1 /* Command line parsing.
2 Copyright (C) 1995, 1996, 1997, 1998 Free Software Foundation, Inc.
4 This file is part of Wget.
6 This program is free software; you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation; either version 2 of the License, or
9 (at your option) any later version.
11 This program is distributed in the hope that it will be useful,
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 GNU General Public License for more details.
16 You should have received a copy of the GNU General Public License
17 along with this program; if not, write to the Free Software
18 Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */
27 #endif /* HAVE_UNISTD_H */
28 #include <sys/types.h>
33 #endif /* HAVE_STRING_H */
40 #endif /* HAVE_LOCALE_H */
43 #define OPTIONS_DEFINED_HERE /* for options.h */
53 #ifndef PATH_SEPARATOR
54 # define PATH_SEPARATOR '/'
57 extern char *version_string;
66 void log_init PARAMS ((const char *, int));
67 void log_close PARAMS ((void));
68 void redirect_output PARAMS ((const char *));
70 static RETSIGTYPE redirect_output_signal PARAMS ((int));
72 const char *exec_name;
74 /* Initialize I18N. The initialization amounts to invoking
75 setlocale(), bindtextdomain() and textdomain().
76 Does nothing if NLS is disabled or missing. */
78 i18n_initialize (void)
80 /* If HAVE_NLS is defined, assume the existence of the three
81 functions invoked here. */
83 /* Set the current locale. */
84 /* Here we use LC_MESSAGES instead of LC_ALL, for two reasons.
85 First, message catalogs are all of I18N Wget uses anyway.
86 Second, setting LC_ALL has a dangerous potential of messing
87 things up. For example, when in a foreign locale, Solaris
88 strptime() fails to handle international dates correctly, which
89 makes http_atotm() malfunction. */
90 setlocale (LC_MESSAGES, "");
91 /* Set the text message domain. */
92 bindtextdomain ("wget", LOCALEDIR);
97 /* Print the usage message. */
101 printf (_("Usage: %s [OPTION]... [URL]...\n"), exec_name);
104 /* Print the help message, describing all the available options. If
105 you add an option, be sure to update this list. */
109 printf (_("GNU Wget %s, a non-interactive network retriever.\n"),
112 /* Had to split this in parts, so the #@@#%# Ultrix compiler and cpp
113 don't bitch. Also, it makes translation much easier. */
114 printf ("%s%s%s%s%s%s%s%s%s%s", _("\
116 Mandatory arguments to long options are mandatory for short options too.\n\
119 -V, --version display the version of Wget and exit.\n\
120 -h, --help print this help.\n\
121 -b, --background go to background after startup.\n\
122 -e, --execute=COMMAND execute a `.wgetrc\' command.\n\
124 Logging and input file:\n\
125 -o, --output-file=FILE log messages to FILE.\n\
126 -a, --append-output=FILE append messages to FILE.\n\
127 -d, --debug print debug output.\n\
128 -q, --quiet quiet (no output).\n\
129 -v, --verbose be verbose (this is the default).\n\
130 -nv, --non-verbose turn off verboseness, without being quiet.\n\
131 -i, --input-file=FILE read URL-s from file.\n\
132 -F, --force-html treat input file as HTML.\n\
135 -t, --tries=NUMBER set number of retries to NUMBER (0 unlimits).\n\
136 -O --output-document=FILE write documents to FILE.\n\
137 -nc, --no-clobber don\'t clobber existing files.\n\
138 -c, --continue restart getting an existing file.\n\
139 --dot-style=STYLE set retrieval display style.\n\
140 -N, --timestamping don\'t retrieve files if older than local.\n\
141 -S, --server-response print server response.\n\
142 --spider don\'t download anything.\n\
143 -T, --timeout=SECONDS set the read timeout to SECONDS.\n\
144 -w, --wait=SECONDS wait SECONDS between retrievals.\n\
145 -Y, --proxy=on/off turn proxy on or off.\n\
146 -Q, --quota=NUMBER set retrieval quota to NUMBER.\n\
149 -nd --no-directories don\'t create directories.\n\
150 -x, --force-directories force creation of directories.\n\
151 -nH, --no-host-directories don\'t create host directories.\n\
152 -P, --directory-prefix=PREFIX save files to PREFIX/...\n\
153 --cut-dirs=NUMBER ignore NUMBER remote directory components.\n\
156 --http-user=USER set http user to USER.\n\
157 --http-passwd=PASS set http password to PASS.\n\
