5. Supported URLs
::WORK:: List supported URL types, specific RFCs, etc.
Uniform Resource Locators (URLs) are a specific form of
Uniform Resource Identifiers (URI) described in RFC2396
which updates RFC1738 and RFC1808.
RFC2016 defines uniform resource agents.
URI have the form scheme:scheme-specific-part, where scheme
appears in the menu below for URLs supported by Emacs/W3.
FTP, NFS, HTTP, HTTPS, rlogin, telnet, tn3270,
IRC and gopher URLs all have the form
scheme://[userinfo@]hostname[:port][/path]
where `[' and `]' delimit optional parts. userinfo
sometimes takes the form userinfo:password but you should
beware of the security risks of sending cleartext passwords.
hostname may be a domain name or a dotted decimal address. If the
`:port' is omitted then Emacs/W3 will use the well known port
for that service. With the possible exception of telnet, it is
very rare for ports to be specified, and it is possible using a
non-standard port may have undesired consequences if a different service
is listening on that port (eg. a gopher URL specifying the
SMTP port can cause mail to be sent), but See section url-bad-port-list. The meaning of the path component depends on
the service.
5.1 File URLs
This allows Emacs/W3 to read arbitary files from hosts. Compressed
files are handled, but support is hard-coded so that
jka-compr-compression-info-list and so on have no affect.
Suffixes recognized are `.z', `.gz' and `.Z'. If the
URL points to a directory, then it will try to retrieve a file
named by url-directory-index-file (`index.html' by default)
and parse it, otherwise you get a directory listing in `dired'
mode. If the URL refers to a file on a remote host, then Emacs/W3
uses `ange-ftp' (@xref{Top, , ange-ftp}) or `efs'
(@xref{Top, , efs}) to retrieve the file. ftp:// and file:// are
synonymous for Emacs/W3.
5.2 FTP URLs
For details of usage see 5.1 File URLs. In Emacs/W3 file and FTP URLs
are synonymous and files on the localhost are retrieved directly rather
than by FTP. Emacs/W3 relies on `ange-ftp'
(@xref{Top, , ange-ftp}) or `efs' (@xref{Top, , efs}) to do the actual
transfers.
5.3 NFS URLs
gdj1: have I misunderstood the point of this?
Since NFS is fairly transparent to the user (at least when it's
working), there isn't very much to say here. An nfs URL is similar
to a file URL except that it points to a file on a remote host that
is handled by the automounter on the local host. The variable
url-nfs-automounter-directory-spec may need to be tweaked
depending on local configuration. The NFS URL is defined in
RFC2224.
5.4 info
Info URLs are not an officially recognised URL (gdj1: is this
right?), but Emacs/W3 will parse them to produce a reference to a
TeXinfo node, or `Top' if one is not specified. `Info-mode'
will be used to browse the document.
5.5 http/https
The HyperText Transfer Protocol is the protocol used to get documents
from the World Wide Web. Emacs/W3 supports HTTP version 1.0 as
defined in RFC1945 -- now superseded by version 1.1 defined in
RFC2068.
If url-honor-refresh-requests is nil then `Refresh'
headers will not be honoured, if t then they will always be
honoured, otherwise the user will be asked for each request. The
default is t. url-be-anal-about-file-attributes controls
whether HTTP is used to discover file attributes, or whether
they're just guessed. The default is nil which means that
Emacs/W3 will make educated guesses.
HTTPS is a secure version of HTTP defined in RFC2069
(gdj1: ?). Emacs/W3 requires SSL to handle this, See section D. Installing SSL.
5.6 mailto
A mailto URL will send an email message to the address in the
URL, for example `mailto:foo@bar.com' would compose a
message to foo@bar.com. Emacs/W3 uses the command specified by the
url-mail-command variable to compose the email, this is url-mail by
default which uses Gnu's `message' mode (@xref{Top, Message, , message}) if
available, otherwise the standard mail command. An X-Url-From
header field containing the URL of the document that contained the
mailto URL is added, as is an X-Mailer header field containing the
version of Emacs/W3 being used.
RFC2368 extends the definition of mailto URLs in RFC1738.
The form of a mailto URL is
`mailto:mailbox[?header=contents[&header=contents]]'
where an arbitary number of headers can be added. If the
header is `body', then contents is put in the body
otherwise a header header field is created with contents as its
contents. Note that Emacs/W3 does not consider any headers as
`dangerous' so you should check them before sending the message.
Email messages are defined in RFC822.
5.7 mailserver
A mailserver URL allows you to send an email to a person, but this
URL optionally specifies a subject and a body. The basic format is
`mailserver:[mailbox/subject[/body]'. Thus,
`mailserver:foo@bar.com/wibble/flibble' will compose a message to
foo@bar.com with subject as the subject and body already in
the body of the email. Note that both the subject and the body are
hexed, but the subject cannot contain newlines.
5.8 news/nntp/snews
If the URL doesn't specify a host, then the host in
url-news-server will be used, and unless the URL has a port
the news port as defined in url-default-ports (119) will be used.
The username and password specified in the URL will be used if
present. The URL may contain a message-id, in which case that
article is displayed; it may contain a newsgroup in which case Gnus is
used to display the newsgroup; or it may by empty in which case Gnus is
called with no arguments. Emacs/W3 requires Gnus v5.x or Red, Quassia
or Pterodactyl Gnus, See section `Top' in Gnus. The variable
url-news-use-article-mode controls the displaying of news
articles; if non-nil then articles are displayed in Gnus article
mode, otherwise they are turned into HTML and rendered by Emacs/W3.
An NNTP URL is the same as a news URL, except that the
URL may specify an article by its number.
5.9 rlogin/telnet/tn3270
To handle rlogin, telnet and tn3270 URLs, Emacs/W3 runs an
rlogin, telnet or tn3270 session (the program names
and arguments are hardcoded) in a terminal-emulator buffer.
Well-known ports are used if the URL does not specify a port.
5.10 irc
Internet Relay Chat (IRC) is handled by handing off the IRC
session to a function named in url-irc-function. This function
must take five argumenst, host, port, channel,
user and password. The channel argument specifies the
channel to join immediately, this can be nil. By default this is
url-irc-zenirc which processes the arguments and lets
zenirc handle the session.
5.11 data
Data URLs contain MIME data in the URL itself, by default
the data is 8bit encoded `text/plain', but the URL can specify either
or both the content-type and the encoding. Emacs/W3 will parse the
URL's data as MIME and display it appropriately. See section 6. MIME Support.
5.12 gopher
Gopher (Go for) was in someways the precurser to the world wide
web and is becoming rarer as the web becomes more powerful.
Nevertheless, there are still many gopher sites around and Emacs/W3
supports this protocol (of course). The variable
url-gopher-labels maps gopher types to something else (gdj: ?)
for displaying the gopher menus. url-gopher-icons maps gopher
types to pictures. url-gopher-to-mime maps gopher types to
MIME types. If url-use-hypertext-gopher is non-nil,
then gopher pages will be turned into HTML for Emacs/W3 to parse
and display normally, otherwise Emacs/W3 will let `gopher.el'
handle all gopher requests which will lose gopher+ support and inlined
searching. This is t by default.
5.13 finger
Finger URLs will finger a given user at a given host, or `localhost'
if no host is specified, processing the results to create an HTML
page for Emacs/W3 to display.
5.14 netrek
This is unsupported at present.
This document was generated by XEmacs shared group account on December, 19 2009
using texi2html 1.65.