1 This is Info file wget.info, produced by Makeinfo version 1.68 from the
2 input file ./wget.texi.
4 INFO-DIR-SECTION Net Utilities
5 INFO-DIR-SECTION World Wide Web
7 * Wget: (wget). The non-interactive network downloader.
10 This file documents the the GNU Wget utility for downloading network
13 Copyright (C) 1996, 1997, 1998, 2000 Free Software Foundation, Inc.
15 Permission is granted to make and distribute verbatim copies of this
16 manual provided the copyright notice and this permission notice are
17 preserved on all copies.
19 Permission is granted to copy and distribute modified versions of
20 this manual under the conditions for verbatim copying, provided also
21 that the sections entitled "Copying" and "GNU General Public License"
22 are included exactly as in the original, and provided that the entire
23 resulting derived work is distributed under the terms of a permission
24 notice identical to this one.
27 File: wget.info, Node: Time-Stamping, Next: Startup File, Prev: Following Links, Up: Top
32 One of the most important aspects of mirroring information from the
33 Internet is updating your archives.
35 Downloading the whole archive again and again, just to replace a few
36 changed files is expensive, both in terms of wasted bandwidth and money,
37 and the time to do the update. This is why all the mirroring tools
38 offer the option of incremental updating.
40 Such an updating mechanism means that the remote server is scanned in
41 search of "new" files. Only those new files will be downloaded in the
42 place of the old ones.
44 A file is considered new if one of these two conditions are met:
46 1. A file of that name does not already exist locally.
48 2. A file of that name does exist, but the remote file was modified
49 more recently than the local file.
51 To implement this, the program needs to be aware of the time of last
52 modification of both remote and local files. Such information are
53 called the "time-stamps".
55 The time-stamping in GNU Wget is turned on using `--timestamping'
56 (`-N') option, or through `timestamping = on' directive in `.wgetrc'.
57 With this option, for each file it intends to download, Wget will check
58 whether a local file of the same name exists. If it does, and the
59 remote file is older, Wget will not download it.
61 If the local file does not exist, or the sizes of the files do not
62 match, Wget will download the remote file no matter what the time-stamps
67 * Time-Stamping Usage::
68 * HTTP Time-Stamping Internals::
69 * FTP Time-Stamping Internals::
72 File: wget.info, Node: Time-Stamping Usage, Next: HTTP Time-Stamping Internals, Prev: Time-Stamping, Up: Time-Stamping
77 The usage of time-stamping is simple. Say you would like to
78 download a file so that it keeps its date of modification.
80 wget -S http://www.gnu.ai.mit.edu/
82 A simple `ls -l' shows that the time stamp on the local file equals
83 the state of the `Last-Modified' header, as returned by the server. As
84 you can see, the time-stamping info is preserved locally, even without
87 Several days later, you would like Wget to check if the remote file
88 has changed, and download it if it has.
90 wget -N http://www.gnu.ai.mit.edu/
92 Wget will ask the server for the last-modified date. If the local
93 file is newer, the remote file will not be re-fetched. However, if the
94 remote file is more recent, Wget will proceed fetching it normally.
96 The same goes for FTP. For example:
98 wget ftp://ftp.ifi.uio.no/pub/emacs/gnus/*
100 `ls' will show that the timestamps are set according to the state on
101 the remote server. Reissuing the command with `-N' will make Wget
102 re-fetch *only* the files that have been modified.
104 In both HTTP and FTP retrieval Wget will time-stamp the local file
105 correctly (with or without `-N') if it gets the stamps, i.e. gets the
106 directory listing for FTP or the `Last-Modified' header for HTTP.
108 If you wished to mirror the GNU archive every week, you would use the
109 following command every week:
111 wget --timestamping -r ftp://prep.ai.mit.edu/pub/gnu/
114 File: wget.info, Node: HTTP Time-Stamping Internals, Next: FTP Time-Stamping Internals, Prev: Time-Stamping Usage, Up: Time-Stamping
116 HTTP Time-Stamping Internals
117 ============================
119 Time-stamping in HTTP is implemented by checking of the
120 `Last-Modified' header. If you wish to retrieve the file `foo.html'
121 through HTTP, Wget will check whether `foo.html' exists locally. If it
122 doesn't, `foo.html' will be retrieved unconditionally.
124 If the file does exist locally, Wget will first check its local
125 time-stamp (similar to the way `ls -l' checks it), and then send a
126 `HEAD' request to the remote server, demanding the information on the
129 The `Last-Modified' header is examined to find which file was
130 modified more recently (which makes it "newer"). If the remote file is
131 newer, it will be downloaded; if it is older, Wget will give up.(1)
133 When `--backup-converted' (`-K') is specified in conjunction with
134 `-N', server file `X' is compared to local file `X.orig', if extant,
135 rather than being compared to local file `X', which will always differ
136 if it's been converted by `--convert-links' (`-k').
138 Arguably, HTTP time-stamping should be implemented using the
139 `If-Modified-Since' request.
141 ---------- Footnotes ----------
143 (1) As an additional check, Wget will look at the `Content-Length'
144 header, and compare the sizes; if they are not the same, the remote
145 file will be downloaded no matter what the time-stamp says.
