[ViewVC] Diff of: cvs/AnyEvent-HTTP/HTTP.pm

Comparing AnyEvent-HTTP/HTTP.pm (file contents):
Revision 1.39 by root, Thu Jun 11 12:46:01 2009 UTC vs.
Revision 1.69 by root, Fri Dec 31 19:32:47 2010 UTC

…		…
39	package AnyEvent::HTTP;	39	package AnyEvent::HTTP;
40		40
41	use strict;	41	use strict;
42	no warnings;	42	no warnings;
43		43
44	use Carp;	44	use Errno ();
45		45
46	use AnyEvent ();	46	use AnyEvent 5.0 ();
47	use AnyEvent::Util ();	47	use AnyEvent::Util ();
48	use AnyEvent::Socket ();
49	use AnyEvent::Handle ();	48	use AnyEvent::Handle ();
50		49
51	use base Exporter::;	50	use base Exporter::;
52		51
53	our $VERSION = '1.12';	52	our $VERSION = '1.5';
54		53
55	our @EXPORT = qw(http_get http_post http_head http_request);	54	our @EXPORT = qw(http_get http_post http_head http_request);
56		55
57	our $USERAGENT = "Mozilla/5.0 (compatible; AnyEvent::HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)";	56	our $USERAGENT = "Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)";
58	our $MAX_RECURSE = 10;	57	our $MAX_RECURSE = 10;
59	our $MAX_PERSISTENT = 8;	58	our $MAX_PERSISTENT = 8;
60	our $PERSISTENT_TIMEOUT = 2;	59	our $PERSISTENT_TIMEOUT = 2;
61	our $TIMEOUT = 300;	60	our $TIMEOUT = 300;
62		61
63	# changing these is evil	62	# changing these is evil
64	our $MAX_PERSISTENT_PER_HOST = 2;	63	our $MAX_PERSISTENT_PER_HOST = 0;
65	our $MAX_PER_HOST = 4;	64	our $MAX_PER_HOST = 4;
66		65
67	our $PROXY;	66	our $PROXY;
68	our $ACTIVE = 0;	67	our $ACTIVE = 0;
69		68
…		…
92	must be an absolute http or https URL.	91	must be an absolute http or https URL.
93		92
94	When called in void context, nothing is returned. In other contexts,	93	When called in void context, nothing is returned. In other contexts,
95	C<http_request> returns a "cancellation guard" - you have to keep the	94	C<http_request> returns a "cancellation guard" - you have to keep the
96	object at least alive until the callback get called. If the object gets	95	object at least alive until the callback get called. If the object gets
97	destroyed before the callbakc is called, the request will be cancelled.	96	destroyed before the callback is called, the request will be cancelled.
98		97
99	The callback will be called with the response data as first argument	98	The callback will be called with the response body data as first argument
100	(or C<undef> if it wasn't available due to errors), and a hash-ref with	99	(or C<undef> if an error occured), and a hash-ref with response headers
101	response headers as second argument.	100	(and trailers) as second argument.
102		101
103	All the headers in that hash are lowercased. In addition to the response	102	All the headers in that hash are lowercased. In addition to the response
104	headers, the "pseudo-headers" C<HTTPVersion>, C<Status> and C<Reason>	103	headers, the "pseudo-headers" (uppercase to avoid clashing with possible
		104	response headers) C<HTTPVersion>, C<Status> and C<Reason> contain the
105	contain the three parts of the HTTP Status-Line of the same name. The	105	three parts of the HTTP Status-Line of the same name. If an error occurs
		106	during the body phase of a request, then the original C<Status> and
		107	C<Reason> values from the header are available as C<OrigStatus> and
		108	C<OrigReason>.
		109
106	pseudo-header C<URL> contains the original URL (which can differ from the	110	The pseudo-header C<URL> contains the actual URL (which can differ from
107	requested URL when following redirects).	111	the requested URL when following redirects - for example, you might get
		112	an error that your URL scheme is not supported even though your URL is a
		113	valid http URL because it redirected to an ftp URL, in which case you can
		114	look at the URL pseudo header).