158 -C, --cache=on/off (dis)allow server-cached data (normally allowed).\n\
159 --ignore-length ignore `Content-Length\' header field.\n\
160 --header=STRING insert STRING among the headers.\n\
161 --proxy-user=USER set USER as proxy username.\n\
162 --proxy-passwd=PASS set PASS as proxy password.\n\
163 -s, --save-headers save the HTTP headers to file.\n\
164 -U, --user-agent=AGENT identify as AGENT instead of Wget/VERSION.\n\
167 --retr-symlinks retrieve FTP symbolic links.\n\
168 -g, --glob=on/off turn file name globbing on or off.\n\
169 --passive-ftp use the \"passive\" transfer mode.\n\
171 Recursive retrieval:\n\
172 -r, --recursive recursive web-suck -- use with care!.\n\
173 -l, --level=NUMBER maximum recursion depth (0 to unlimit).\n\
174 --delete-after delete downloaded files.\n\
175 -k, --convert-links convert non-relative links to relative.\n\
176 -K, --backup-converted before converting file X, back up as X.orig.\n\
177 -m, --mirror turn on options suitable for mirroring.\n\
178 -nr, --dont-remove-listing don\'t remove `.listing\' files.\n\
180 Recursive accept/reject:\n\
181 -A, --accept=LIST list of accepted extensions.\n\
182 -R, --reject=LIST list of rejected extensions.\n\
183 -D, --domains=LIST list of accepted domains.\n\
184 --exclude-domains=LIST comma-separated list of rejected domains.\n\
185 -L, --relative follow relative links only.\n\
186 --follow-ftp follow FTP links from HTML documents.\n\
187 -H, --span-hosts go to foreign hosts when recursive.\n\
188 -I, --include-directories=LIST list of allowed directories.\n\
189 -X, --exclude-directories=LIST list of excluded directories.\n\
190 -nh, --no-host-lookup don\'t DNS-lookup hosts.\n\
191 -np, --no-parent don\'t ascend to the parent directory.\n\
192 \n"), _("Mail bug reports and suggestions to <bug-wget@gnu.org>.\n"));
196 main (int argc, char *const *argv)
199 int i, c, nurl, status, append_to_log;
201 static struct option long_options[] =
203 { "background", no_argument, NULL, 'b' },
204 { "continue", no_argument, NULL, 'c' },
205 { "convert-links", no_argument, NULL, 'k' },
206 { "backup-converted", no_argument, NULL, 'K' },
207 { "debug", no_argument, NULL, 'd' },
208 { "dont-remove-listing", no_argument, NULL, 21 },
209 { "email-address", no_argument, NULL, 'E' }, /* undocumented (debug) */
210 { "follow-ftp", no_argument, NULL, 14 },
211 { "force-directories", no_argument, NULL, 'x' },
212 { "force-hier", no_argument, NULL, 'x' }, /* obsolete */
213 { "force-html", no_argument, NULL, 'F'},
214 { "help", no_argument, NULL, 'h' },
215 { "ignore-length", no_argument, NULL, 10 },
216 { "mirror", no_argument, NULL, 'm' },
217 { "no-clobber", no_argument, NULL, 13 },
218 { "no-directories", no_argument, NULL, 19 },
219 { "no-host-directories", no_argument, NULL, 20 },
220 { "no-host-lookup", no_argument, NULL, 22 },
221 { "no-parent", no_argument, NULL, 5 },
222 { "non-verbose", no_argument, NULL, 18 },
223 { "passive-ftp", no_argument, NULL, 11 },
224 { "quiet", no_argument, NULL, 'q' },
225 { "recursive", no_argument, NULL, 'r' },
226 { "relative", no_argument, NULL, 'L' },
227 { "retr-symlinks", no_argument, NULL, 9 },
228 { "save-headers", no_argument, NULL, 's' },
229 { "server-response", no_argument, NULL, 'S' },
230 { "span-hosts", no_argument, NULL, 'H' },
231 { "spider", no_argument, NULL, 4 },
232 { "timestamping", no_argument, NULL, 'N' },
233 { "verbose", no_argument, NULL, 'v' },
234 { "version", no_argument, NULL, 'V' },
236 { "accept", required_argument, NULL, 'A' },
237 { "append-output", required_argument, NULL, 'a' },
238 { "backups", required_argument, NULL, 23 }, /* undocumented */
239 { "base", required_argument, NULL, 'B' },
240 { "cache", required_argument, NULL, 'C' },
241 { "cut-dirs", required_argument, NULL, 17 },
242 { "delete-after", no_argument, NULL, 8 },
243 { "directory-prefix", required_argument, NULL, 'P' },
244 { "domains", required_argument, NULL, 'D' },
245 { "dot-style", required_argument, NULL, 6 },
246 { "execute", required_argument, NULL, 'e' },
247 { "exclude-directories", required_argument, NULL, 'X' },
248 { "exclude-domains", required_argument, NULL, 12 },