148 File: wget.info, Node: FTP Time-Stamping Internals, Prev: HTTP Time-Stamping Internals, Up: Time-Stamping
150 FTP Time-Stamping Internals
151 ===========================
153 In theory, FTP time-stamping works much the same as HTTP, only FTP
154 has no headers--time-stamps must be received from the directory
157 For each directory files must be retrieved from, Wget will use the
158 `LIST' command to get the listing. It will try to analyze the listing,
159 assuming that it is a Unix `ls -l' listing, and extract the
160 time-stamps. The rest is exactly the same as for HTTP.
162 Assumption that every directory listing is a Unix-style listing may
163 sound extremely constraining, but in practice it is not, as many
164 non-Unix FTP servers use the Unixoid listing format because most (all?)
165 of the clients understand it. Bear in mind that RFC959 defines no
166 standard way to get a file list, let alone the time-stamps. We can
167 only hope that a future standard will define this.
169 Another non-standard solution includes the use of `MDTM' command
170 that is supported by some FTP servers (including the popular
171 `wu-ftpd'), which returns the exact time of the specified file. Wget
172 may support this command in the future.
175 File: wget.info, Node: Startup File, Next: Examples, Prev: Time-Stamping, Up: Top
180 Once you know how to change default settings of Wget through command
181 line arguments, you may wish to make some of those settings permanent.
182 You can do that in a convenient way by creating the Wget startup
185 Besides `.wgetrc' is the "main" initialization file, it is
186 convenient to have a special facility for storing passwords. Thus Wget
187 reads and interprets the contents of `$HOME/.netrc', if it finds it.
188 You can find `.netrc' format in your system manuals.
190 Wget reads `.wgetrc' upon startup, recognizing a limited set of
195 * Wgetrc Location:: Location of various wgetrc files.
196 * Wgetrc Syntax:: Syntax of wgetrc.
197 * Wgetrc Commands:: List of available commands.
198 * Sample Wgetrc:: A wgetrc example.
201 File: wget.info, Node: Wgetrc Location, Next: Wgetrc Syntax, Prev: Startup File, Up: Startup File
206 When initializing, Wget will look for a "global" startup file,
207 `/usr/local/etc/wgetrc' by default (or some prefix other than
208 `/usr/local', if Wget was not installed there) and read commands from
211 Then it will look for the user's file. If the environmental variable
212 `WGETRC' is set, Wget will try to load that file. Failing that, no
213 further attempts will be made.
215 If `WGETRC' is not set, Wget will try to load `$HOME/.wgetrc'.
217 The fact that user's settings are loaded after the system-wide ones
218 means that in case of collision user's wgetrc *overrides* the
219 system-wide wgetrc (in `/usr/local/etc/wgetrc' by default). Fascist
223 File: wget.info, Node: Wgetrc Syntax, Next: Wgetrc Commands, Prev: Wgetrc Location, Up: Startup File
228 The syntax of a wgetrc command is simple:
232 The "variable" will also be called "command". Valid "values" are
233 different for different commands.
235 The commands are case-insensitive and underscore-insensitive. Thus
236 `DIr__PrefiX' is the same as `dirprefix'. Empty lines, lines beginning
237 with `#' and lines containing white-space only are discarded.
239 Commands that expect a comma-separated list will clear the list on an
240 empty command. So, if you wish to reset the rejection list specified in
241 global `wgetrc', you can do it with:
246 File: wget.info, Node: Wgetrc Commands, Next: Sample Wgetrc, Prev: Wgetrc Syntax, Up: Startup File
251 The complete set of commands is listed below, the letter after `='
252 denoting the value the command takes. It is `on/off' for `on' or `off'
253 (which can also be `1' or `0'), STRING for any non-empty string or N
254 for a positive integer. For example, you may specify `use_proxy = off'
255 to disable use of proxy servers by default. You may use `inf' for
256 infinite values, where appropriate.
258 Most of the commands have their equivalent command-line option
259 (*Note Invoking::), except some more obscure or rarely used ones.
261 accept/reject = STRING
262 Same as `-A'/`-R' (*Note Types of Files::).
265 Enable/disable host-prefixed file names. `-nH' disables it.
268 Enable/disable continuation of the retrieval - the same as `-c'
272 Enable/disable going to background - the same as `-b' (which
275 backup_converted = on/off
276 Enable/disable saving pre-converted files with the suffix `.orig'
277 - the same as `-K' (which enables it).
280 Consider relative URLs in URL input files forced to be interpreted
281 as HTML as being relative to STRING - the same as `-B'.
284 When set to off, disallow server-caching. See the `-C' option.
286 convert links = on/off
287 Convert non-relative links locally. The same as `-k'.
290 Ignore N remote directory components.
293 Debug mode, same as `-d'.
295 delete_after = on/off
296 Delete after download - the same as `--delete-after'.
299 Top of directory tree - the same as `-P'.
302 Turning dirstruct on or off - the same as `-x' or `-nd',
306 Same as `-D' (*Note Domain Acceptance::).
309 Specify the number of bytes "contained" in a dot, as seen
310 throughout the retrieval (1024 by default). You can postfix the
311 value with `k' or `m', representing kilobytes and megabytes,
312 respectively. With dot settings you can tailor the dot retrieval
313 to suit your needs, or you can use the predefined "styles" (*Note
317 Specify the number of dots that will be printed in each line
318 throughout the retrieval (50 by default).
321 Specify the number of dots in a single cluster (10 by default).
324 Specify the dot retrieval "style", as with `--dot-style'.
326 exclude_directories = STRING
327 Specify a comma-separated list of directories you wish to exclude
328 from download - the same as `-X' (*Note Directory-Based Limits::).