		115
		116	The pseudo-header C<Redirect> only exists when the request was a result
		117	of an internal redirect. In that case it is an array reference with
		118	the C<($data, $headers)> from the redirect response. Note that this
		119	response could in turn be the result of a redirect itself, and C<<
		120	$headers->{Redirect}[1]{Redirect} >> will then contain the original
		121	response, and so on.
108		122
109	If the server sends a header multiple times, then their contents will be	123	If the server sends a header multiple times, then their contents will be
110	joined together with a comma (C<,>), as per the HTTP spec.	124	joined together with a comma (C<,>), as per the HTTP spec.
111		125
112	If an internal error occurs, such as not being able to resolve a hostname,	126	If an internal error occurs, such as not being able to resolve a hostname,
113	then C<$data> will be C<undef>, C<< $headers->{Status} >> will be C<599>	127	then C<$data> will be C<undef>, C<< $headers->{Status} >> will be C<59x>
114	and the C<Reason> pseudo-header will contain an error message.	128	(usually C<599>) and the C<Reason> pseudo-header will contain an error
		129	message.
115		130
116	A typical callback might look like this:	131	A typical callback might look like this:
117		132
118	sub {	133	sub {
119	my ($body, $hdr) = @_;	134	my ($body, $hdr) = @_;
…		…
135	Whether to recurse requests or not, e.g. on redirects, authentication	150	Whether to recurse requests or not, e.g. on redirects, authentication
136	retries and so on, and how often to do so.	151	retries and so on, and how often to do so.
137		152
138	=item headers => hashref	153	=item headers => hashref
139		154
140	The request headers to use. Currently, C<http_request> may provide its	155	The request headers to use. Currently, C<http_request> may provide its own
141	own C<Host:>, C<Content-Length:>, C<Connection:> and C<Cookie:> headers	156	C<Host:>, C<Content-Length:>, C<Connection:> and C<Cookie:> headers and
142	and will provide defaults for C<User-Agent:> and C<Referer:>.	157	will provide defaults at least for C<TE:>, C<Referer:> and C<User-Agent:>
		158	(this can be suppressed by using C<undef> for these headers in which case
		159	they won't be sent at all).
143		160
144	=item timeout => $seconds	161	=item timeout => $seconds
145		162
146	The time-out to use for various stages - each connect attempt will reset	163	The time-out to use for various stages - each connect attempt will reset
147	the timeout, as will read or write activity. Default timeout is 5 minutes.	164	the timeout, as will read or write activity, i.e. this is not an overall
		165	timeout.
		166
		167	Default timeout is 5 minutes.
148		168
149	=item proxy => [$host, $port[, $scheme]] or undef	169	=item proxy => [$host, $port[, $scheme]] or undef
150		170
151	Use the given http proxy for all requests. If not specified, then the	171	Use the given http proxy for all requests. If not specified, then the
152	default proxy (as specified by C<$ENV{http_proxy}>) is used.	172	default proxy (as specified by C<$ENV{http_proxy}>) is used.
153		173
154	C<$scheme> must be either missing or C<http> for HTTP, or C<https> for	174	C<$scheme> must be either missing, C<http> for HTTP or C<https> for
155	HTTPS.	175	HTTPS.
156		176
157	=item body => $string	177	=item body => $string
158		178
159	The request body, usually empty. Will be-sent as-is (future versions of	179	The request body, usually empty. Will be sent as-is (future versions of
160	this module might offer more options).	180	this module might offer more options).
161		181
162	=item cookie_jar => $hash_ref	182	=item cookie_jar => $hash_ref
163		183
164	Passing this parameter enables (simplified) cookie-processing, loosely	184	Passing this parameter enables (simplified) cookie-processing, loosely
165	based on the original netscape specification.	185	based on the original netscape specification.
166		186
167	The C<$hash_ref> must be an (initially empty) hash reference which will	187	The C<$hash_ref> must be an (initially empty) hash reference which will
168	get updated automatically. It is possible to save the cookie_jar to	188	get updated automatically. It is possible to save the cookie_jar to
169	persistent storage with something like JSON or Storable, but this is not	189	persistent storage with something like JSON or Storable, but this is not
170	recommended, as expire times are currently being ignored.	190	recommended, as expiry times are currently being ignored.
