restrict_os = opt.restrict_files_os;
1405 int restrict_ctrl = opt.restrict_files_ctrl;
1406 int restrict_case = opt.restrict_files_case;
1407 int restrict_nonascii = opt.restrict_files_nonascii;
1411 #define VAL_IS(string_literal) BOUNDED_EQUAL (val, end, string_literal)
1415 end = strchr (val, ',');
1417 end = val + strlen (val);
1419 if (VAL_IS ("unix"))
1420 restrict_os = restrict_unix;
1421 else if (VAL_IS ("windows"))
1422 restrict_os = restrict_windows;
1423 else if (VAL_IS ("lowercase"))
1424 restrict_case = restrict_lowercase;
1425 else if (VAL_IS ("uppercase"))
1426 restrict_case = restrict_uppercase;
1427 else if (VAL_IS ("nocontrol"))
1428 restrict_ctrl = false;
1429 else if (VAL_IS ("ascii"))
1430 restrict_nonascii = true;
1433 fprintf (stderr, _("\
1434 %s: %s: Invalid restriction %s,\n\
1435 use [unix|windows],[lowercase|uppercase],[nocontrol],[ascii].\n"),
1436 exec_name, com, quote (val));
1443 while (*val && *end);
1447 opt.restrict_files_os = restrict_os;
1448 opt.restrict_files_ctrl = restrict_ctrl;
1449 opt.restrict_files_case = restrict_case;
1450 opt.restrict_files_nonascii = restrict_nonascii;
1456 cmd_spec_report_speed (const char *com, const char *val, void *place_ignored)
1458 opt.report_bps = strcasecmp (val, "bits") == 0;
1459 if (!opt.report_bps)
1460 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1461 return opt.report_bps;
1466 cmd_spec_secure_protocol (const char *com, const char *val, void *place)
1468 static const struct decode_item choices[] = {
1469 { "auto", secure_protocol_auto },
1470 { "sslv2", secure_protocol_sslv2 },
1471 { "sslv3", secure_protocol_sslv3 },
1472 { "tlsv1", secure_protocol_tlsv1 },
1474 int ok = decode_string (val, choices, countof (choices), place);
1476 fprintf (stderr, _("%s: %s: Invalid value %s.\n"), exec_name, com, quote (val));
1481 /* Set all three timeout values. */
1484 cmd_spec_timeout (const char *com, const char *val, void *place_ignored)
1487 if (!cmd_time (com, val, &value))
1489 opt.read_timeout = value;
1490 opt.connect_timeout = value;
1491 opt.dns_timeout = value;
1496 cmd_spec_useragent (const char *com, const char *val, void *place_ignored)
1498 /* Disallow embedded newlines. */
1499 if (strchr (val, '\n'))
1501 fprintf (stderr, _("%s: %s: Invalid value %s.\n"),
1502 exec_name, com, quote (val));
1505 xfree_null (opt.useragent);
1506 opt.useragent = xstrdup (val);
1510 /* The "verbose" option cannot be cmd_boolean because the variable is
1511 not bool -- it's of type int (-1 means uninitialized because of
1512 some random hackery for disallowing -q -v). */
1515 cmd_spec_verbose (const char *com, const char *val, void *place_ignored)
1518 if (cmd_boolean (com, val, &flag))
1526 /* Miscellaneous useful routines. */
1528 /* A very simple atoi clone, more useful than atoi because it works on
1529 delimited strings, and has error reportage. Returns true on success,
1530 false on failure. If successful, stores result to *DEST. */
1533 simple_atoi (const char *beg, const char *end, int *dest)
1536 bool negative = false;
1537 const char *p = beg;
1539 while (p < end && c_isspace (*p))
1541 if (p < end && (*p == '-' || *p == '+'))
1543 negative = (*p == '-');
1549 /* Read negative numbers in a separate loop because the most
1550 negative integer cannot be represented as a positive number. */
1553 for (; p < end && c_isdigit (*p); p++)
1555 int next = (10 * result) + (*p - '0');
1557 return false; /* overflow */
1561 for (; p < end && c_isdigit (*p); p++)
1563 int next = (10 * result) - (*p - '0');
1565 return false; /* underflow */
1576 /* Trivial atof, with error reporting. Handles "<digits>[.<digits>]",
1577 doesn't handle exponential notation. Returns true on success,
1578 false on failure. In case of success, stores its result to
1582 simple_atof (const char *beg, const char *end, double *dest)
1586 bool negative = false;
1587 bool seen_dot = false;
1588 bool seen_digit = false;
1591 const char *p = beg;
1593 while (p < end && c_isspace (*p))
1595 if (p < end && (*p == '-' || *p == '+'))
1597 negative = (*p == '-');
1601 for (; p < end; p++)
1607 result = (10 * result) + (ch - '0');
1609 result += (ch - '0') / (divider *= 10);
1631 /* Verify that the user-specified header in S is valid. It must