330 exclude_domains = STRING
331 Same as `--exclude-domains' (*Note Domain Acceptance::).
334 Follow FTP links from HTML documents - the same as `-f'.
337 Only follow certain HTML tags when doing a recursive retrieval,
338 just like `--follow-tags'.
341 If set to on, force the input filename to be regarded as an HTML
342 document - the same as `-F'.
345 Use STRING as FTP proxy, instead of the one specified in
349 Turn globbing on/off - the same as `-g'.
352 Define an additional header, like `--header'.
358 Use STRING as HTTP proxy, instead of the one specified in
362 Set HTTP user to STRING.
364 ignore_length = on/off
365 When set to on, ignore `Content-Length' header; the same as
369 Ignore certain HTML tags when doing a recursive retrieval, just
370 like `-G' / `--ignore-tags'.
372 include_directories = STRING
373 Specify a comma-separated list of directories you wish to follow
374 when downloading - the same as `-I'.
377 Read the URLs from STRING, like `-i'.
380 Consider data longer than specified in content-length header as
381 invalid (and retry getting it). The default behaviour is to save
382 as much data as there is, provided there is more than or equal to
383 the value in `Content-Length'.
386 Set logfile - the same as `-o'.
389 Your user name on the remote machine, for FTP. Defaults to
393 Turn mirroring on/off. The same as `-m'.
396 Turn reading netrc on or off.
402 Disallow retrieving outside the directory hierarchy, like
403 `--no-parent' (*Note Directory-Based Limits::).
406 Use STRING as the comma-separated list of domains to avoid in
407 proxy loading, instead of the one specified in environment.
409 output_document = STRING
410 Set the output filename - the same as `-O'.
412 page_requisites = on/off
413 Download all ancillary documents necessary for a single HTML page
414 to display properly - the same as `-p'.
417 Set passive FTP - the same as `--passive-ftp'.
420 Set your FTP password to PASSWORD. Without this setting, the
421 password defaults to `username@hostname.domainname'.
424 Set proxy authentication user name to STRING, like `--proxy-user'.
426 proxy_passwd = STRING
427 Set proxy authentication password to STRING, like `--proxy-passwd'.
430 Quiet mode - the same as `-q'.
433 Specify the download quota, which is useful to put in the global
434 `wgetrc'. When download quota is specified, Wget will stop
435 retrieving after the download sum has become greater than quota.
436 The quota can be specified in bytes (default), kbytes `k'
437 appended) or mbytes (`m' appended). Thus `quota = 5m' will set
438 the quota to 5 mbytes. Note that the user's startup file overrides
442 Recursion level - the same as `-l'.
445 Recursive on/off - the same as `-r'.
447 relative_only = on/off
448 Follow only relative links - the same as `-L' (*Note Relative
451 remove_listing = on/off
452 If set to on, remove FTP listings downloaded by Wget. Setting it
453 to off is the same as `-nr'.
455 retr_symlinks = on/off
456 When set to on, retrieve symbolic links as if they were plain
457 files; the same as `--retr-symlinks'.
460 Use (or not) `/robots.txt' file (*Note Robots::). Be sure to know
461 what you are doing before changing the default (which is `on').
463 server_response = on/off
464 Choose whether or not to print the HTTP and FTP server responses -
467 simple_host_check = on/off
468 Same as `-nh' (*Note Host Checking::).
474 Set timeout value - the same as `-T'.
476 timestamping = on/off
477 Turn timestamping on/off. The same as `-N' (*Note Time-Stamping::).
480 Set number of retries per URL - the same as `-t'.
483 Turn proxy support on/off. The same as `-Y'.
486 Turn verbose on/off - the same as `-v'/`-nv'.
489 Wait N seconds between retrievals - the same as `-w'.
492 Wait up to N seconds between retries of failed retrievals only -
493 the same as `--waitretry'. Note that this is turned on by default
494 in the global `wgetrc'.
497 File: wget.info, Node: Sample Wgetrc, Prev: Wgetrc Commands, Up: Startup File
502 This is the sample initialization file, as given in the distribution.
503 It is divided in two section--one for global usage (suitable for global
504 startup file), and one for local usage (suitable for `$HOME/.wgetrc').
505 Be careful about the things you change.
507 Note that almost all the lines are commented out. For a command to
508 have any effect, you must remove the `#' character at the beginning of
512 ### Sample Wget initialization file .wgetrc
515 ## You can use this file to change the default behaviour of wget or to
516 ## avoid having to type many many command-line options. This file does
517 ## not contain a comprehensive list of commands -- look at the manual
518 ## to find out what you can put into this file.
520 ## Wget initialization file can reside in /usr/local/etc/wgetrc
521 ## (global, for all users) or $HOME/.wgetrc (for a single user).
523 ## To use the settings in this file, you will have to uncomment them,
524 ## as well as change them, in most cases, as the values on the
525 ## commented-out lines are the default values (e.g. "off").
529 ## Global settings (useful for setting up in /usr/local/etc/wgetrc).
530 ## Think well before you change them, since they may reduce wget's
531 ## functionality, and make it behave contrary to the documentation:
534 # You can set retrieve quota for beginners by specifying a value
535 # optionally followed by 'K' (kilobytes) or 'M' (megabytes). The
536 # default quota is unlimited.
539 # You can lower (or raise) the default number of retries when
540 # downloading a file (default is 20).