171		191
172	Note that this cookie implementation is not of very high quality, nor	192	Note that this cookie implementation is not of very high quality, nor
173	meant to be complete. If you want complete cookie management you have to	193	meant to be complete. If you want complete cookie management you have to
174	do that on your own. C<cookie_jar> is meant as a quick fix to get some	194	do that on your own. C<cookie_jar> is meant as a quick fix to get some
175	cookie-using sites working. Cookies are a privacy disaster, do not use	195	cookie-using sites working. Cookies are a privacy disaster, do not use
176	them unless required to.	196	them unless required to.
177		197
		198	When cookie processing is enabled, the C<Cookie:> and C<Set-Cookie:>
		199	headers will be ste and handled by this module, otherwise they will be
		200	left untouched.
		201
		202	=item tls_ctx => $scheme \| $tls_ctx
		203
		204	Specifies the AnyEvent::TLS context to be used for https connections. This
		205	parameter follows the same rules as the C<tls_ctx> parameter to
		206	L<AnyEvent::Handle>, but additionally, the two strings C<low> or
		207	C<high> can be specified, which give you a predefined low-security (no
		208	verification, highest compatibility) and high-security (CA and common-name
		209	verification) TLS context.
		210
		211	The default for this option is C<low>, which could be interpreted as "give
		212	me the page, no matter what".
		213
		214	=item on_prepare => $callback->($fh)
		215
		216	In rare cases you need to "tune" the socket before it is used to
		217	connect (for exmaple, to bind it on a given IP address). This parameter
		218	overrides the prepare callback passed to C<AnyEvent::Socket::tcp_connect>
		219	and behaves exactly the same way (e.g. it has to provide a
		220	timeout). See the description for the C<$prepare_cb> argument of
		221	C<AnyEvent::Socket::tcp_connect> for details.
		222
		223	=item tcp_connect => $callback->($host, $service, $connect_cb, $prepare_cb)
		224
		225	In even rarer cases you want total control over how AnyEvent::HTTP
		226	establishes connections. Normally it uses L<AnyEvent::Socket::tcp_connect>
		227	to do this, but you can provide your own C<tcp_connect> function -
		228	obviously, it has to follow the same calling conventions, except that it
		229	may always return a connection guard object.
		230
		231	There are probably lots of weird uses for this function, starting from
		232	tracing the hosts C<http_request> actually tries to connect, to (inexact
		233	but fast) host => IP address caching or even socks protocol support.
		234
		235	=item on_header => $callback->($headers)
		236
		237	When specified, this callback will be called with the header hash as soon
		238	as headers have been successfully received from the remote server (not on
		239	locally-generated errors).
		240
		241	It has to return either true (in which case AnyEvent::HTTP will continue),
		242	or false, in which case AnyEvent::HTTP will cancel the download (and call
		243	the finish callback with an error code of C<598>).
		244
		245	This callback is useful, among other things, to quickly reject unwanted
		246	content, which, if it is supposed to be rare, can be faster than first
		247	doing a C<HEAD> request.
		248
		249	The downside is that cancelling the request makes it impossible to re-use
		250	the connection. Also, the C<on_header> callback will not receive any
		251	trailer (headers sent after the response body).
		252
		253	Example: cancel the request unless the content-type is "text/html".
		254
		255	on_header => sub {
		256	$_[0]{"content-type"} =~ /^text\/html\s*(?:;\|$)/
		257	},
		258
		259	=item on_body => $callback->($partial_body, $headers)
		260
		261	When specified, all body data will be passed to this callback instead of
		262	to the completion callback. The completion callback will get the empty
		263	string instead of the body data.
		264
		265	It has to return either true (in which case AnyEvent::HTTP will continue),
		266	or false, in which case AnyEvent::HTTP will cancel the download (and call
		267	the completion callback with an error code of C<598>).
		268
		269	The downside to cancelling the request is that it makes it impossible to
		270	re-use the connection.
		271
		272	This callback is useful when the data is too large to be held in memory
		273	(so the callback writes it to a file) or when only some information should
		274	be extracted, or when the body should be processed incrementally.