249 { "glob", required_argument, NULL, 'g' },
250 { "header", required_argument, NULL, 3 },
251 { "htmlify", required_argument, NULL, 7 },
252 { "http-passwd", required_argument, NULL, 2 },
253 { "http-user", required_argument, NULL, 1 },
254 { "include-directories", required_argument, NULL, 'I' },
255 { "input-file", required_argument, NULL, 'i' },
256 { "level", required_argument, NULL, 'l' },
257 { "no", required_argument, NULL, 'n' },
258 { "output-document", required_argument, NULL, 'O' },
259 { "output-file", required_argument, NULL, 'o' },
260 { "proxy", required_argument, NULL, 'Y' },
261 { "proxy-passwd", required_argument, NULL, 16 },
262 { "proxy-user", required_argument, NULL, 15 },
263 { "quota", required_argument, NULL, 'Q' },
264 { "reject", required_argument, NULL, 'R' },
265 { "timeout", required_argument, NULL, 'T' },
266 { "tries", required_argument, NULL, 't' },
267 { "user-agent", required_argument, NULL, 'U' },
268 { "referer", required_argument, NULL, 129 },
269 { "use-proxy", required_argument, NULL, 'Y' },
270 { "wait", required_argument, NULL, 'w' },
278 /* Construct the name of the executable, without the directory part. */
279 exec_name = strrchr (argv[0], PATH_SEPARATOR);
286 windows_main_junk (&argc, (char **) argv, (char **) &exec_name);
291 while ((c = getopt_long (argc, argv, "\
292 hVqvdkKsxmNWrHSLcFbEY:g:T:U:O:l:n:i:o:a:t:D:A:R:P:B:e:Q:X:I:w:",
293 long_options, (int *)0)) != EOF)
297 /* Options without arguments: */
299 setval ("spider", "on");
302 setval ("noparent", "on");
305 setval ("deleteafter", "on");
308 setval ("retrsymlinks", "on");
311 setval ("ignorelength", "on");
314 setval ("passiveftp", "on");
317 setval ("noclobber", "on");
320 setval ("followftp", "on");
323 setval ("cutdirs", optarg);
326 setval ("verbose", "off");
329 setval ("dirstruct", "off");
332 setval ("addhostdir", "off");
335 setval ("removelisting", "off");
338 setval ("simplehostcheck", "on");
341 setval ("background", "on");
344 setval ("continue", "on");
348 setval ("debug", "on");
349 #else /* not DEBUG */
350 fprintf (stderr, _("%s: debug support not compiled in.\n"),
352 #endif /* not DEBUG */
355 /* For debugging purposes. */
356 printf ("%s\n", ftp_getaddress ());
360 setval ("forcehtml", "on");
363 setval ("spanhosts", "on");
373 setval ("convertlinks", "on");
376 setval ("backupconverted", "on");
379 setval ("relativeonly", "on");
382 setval ("mirror", "on");
385 setval ("timestamping", "on");
388 setval ("serverresponse", "on");
391 setval ("saveheaders", "on");
394 setval ("quiet", "on");
397 setval ("recursive", "on");
400 printf ("GNU Wget %s\n\n", version_string);
402 Copyright (C) 1995, 1996, 1997, 1998 Free Software Foundation, Inc.\n\
403 This program is distributed in the hope that it will be useful,\n\
404 but WITHOUT ANY WARRANTY; without even the implied warranty of\n\
405 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the\n\
406 GNU General Public License for more details.\n"));
407 printf (_("\nWritten by Hrvoje Niksic <hniksic@srce.hr>.\n"));
411 setval ("verbose", "on");
414 setval ("dirstruct", "on");
417 /* Options accepting an argument: */
419 setval ("httpuser", optarg);
422 setval ("httppasswd", optarg);
425 setval ("header", optarg);
428 setval ("dotstyle", optarg);
431 setval ("htmlify", optarg);
434 setval ("excludedomains", optarg);
437 setval ("proxyuser", optarg);
440 setval ("proxypasswd", optarg);
443 setval ("backups", optarg);
446 setval ("accept", optarg);
449 setval ("logfile", optarg);
453 setval ("base", optarg);
456 setval ("cache", optarg);
459 setval ("domains", optarg);
464 if (parse_line (optarg, &com, &val))
466 if (!setval (com, val))
471 fprintf (stderr, _("%s: %s: invalid command\n"), exec_name,
480 setval ("glob", optarg);
483 setval ("includedirectories", optarg);
486 setval ("input", optarg);
489 setval ("reclevel", optarg);
493 /* #### The n? options are utter crock! */
496 for (p = optarg; *p; p++)
500 setval ("verbose", "off");
503 setval ("simplehostcheck", "on");
506 setval ("addhostdir", "off");
509 setval ("dirstruct", "off");
512 setval ("noclobber", "on");
515 setval ("removelisting", "off");
518 setval ("noparent", "on");