543 # Lowering the maximum depth of the recursive retrieval is handy to
544 # prevent newbies from going too "deep" when they unwittingly start
545 # the recursive retrieval. The default is 5.
548 # Many sites are behind firewalls that do not allow initiation of
549 # connections from the outside. On these sites you have to use the
550 # `passive' feature of FTP. If you are behind such a firewall, you
551 # can turn this on to make Wget use passive FTP by default.
554 # The "wait" command below makes Wget wait between every connection.
555 # If, instead, you want Wget to wait only between retries of failed
556 # downloads, set waitretry to maximum number of seconds to wait (Wget
557 # will use "linear backoff", waiting 1 second after the first failure
558 # on a file, 2 seconds after the second failure, etc. up to this max).
563 ## Local settings (for a user to set in his $HOME/.wgetrc). It is
564 ## *highly* undesirable to put these settings in the global file, since
565 ## they are potentially dangerous to "normal" users.
567 ## Even when setting up your own ~/.wgetrc, you should know what you
568 ## are doing before doing so.
571 # Set this to on to use timestamping by default:
574 # It is a good idea to make Wget send your email address in a `From:'
575 # header with your request (so that server administrators can contact
576 # you in case of errors). Wget does *not* send `From:' by default.
577 #header = From: Your Name <username@site.domain>
579 # You can set up other headers, like Accept-Language. Accept-Language
580 # is *not* sent by default.
581 #header = Accept-Language: en
583 # You can set the default proxy for Wget to use. It will override the
584 # value in the environment.
585 #http_proxy = http://proxy.yoyodyne.com:18023/
587 # If you do not want to use proxy at all, set this to off.
590 # You can customize the retrieval outlook. Valid options are default,
591 # binary, mega and micro.
594 # Setting this to off makes Wget not download /robots.txt. Be sure to
595 # know *exactly* what /robots.txt is and how it is used before changing
599 # It can be useful to make Wget wait between connections. Set this to
600 # the number of seconds you want Wget to wait.
603 # You can force creating directory structure, even if a single is being
604 # retrieved, by setting this to on.
607 # You can turn on recursive retrieving by default (don't do this if
608 # you are not sure you know what it means) by setting this to on.
611 # To always back up file X as X.orig before converting its links (due
612 # to -k / --convert-links / convert_links = on having been specified),
613 # set this variable to on:
614 #backup_converted = off
616 # To have Wget follow FTP links from HTML files by default, set this
621 File: wget.info, Node: Examples, Next: Various, Prev: Startup File, Up: Top
626 The examples are classified into three sections, because of clarity.
627 The first section is a tutorial for beginners. The second section
628 explains some of the more complex program features. The third section
629 contains advice for mirror administrators, as well as even more complex
630 features (that some would call perverted).
634 * Simple Usage:: Simple, basic usage of the program.
635 * Advanced Usage:: Advanced techniques of usage.
636 * Guru Usage:: Mirroring and the hairy stuff.
639 File: wget.info, Node: Simple Usage, Next: Advanced Usage, Prev: Examples, Up: Examples
644 * Say you want to download a URL. Just type:
646 wget http://fly.cc.fer.hr/
648 The response will be something like:
650 --13:30:45-- http://fly.cc.fer.hr:80/en/
652 Connecting to fly.cc.fer.hr:80... connected!
653 HTTP request sent, awaiting response... 200 OK
654 Length: 4,694 [text/html]
658 13:30:46 (23.75 KB/s) - `index.html' saved [4694/4694]
660 * But what will happen if the connection is slow, and the file is
661 lengthy? The connection will probably fail before the whole file
662 is retrieved, more than once. In this case, Wget will try getting
663 the file until it either gets the whole of it, or exceeds the
664 default number of retries (this being 20). It is easy to change
665 the number of tries to 45, to insure that the whole file will
668 wget --tries=45 http://fly.cc.fer.hr/jpg/flyweb.jpg
670 * Now let's leave Wget to work in the background, and write its
671 progress to log file `log'. It is tiring to type `--tries', so we
674 wget -t 45 -o log http://fly.cc.fer.hr/jpg/flyweb.jpg &
676 The ampersand at the end of the line makes sure that Wget works in
677 the background. To unlimit the number of retries, use `-t inf'.
679 * The usage of FTP is as simple. Wget will take care of login and
682 $ wget ftp://gnjilux.cc.fer.hr/welcome.msg
683 --10:08:47-- ftp://gnjilux.cc.fer.hr:21/welcome.msg
685 Connecting to gnjilux.cc.fer.hr:21... connected!
686 Logging in as anonymous ... Logged in!
687 ==> TYPE I ... done. ==> CWD not needed.
688 ==> PORT ... done. ==> RETR welcome.msg ... done.