		275
		276	It is usually preferred over doing your own body handling via
		277	C<want_body_handle>, but in case of streaming APIs, where HTTP is
		278	only used to create a connection, C<want_body_handle> is the better
		279	alternative, as it allows you to install your own event handler, reducing
		280	resource usage.
		281
		282	=item want_body_handle => $enable
		283
		284	When enabled (default is disabled), the behaviour of AnyEvent::HTTP
		285	changes considerably: after parsing the headers, and instead of
		286	downloading the body (if any), the completion callback will be
		287	called. Instead of the C<$body> argument containing the body data, the
		288	callback will receive the L<AnyEvent::Handle> object associated with the
		289	connection. In error cases, C<undef> will be passed. When there is no body
		290	(e.g. status C<304>), the empty string will be passed.
		291
		292	The handle object might or might not be in TLS mode, might be connected to
		293	a proxy, be a persistent connection etc., and configured in unspecified
		294	ways. The user is responsible for this handle (it will not be used by this
		295	module anymore).
		296
		297	This is useful with some push-type services, where, after the initial
		298	headers, an interactive protocol is used (typical example would be the
		299	push-style twitter API which starts a JSON/XML stream).
		300
		301	If you think you need this, first have a look at C<on_body>, to see if
		302	that doesn't solve your problem in a better way.
		303
178	=back	304	=back
179		305
180	Example: make a simple HTTP GET request for http://www.nethype.de/	306	Example: do a simple HTTP GET request for http://www.nethype.de/ and print
		307	the response body.
181		308
182	http_request GET => "http://www.nethype.de/", sub {	309	http_request GET => "http://www.nethype.de/", sub {
183	my ($body, $hdr) = @_;	310	my ($body, $hdr) = @_;
184	print "$body\n";	311	print "$body\n";
185	};	312	};
186		313
187	Example: make a HTTP HEAD request on https://www.google.com/, use a	314	Example: do a HTTP HEAD request on https://www.google.com/, use a
188	timeout of 30 seconds.	315	timeout of 30 seconds.
189		316
190	http_request	317	http_request
191	GET => "https://www.google.com",	318	GET => "https://www.google.com",
192	timeout => 30,	319	timeout => 30,
…		…
195	use Data::Dumper;	322	use Data::Dumper;
196	print Dumper $hdr;	323	print Dumper $hdr;
197	}	324	}
198	;	325	;
199		326
200	Example: make another simple HTTP GET request, but immediately try to	327	Example: do another simple HTTP GET request, but immediately try to
201	cancel it.	328	cancel it.
202		329
203	my $request = http_request GET => "http://www.nethype.de/", sub {	330	my $request = http_request GET => "http://www.nethype.de/", sub {
204	my ($body, $hdr) = @_;	331	my ($body, $hdr) = @_;
205	print "$body\n";	332	print "$body\n";
…		…
237	push @{ $CO_SLOT{$_[0]}[1] }, $_[1];	364	push @{ $CO_SLOT{$_[0]}[1] }, $_[1];
238		365
239	_slot_schedule $_[0];	366	_slot_schedule $_[0];
240	}	367	}
241		368
242	our $qr_nl = qr<\015?\012>;	369	# continue to parse $_ for headers and place them into the arg
243	our $qr_nlnl = qr<\015?\012\015?\012>;	370	sub parse_hdr() {
		371	my %hdr;
		372
		373	# things seen, not parsed:
		374	# p3pP="NON CUR OTPi OUR NOR UNI"
		375
		376	$hdr{lc $1} .= ",$2"
		377	while /\G