1632 contain a colon preceded by non-white-space characters and must not
1633 contain newlines. */
1636 check_user_specified_header (const char *s)
1640 for (p = s; *p && *p != ':' && !c_isspace (*p); p++)
1642 /* The header MUST contain `:' preceded by at least one
1643 non-whitespace character. */
1644 if (*p != ':' || p == s)
1646 /* The header MUST NOT contain newlines. */
1647 if (strchr (s, '\n'))
1652 /* Decode VAL into a number, according to ITEMS. */
1655 decode_string (const char *val, const struct decode_item *items, int itemcount,
1659 for (i = 0; i < itemcount; i++)
1660 if (0 == strcasecmp (val, items[i].name))
1662 *place = items[i].code;
1669 void cleanup_html_url (void);
1672 /* Free the memory allocated by global variables. */
1676 /* Free external resources, close files, etc. */
1679 fclose (output_stream);
1680 /* No need to check for error because Wget flushes its output (and
1681 checks for errors) after any data arrives. */
1683 /* We're exiting anyway so there's no real need to call free()
1684 hundreds of times. Skipping the frees will make Wget exit
1687 However, when detecting leaks, it's crucial to free() everything
1688 because then you can find the real leaks, i.e. the allocated
1689 memory which grows with the size of the program. */
1695 cleanup_html_url ();
1700 extern acc_t *netrc_list;
1701 free_netrc (netrc_list);
1703 xfree_null (opt.choose_config);
1704 xfree_null (opt.lfilename);
1705 xfree_null (opt.dir_prefix);
1706 xfree_null (opt.input_filename);
1707 xfree_null (opt.output_document);
1708 free_vec (opt.accepts);
1709 free_vec (opt.rejects);
1710 free_vec (opt.excludes);
1711 free_vec (opt.includes);
1712 free_vec (opt.domains);
1713 free_vec (opt.follow_tags);
1714 free_vec (opt.ignore_tags);
1715 xfree_null (opt.progress_type);
1716 xfree_null (opt.ftp_user);
1717 xfree_null (opt.ftp_passwd);
1718 xfree_null (opt.ftp_proxy);
1719 xfree_null (opt.https_proxy);
1720 xfree_null (opt.http_proxy);
1721 free_vec (opt.no_proxy);
1722 xfree_null (opt.useragent);
1723 xfree_null (opt.referer);
1724 xfree_null (opt.http_user);
1725 xfree_null (opt.http_passwd);
1726 free_vec (opt.user_headers);
1727 free_vec (opt.warc_user_headers);
1729 xfree_null (opt.cert_file);
1730 xfree_null (opt.private_key);
1731 xfree_null (opt.ca_directory);
1732 xfree_null (opt.ca_cert);
1733 xfree_null (opt.random_file);
1734 xfree_null (opt.egd_file);
1736 xfree_null (opt.bind_address);
1737 xfree_null (opt.cookies_input);
1738 xfree_null (opt.cookies_output);
1739 xfree_null (opt.user);
1740 xfree_null (opt.passwd);
1741 xfree_null (opt.base_href);
1743 #endif /* DEBUG_MALLOC */
1746 /* Unit testing routines. */
1751 test_commands_sorted()
1753 int prev_idx = 0, next_idx = 1;
1754 int command_count = countof (commands) - 1;
1756 while (next_idx <= command_count)
1758 cmp = strcasecmp (commands[prev_idx].name, commands[next_idx].name);
1761 mu_assert ("FAILED", false);
1774 test_cmd_spec_restrict_file_names()
1779 int expected_restrict_files_os;
1780 int expected_restrict_files_ctrl;
1781 int expected_restrict_files_case;
1784 { "windows", restrict_windows, true, restrict_no_case_restriction, true },
1785 { "windows,", restrict_windows, true, restrict_no_case_restriction, true },
1786 { "windows,lowercase", restrict_windows, true, restrict_lowercase, true },
1787 { "unix,nocontrol,lowercase,", restrict_unix, false, restrict_lowercase, true },
1790 for (i = 0; i < sizeof(test_array)/sizeof(test_array[0]); ++i)
1795 res = cmd_spec_restrict_file_names ("dummy", test_array[i].val, NULL);
1798 fprintf (stderr, "test_cmd_spec_restrict_file_names: TEST %d\n", i); fflush (stderr);
1799 fprintf (stderr, "opt.restrict_files_os: %d\n", opt.restrict_files_os); fflush (stderr);
1800 fprintf (stderr, "opt.restrict_files_ctrl: %d\n", opt.restrict_files_ctrl); fflush (stderr);
1801 fprintf (stderr, "opt.restrict_files_case: %d\n", opt.restrict_files_case); fflush (stderr);
1803 mu_assert ("test_cmd_spec_restrict_file_names: wrong result",
1804 res == test_array[i].result
1805 && opt.restrict_files_os == test_array[i].expected_restrict_files_os
1806 && opt.restrict_files_ctrl == test_array[i].expected_restrict_files_ctrl
1807 && opt.restrict_files_case == test_array[i].expected_restrict_files_case);
1813 #endif /* TESTING */