689 Length: 1,340 (unauthoritative)
693 10:08:48 (1.28 MB/s) - `welcome.msg' saved [1340]
695 * If you specify a directory, Wget will retrieve the directory
696 listing, parse it and convert it to HTML. Try:
698 wget ftp://prep.ai.mit.edu/pub/gnu/
702 File: wget.info, Node: Advanced Usage, Next: Guru Usage, Prev: Simple Usage, Up: Examples
707 * You would like to read the list of URLs from a file? Not a problem
712 If you specify `-' as file name, the URLs will be read from
715 * Create a mirror image of GNU WWW site (with the same directory
716 structure the original has) with only one try per document, saving
717 the log of the activities to `gnulog':
719 wget -r -t1 http://www.gnu.ai.mit.edu/ -o gnulog
721 * Retrieve the first layer of yahoo links:
723 wget -r -l1 http://www.yahoo.com/
725 * Retrieve the index.html of `www.lycos.com', showing the original
728 wget -S http://www.lycos.com/
730 * Save the server headers with the file:
731 wget -s http://www.lycos.com/
734 * Retrieve the first two levels of `wuarchive.wustl.edu', saving them
737 wget -P/tmp -l2 ftp://wuarchive.wustl.edu/
739 * You want to download all the GIFs from an HTTP directory. `wget
740 http://host/dir/*.gif' doesn't work, since HTTP retrieval does not
741 support globbing. In that case, use:
743 wget -r -l1 --no-parent -A.gif http://host/dir/
745 It is a bit of a kludge, but it works. `-r -l1' means to retrieve
746 recursively (*Note Recursive Retrieval::), with maximum depth of 1.
747 `--no-parent' means that references to the parent directory are
748 ignored (*Note Directory-Based Limits::), and `-A.gif' means to
749 download only the GIF files. `-A "*.gif"' would have worked too.
751 * Suppose you were in the middle of downloading, when Wget was
752 interrupted. Now you do not want to clobber the files already
753 present. It would be:
755 wget -nc -r http://www.gnu.ai.mit.edu/
757 * If you want to encode your own username and password to HTTP or
758 FTP, use the appropriate URL syntax (*Note URL Format::).
760 wget ftp://hniksic:mypassword@jagor.srce.hr/.emacs
762 * If you do not like the default retrieval visualization (1K dots
763 with 10 dots per cluster and 50 dots per line), you can customize
764 it through dot settings (*Note Wgetrc Commands::). For example,
765 many people like the "binary" style of retrieval, with 8K dots and
768 wget --dot-style=binary ftp://prep.ai.mit.edu/pub/gnu/README
770 You can experiment with other styles, like:
772 wget --dot-style=mega ftp://ftp.xemacs.org/pub/xemacs/xemacs-20.4/xemacs-20.4.tar.gz
773 wget --dot-style=micro http://fly.cc.fer.hr/
775 To make these settings permanent, put them in your `.wgetrc', as
776 described before (*Note Sample Wgetrc::).
779 File: wget.info, Node: Guru Usage, Prev: Advanced Usage, Up: Examples
784 * If you wish Wget to keep a mirror of a page (or FTP
785 subdirectories), use `--mirror' (`-m'), which is the shorthand for
786 `-r -N'. You can put Wget in the crontab file asking it to
787 recheck a site each Sunday:
790 0 0 * * 0 wget --mirror ftp://ftp.xemacs.org/pub/xemacs/ -o /home/me/weeklog
792 * You may wish to do the same with someone's home page. But you do
793 not want to download all those images--you're only interested in
796 wget --mirror -A.html http://www.w3.org/
798 * But what about mirroring the hosts networkologically close to you?
799 It seems so awfully slow because of all that DNS resolving. Just
800 use `-D' (*Note Domain Acceptance::).
802 wget -rN -Dsrce.hr http://www.srce.hr/
804 Now Wget will correctly find out that `regoc.srce.hr' is the same
805 as `www.srce.hr', but will not even take into consideration the
806 link to `www.mit.edu'.
808 * You have a presentation and would like the dumb absolute links to
809 be converted to relative? Use `-k':
813 * You would like the output documents to go to standard output
814 instead of to files? OK, but Wget will automatically shut up
815 (turn on `--quiet') to prevent mixing of Wget output and the
818 wget -O - http://jagor.srce.hr/ http://www.srce.hr/
820 You can also combine the two options and make weird pipelines to
821 retrieve the documents from remote hotlists:
823 wget -O - http://cool.list.com/ | wget --force-html -i -
826 File: wget.info, Node: Various, Next: Appendices, Prev: Examples, Up: Top
831 This chapter contains all the stuff that could not fit anywhere else.
835 * Proxies:: Support for proxy servers
836 * Distribution:: Getting the latest version.
837 * Mailing List:: Wget mailing list for announcements and discussion.
838 * Reporting Bugs:: How and where to report bugs.
839 * Portability:: The systems Wget works on.
840 * Signals:: Signal-handling performed by Wget.
843 File: wget.info, Node: Proxies, Next: Distribution, Prev: Various, Up: Various
848 "Proxies" are special-purpose HTTP servers designed to transfer data
849 from remote servers to local clients. One typical use of proxies is
850 lightening network load for users behind a slow connection. This is
851 achieved by channeling all HTTP and FTP requests through the proxy
852 which caches the transferred data. When a cached resource is requested
853 again, proxy will return the data from cache. Another use for proxies
854 is for companies that separate (for security reasons) their internal
855 networks from the rest of Internet. In order to obtain information
856 from the Web, their users connect and retrieve remote data using an
859 Wget supports proxies for both HTTP and FTP retrievals. The
860 standard way to specify proxy location, which Wget recognizes, is using
861 the following environment variables:
864 This variable should contain the URL of the proxy for HTTP
868 This variable should contain the URL of the proxy for HTTP
869 connections. It is quite common that HTTP_PROXY and FTP_PROXY are
873 This variable should contain a comma-separated list of domain
874 extensions proxy should *not* be used for. For instance, if the
875 value of `no_proxy' is `.mit.edu', proxy will not be used to
876 retrieve documents from MIT.