521 printf (_("%s: illegal option -- `-n%c'\n"), exec_name, *p);
524 printf (_("Try `%s --help\' for more options.\n"), exec_name);
530 setval ("outputdocument", optarg);
533 setval ("logfile", optarg);
536 setval ("dirprefix", optarg);
539 setval ("quota", optarg);
542 setval ("reject", optarg);
545 setval ("timeout", optarg);
548 setval ("tries", optarg);
551 setval ("useragent", optarg);
554 setval ("referer", optarg);
557 setval ("wait", optarg);
560 setval ("excludedirectories", optarg);
563 setval ("useproxy", optarg);
569 printf (_("Try `%s --help' for more options.\n"), exec_name);
574 if (opt.verbose == -1)
575 opt.verbose = !opt.quiet;
578 if (opt.verbose && opt.quiet)
580 printf (_("Can't be verbose and quiet at the same time.\n"));
584 if (opt.timestamping && opt.noclobber)
587 Can't timestamp and not clobber old files at the same time.\n"));
591 nurl = argc - optind;
592 if (!nurl && !opt.input_filename)
594 /* No URL specified. */
595 printf (_("%s: missing URL\n"), exec_name);
598 /* #### Something nicer should be printed here -- similar to the
599 pre-1.5 `--help' page. */
600 printf (_("Try `%s --help' for more options.\n"), exec_name);
605 fork_to_background ();
607 /* Allocate basic pointer. */
608 url = ALLOCA_ARRAY (char *, nurl + 1);
609 /* Fill in the arguments. */
610 for (i = 0; i < nurl; i++, optind++)
612 char *irix4_cc_needs_this;
613 STRDUP_ALLOCA (irix4_cc_needs_this, argv[optind]);
614 url[i] = irix4_cc_needs_this;
618 /* Change the title of console window on Windows. #### I think this
619 statement should belong to retrieve_url(). --hniksic. */
621 ws_changetitle (*url, nurl);
624 /* Initialize logging. */
625 log_init (opt.lfilename, append_to_log);
627 DEBUGP (("DEBUG output created by Wget %s on %s.\n\n", version_string,
629 /* Open the output filename if necessary. */
630 if (opt.output_document)
632 if (HYPHENP (opt.output_document))
636 opt.dfp = fopen (opt.output_document, "wb");
639 perror (opt.output_document);
649 /* Setup the signal handler to redirect output when hangup is
652 if (signal(SIGHUP, SIG_IGN) != SIG_IGN)
653 signal(SIGHUP, redirect_output_signal);
654 /* ...and do the same for SIGUSR1. */
655 signal (SIGUSR1, redirect_output_signal);
656 /* Writing to a closed socket normally signals SIGPIPE, and the
657 process exits. What we want is to ignore SIGPIPE and just check
658 for the return value of write(). */
659 signal (SIGPIPE, SIG_IGN);
660 #endif /* HAVE_SIGNAL */
662 status = RETROK; /* initialize it, just-in-case */
664 /* Retrieve the URLs from argument list. */
665 for (t = url; *t; t++)
667 char *filename, *new_file;
670 status = retrieve_url (*t, &filename, &new_file, NULL, &dt);
671 if (opt.recursive && status == RETROK && (dt & TEXTHTML))
672 status = recursive_retrieve (filename, new_file ? new_file : *t);
673 FREE_MAYBE (new_file);
674 FREE_MAYBE (filename);
677 /* And then from the input file, if any. */
678 if (opt.input_filename)
681 status = retrieve_from_file (opt.input_filename, opt.force_html, &count);
683 logprintf (LOG_NOTQUIET, _("No URLs found in %s.\n"),
686 /* Print the downloaded sum. */
689 || (opt.input_filename && opt.downloaded != 0))
691 logprintf (LOG_NOTQUIET,
692 _("\nFINISHED --%s--\nDownloaded: %s bytes in %d files\n"),
693 time_str (NULL), legible (opt.downloaded), opt.numurls);
694 /* Print quota warning, if exceeded. */
695 if (opt.quota && opt.downloaded > opt.quota)
696 logprintf (LOG_NOTQUIET,
697 _("Download quota (%s bytes) EXCEEDED!\n"),
698 legible (opt.quota));
700 if (opt.convert_links)
702 convert_all_links ();
706 if (status == RETROK)
712 /* Hangup signal handler. When wget receives SIGHUP or SIGUSR1, it
713 will proceed operation as usual, trying to write into a log file.
714 If that is impossible, the output will be turned off. */
718 redirect_output_signal (int sig)
721 signal (sig, redirect_output_signal);
722 /* Please note that the double `%' in `%%s' is intentional, because
723 redirect_output passes tmp through printf. */
724 sprintf (tmp, _("%s received, redirecting output to `%%s'.\n"),
725 (sig == SIGHUP ? "SIGHUP" :
726 (sig == SIGUSR1 ? "SIGUSR1" :
728 redirect_output (tmp);
730 #endif /* HAVE_SIGNAL */