878 In addition to the environment variables, proxy location and settings
879 may be specified from within Wget itself.
884 This option may be used to turn the proxy support on or off. Proxy
885 support is on by default, provided that the appropriate environment
891 These startup file variables allow you to override the proxy
892 settings specified by the environment.
894 Some proxy servers require authorization to enable you to use them.
895 The authorization consists of "username" and "password", which must be
896 sent by Wget. As with HTTP authorization, several authentication
897 schemes exist. For proxy authorization only the `Basic' authentication
898 scheme is currently implemented.
900 You may specify your username and password either through the proxy
901 URL or through the command-line options. Assuming that the company's
902 proxy is located at `proxy.srce.hr' at port 8001, a proxy URL location
903 containing authorization data might look like this:
905 http://hniksic:mypassword@proxy.company.com:8001/
907 Alternatively, you may use the `proxy-user' and `proxy-password'
908 options, and the equivalent `.wgetrc' settings `proxy_user' and
909 `proxy_passwd' to set the proxy username and password.
912 File: wget.info, Node: Distribution, Next: Mailing List, Prev: Proxies, Up: Various
917 Like all GNU utilities, the latest version of Wget can be found at
918 the master GNU archive site prep.ai.mit.edu, and its mirrors. For
919 example, Wget 1.5.3+dev can be found at
920 `ftp://prep.ai.mit.edu/gnu/wget/wget-1.5.3+dev.tar.gz'
923 File: wget.info, Node: Mailing List, Next: Reporting Bugs, Prev: Distribution, Up: Various
928 Wget has its own mailing list at <wget@sunsite.auc.dk>, thanks to
929 Karsten Thygesen. The mailing list is for discussion of Wget features
930 and web, reporting Wget bugs (those that you think may be of interest
931 to the public) and mailing announcements. You are welcome to
932 subscribe. The more people on the list, the better!
934 To subscribe, send mail to <wget-subscribe@sunsite.auc.dk>. the
935 magic word `subscribe' in the subject line. Unsubscribe by mailing to
936 <wget-unsubscribe@sunsite.auc.dk>.
938 The mailing list is archived at `http://fly.cc.fer.hr/archive/wget'.
941 File: wget.info, Node: Reporting Bugs, Next: Portability, Prev: Mailing List, Up: Various
946 You are welcome to send bug reports about GNU Wget to
947 <bug-wget@gnu.org>. The bugs that you think are of the interest to the
948 public (i.e. more people should be informed about them) can be Cc-ed to
949 the mailing list at <wget@sunsite.auc.dk>.
951 Before actually submitting a bug report, please try to follow a few
954 1. Please try to ascertain that the behaviour you see really is a
955 bug. If Wget crashes, it's a bug. If Wget does not behave as
956 documented, it's a bug. If things work strange, but you are not
957 sure about the way they are supposed to work, it might well be a
960 2. Try to repeat the bug in as simple circumstances as possible.
961 E.g. if Wget crashes on `wget -rLl0 -t5 -Y0 http://yoyodyne.com -o
962 /tmp/log', you should try to see if it will crash with a simpler
965 Also, while I will probably be interested to know the contents of
966 your `.wgetrc' file, just dumping it into the debug message is
967 probably a bad idea. Instead, you should first try to see if the
968 bug repeats with `.wgetrc' moved out of the way. Only if it turns
969 out that `.wgetrc' settings affect the bug, should you mail me the
970 relevant parts of the file.
972 3. Please start Wget with `-d' option and send the log (or the
973 relevant parts of it). If Wget was compiled without debug support,
974 recompile it. It is *much* easier to trace bugs with debug support
977 4. If Wget has crashed, try to run it in a debugger, e.g. `gdb `which
978 wget` core' and type `where' to get the backtrace.
980 5. Find where the bug is, fix it and send me the patches. :-)
983 File: wget.info, Node: Portability, Next: Signals, Prev: Reporting Bugs, Up: Various
988 Since Wget uses GNU Autoconf for building and configuring, and avoids
989 using "special" ultra-mega-cool features of any particular Unix, it
990 should compile (and work) on all common Unix flavors.
992 Various Wget versions have been compiled and tested under many kinds
993 of Unix systems, including Solaris, Linux, SunOS, OSF (aka Digital
994 Unix), Ultrix, *BSD, IRIX, and others; refer to the file `MACHINES' in
995 the distribution directory for a comprehensive list. If you compile it
996 on an architecture not listed there, please let me know so I can update
999 Wget should also compile on the other Unix systems, not listed in
1000 `MACHINES'. If it doesn't, please let me know.
1002 Thanks to kind contributors, this version of Wget compiles and works
1003 on Microsoft Windows 95 and Windows NT platforms. It has been compiled
1004 successfully using MS Visual C++ 4.0, Watcom, and Borland C compilers,
1005 with Winsock as networking software. Naturally, it is crippled of some
1006 features available on Unix, but it should work as a substitute for
1007 people stuck with Windows. Note that the Windows port is *neither
1008 tested nor maintained* by me--all questions and problems should be
1009 reported to Wget mailing list at <wget@sunsite.auc.dk> where the
1010 maintainers will look at them.
1013 File: wget.info, Node: Signals, Prev: Portability, Up: Various
1018 Since the purpose of Wget is background work, it catches the hangup
1019 signal (`SIGHUP') and ignores it. If the output was on standard
1020 output, it will be redirected to a file named `wget-log'. Otherwise,
1021 `SIGHUP' is ignored. This is convenient when you wish to redirect the
1022 output of Wget after having started it.
1024 $ wget http://www.ifi.uio.no/~larsi/gnus.tar.gz &
1025 $ kill -HUP %% # Redirect the output to wget-log
1027 Other than that, Wget will not try to interfere with signals in any
1028 way. `C-c', `kill -TERM' and `kill -KILL' should kill it alike.
1031 File: wget.info, Node: Appendices, Next: Copying, Prev: Various, Up: Top
1036 This chapter contains some references I consider useful, like the
1037 Robots Exclusion Standard specification, as well as a list of
1038 contributors to GNU Wget.
1042 * Robots:: Wget as a WWW robot.
1043 * Security Considerations:: Security with Wget.
1044 * Contributors:: People who helped.
1047 File: wget.info, Node: Robots, Next: Security Considerations, Prev: Appendices, Up: Appendices
1052 Since Wget is able to traverse the web, it counts as one of the Web
1053 "robots". Thus Wget understands "Robots Exclusion Standard"
1054 (RES)--contents of `/robots.txt', used by server administrators to
1055 shield parts of their systems from wanderings of Wget.
1057 Norobots support is turned on only when retrieving recursively, and
1058 *never* for the first page. Thus, you may issue:
1060 wget -r http://fly.cc.fer.hr/
1062 First the index of fly.cc.fer.hr will be downloaded. If Wget finds
1063 anything worth downloading on the same host, only *then* will it load
1064 the robots, and decide whether or not to load the links after all.
1065 `/robots.txt' is loaded only once per host. Wget does not support the
1068 The description of the norobots standard was written, and is
1069 maintained by Martijn Koster <m.koster@webcrawler.com>. With his
1070 permission, I contribute a (slightly modified) TeXified version of the
1075 * Introduction to RES::
1077 * User-Agent Field::
1079 * Norobots Examples::
1082 File: wget.info, Node: Introduction to RES, Next: RES Format, Prev: Robots, Up: Robots
1087 "WWW Robots" (also called "wanderers" or "spiders") are programs
1088 that traverse many pages in the World Wide Web by recursively
1089 retrieving linked pages. For more information see the robots page.
1091 In 1993 and 1994 there have been occasions where robots have visited
1092 WWW servers where they weren't welcome for various reasons. Sometimes
1093 these reasons were robot specific, e.g. certain robots swamped servers
1094 with rapid-fire requests, or retrieved the same files repeatedly. In
1095 other situations robots traversed parts of WWW servers that weren't
1096 suitable, e.g. very deep virtual trees, duplicated information,
1097 temporary information, or cgi-scripts with side-effects (such as
1100 These incidents indicated the need for established mechanisms for
1101 WWW servers to indicate to robots which parts of their server should
1102 not be accessed. This standard addresses this need with an operational
1105 This document represents a consensus on 30 June 1994 on the robots
1106 mailing list (`robots@webcrawler.com'), between the majority of robot
1107 authors and other people with an interest in robots. It has also been
1108 open for discussion on the Technical World Wide Web mailing list
1109 (`www-talk@info.cern.ch'). This document is based on a previous working
1110 draft under the same title.
1112 It is not an official standard backed by a standards body, or owned
1113 by any commercial organization. It is not enforced by anybody, and there
1114 no guarantee that all current and future robots will use it. Consider
1115 it a common facility the majority of robot authors offer the WWW
1116 community to protect WWW server against unwanted accesses by their
1119 The latest version of this document can be found at
1120 `http://info.webcrawler.com/mak/projects/robots/norobots.html'.
1123 File: wget.info, Node: RES Format, Next: User-Agent Field, Prev: Introduction to RES, Up: Robots
1128 The format and semantics of the `/robots.txt' file are as follows:
1130 The file consists of one or more records separated by one or more
1131 blank lines (terminated by `CR', `CR/NL', or `NL'). Each record
1132 contains lines of the form:
1134 <field>:<optionalspace><value><optionalspace>
1136 The field name is case insensitive.
1138 Comments can be included in file using UNIX Bourne shell conventions:
1139 the `#' character is used to indicate that preceding space (if any) and
1140 the remainder of the line up to the line termination is discarded.
1141 Lines containing only a comment are discarded completely, and therefore
1142 do not indicate a record boundary.
1144 The record starts with one or more User-agent lines, followed by one
1145 or more Disallow lines, as detailed below. Unrecognized headers are
1148 The presence of an empty `/robots.txt' file has no explicit
1149 associated semantics, it will be treated as if it was not present, i.e.
1150 all robots will consider themselves welcome.
1153 File: wget.info, Node: User-Agent Field, Next: Disallow Field, Prev: RES Format, Up: Robots
1158 The value of this field is the name of the robot the record is
1159 describing access policy for.
1161 If more than one User-agent field is present the record describes an
1162 identical access policy for more than one robot. At least one field
1163 needs to be present per record.
1165 The robot should be liberal in interpreting this field. A case
1166 insensitive substring match of the name without version information is
1169 If the value is `*', the record describes the default access policy
1170 for any robot that has not matched any of the other records. It is not
1171 allowed to have multiple such records in the `/robots.txt' file.
1174 File: wget.info, Node: Disallow Field, Next: Norobots Examples, Prev: User-Agent Field, Up: Robots
1179 The value of this field specifies a partial URL that is not to be
1180 visited. This can be a full path, or a partial path; any URL that
1181 starts with this value will not be retrieved. For example,
1182 `Disallow: /help' disallows both `/help.html' and `/help/index.html',
1183 whereas `Disallow: /help/' would disallow `/help/index.html' but allow
1186 Any empty value, indicates that all URLs can be retrieved. At least
1187 one Disallow field needs to be present in a record.
1190 File: wget.info, Node: Norobots Examples, Prev: Disallow Field, Up: Robots
1195 The following example `/robots.txt' file specifies that no robots
1196 should visit any URL starting with `/cyberworld/map/' or `/tmp/':
1198 # robots.txt for http://www.site.com/
1201 Disallow: /cyberworld/map/ # This is an infinite virtual URL space
1202 Disallow: /tmp/ # these will soon disappear
1204 This example `/robots.txt' file specifies that no robots should
1205 visit any URL starting with `/cyberworld/map/', except the robot called
1208 # robots.txt for http://www.site.com/
1211 Disallow: /cyberworld/map/ # This is an infinite virtual URL space
1213 # Cybermapper knows where to go.
1214 User-agent: cybermapper
1217 This example indicates that no robots should visit this site further:
1224 File: wget.info, Node: Security Considerations, Next: Contributors, Prev: Robots, Up: Appendices
1226 Security Considerations
1227 =======================
1229 When using Wget, you must be aware that it sends unencrypted
1230 passwords through the network, which may present a security problem.
1231 Here are the main issues, and some solutions.
1233 1. The passwords on the command line are visible using `ps'. If this
1234 is a problem, avoid putting passwords from the command line--e.g.
1235 you can use `.netrc' for this.
1237 2. Using the insecure "basic" authentication scheme, unencrypted
1238 passwords are transmitted through the network routers and gateways.
1240 3. The FTP passwords are also in no way encrypted. There is no good
1241 solution for this at the moment.
1243 4. Although the "normal" output of Wget tries to hide the passwords,
1244 debugging logs show them, in all forms. This problem is avoided by
1245 being careful when you send debug logs (yes, even when you send
1249 File: wget.info, Node: Contributors, Prev: Security Considerations, Up: Appendices
1254 GNU Wget was written by Hrvoje Niksic <hniksic@iskon.hr>. However,
1255 its development could never have gone as far as it has, were it not for
1256 the help of many people, either with bug reports, feature proposals,
1257 patches, or letters saying "Thanks!".
1259 Special thanks goes to the following people (no particular order):
1261 * Karsten Thygesen--donated the mailing list and the initial FTP
1264 * Shawn McHorse--bug reports and patches.
1266 * Kaveh R. Ghazi--on-the-fly `ansi2knr'-ization.
1268 * Gordon Matzigkeit--`.netrc' support.
1270 * Zlatko Calusic, Tomislav Vujec and Drazen Kacar--feature
1271 suggestions and "philosophical" discussions.
1273 * Darko Budor--initial port to Windows.
1275 * Antonio Rosella--help and suggestions, plus the Italian
1278 * Tomislav Petrovic, Mario Mikocevic--many bug reports and
1281 * Francois Pinard--many thorough bug reports and discussions.
1283 * Karl Eichwalder--lots of help with internationalization and other
1286 * Junio Hamano--donated support for Opie and HTTP `Digest'
1289 * Brian Gough--a generous donation.
1291 The following people have provided patches, bug/build reports, useful
1292 suggestions, beta testing services, fan mail and all the other things
1293 that make maintenance so much fun:
1295 Tim Adam, Martin Baehr, Dieter Baron, Roger Beeman and the Gurus at
1296 Cisco, Dan Berger, Mark Boyns, John Burden, Wanderlei Cavassin, Gilles
1297 Cedoc, Tim Charron, Noel Cragg, Kristijan Conkas, Andrew Deryabin,
1298 Damir Dzeko, Andrew Davison, Ulrich Drepper, Marc Duponcheel,
1299 Aleksandar Erkalovic, Andy Eskilsson, Masashi Fujita, Howard Gayle,
1300 Marcel Gerrits, Hans Grobler, Mathieu Guillaume, Dan Harkless, Heiko
1301 Herold, Karl Heuer, HIROSE Masaaki, Gregor Hoffleit, Erik Magnus
1302 Hulthen, Richard Huveneers, Simon Josefsson, Mario Juric, Goran
1303 Kezunovic, Robert Kleine, Fila Kolodny, Alexander Kourakos, Martin
1304 Kraemer, Simos KSenitellis, Hrvoje Lacko, Daniel S. Lewart, Dave Love,
1305 Jordan Mendelson, Lin Zhe Min, Charlie Negyesi, Andrew Pollock, Steve
1306 Pothier, Jan Prikryl, Marin Purgar, Keith Refson, Tobias Ringstrom,
1307 Juan Jose Rodrigues, Edward J. Sabol, Heinz Salzmann, Robert Schmidt,
1308 Toomas Soome, Tage Stabell-Kulo, Sven Sternberger, Markus Strasser,
1309 Szakacsits Szabolcs, Mike Thomas, Russell Vincent, Charles G Waldman,
1310 Douglas E. Wegscheid, Jasmin Zainul, Bojan Zdrnja, Kristijan Zimmer.
1312 Apologies to all who I accidentally left out, and many thanks to all
1313 the subscribers of the Wget mailing list.