		378	([^:\000-\037]*):
		379	[\011\040]*
		380	((?: [^\012]+ \| \012[\011\040] )*)
		381	\012
		382	/gxc;
		383
		384	/\G$/
		385	or return;
		386
		387	# remove the "," prefix we added to all headers above
		388	substr $_, 0, 1, ""
		389	for values %hdr;
		390
		391	\%hdr
		392	}
		393
		394	our $qr_nlnl = qr{(?<![^\012])\015?\012};
		395
		396	our $TLS_CTX_LOW = { cache => 1, sslv2 => 1 };
		397	our $TLS_CTX_HIGH = { cache => 1, verify => 1, verify_peername => "https" };
244		398
245	sub http_request($$@) {	399	sub http_request($$@) {
246	my $cb = pop;	400	my $cb = pop;
247	my ($method, $url, %arg) = @_;	401	my ($method, $url, %arg) = @_;
248		402
249	my %hdr;	403	my %hdr;
		404
		405	$arg{tls_ctx} = $TLS_CTX_LOW if $arg{tls_ctx} eq "low" \|\| !exists $arg{tls_ctx};
		406	$arg{tls_ctx} = $TLS_CTX_HIGH if $arg{tls_ctx} eq "high";
250		407
251	$method = uc $method;	408	$method = uc $method;
252		409
253	if (my $hdr = $arg{headers}) {	410	if (my $hdr = $arg{headers}) {
254	while (my ($k, $v) = each %$hdr) {	411	while (my ($k, $v) = each %$hdr) {
255	$hdr{lc $k} = $v;	412	$hdr{lc $k} = $v;
256	}	413	}
257	}	414	}
258		415
		416	# pseudo headers for all subsequent responses
		417	my @pseudo = (URL => $url);
		418	push @pseudo, Redirect => delete $arg{Redirect} if exists $arg{Redirect};
		419
259	my $recurse = exists $arg{recurse} ? delete $arg{recurse} : $MAX_RECURSE;	420	my $recurse = exists $arg{recurse} ? delete $arg{recurse} : $MAX_RECURSE;
260		421
261	return $cb->(undef, { Status => 599, Reason => "recursion limit reached", URL => $url })	422	return $cb->(undef, { @pseudo, Status => 599, Reason => "Too many redirections" })
262	if $recurse < 0;	423	if $recurse < 0;
263		424
264	my $proxy = $arg{proxy} \|\| $PROXY;	425	my $proxy = $arg{proxy} \|\| $PROXY;
265	my $timeout = $arg{timeout} \|\| $TIMEOUT;	426	my $timeout = $arg{timeout} \|\| $TIMEOUT;
266		427
267	$hdr{"user-agent"} \|\|= $USERAGENT;
268
269	my ($uscheme, $uauthority, $upath, $query, $fragment) =	428	my ($uscheme, $uauthority, $upath, $query, $fragment) =
270	$url =~ m\|(?:([^:/?#]+):)?(?://([^/?#]))?([^?#])(?:\?([^#]))?(?:#(.))?\|;	429	$url =~ m\|(?:([^:/?#]+):)?(?://([^/?#]))?([^?#])(?:(\?[^#]))?(?:#(.))?\|;
271		430
272	$uscheme = lc $uscheme;	431	$uscheme = lc $uscheme;
273		432
274	my $uport = $uscheme eq "http" ? 80	433	my $uport = $uscheme eq "http" ? 80
275	: $uscheme eq "https" ? 443	434	: $uscheme eq "https" ? 443
276	: return $cb->(undef, { Status => 599, Reason => "only http and https URL schemes supported", URL => $url });	435	: return $cb->(undef, { @pseudo, Status => 599, Reason => "Only http and https URL schemes supported" });
277
278	$hdr{referer} \|\|= "$uscheme://$uauthority$upath"; # leave out fragment and query string, just a heuristic
279		436
280	$uauthority =~ /^(?: .*\@ )? ([^\@:]+) (?: : (\d+) )?$/x	437	$uauthority =~ /^(?: .*\@ )? ([^\@:]+) (?: : (\d+) )?$/x
281	or return $cb->(undef, { Status => 599, Reason => "unparsable URL", URL => $url });	438	or return $cb->(undef, { @pseudo, Status => 599, Reason => "Unparsable URL" });
282		439
283	my $uhost = $1;	440	my $uhost = $1;
284	$uport = $2 if defined $2;	441	$uport = $2 if defined $2;
285		442
		443	$hdr{host} = defined $2 ? "$uhost:$2" : "$uhost"
		444	unless exists $hdr{host};
		445
286	$uhost =~ s/^\[(.*)\]$/$1/;	446	$uhost =~ s/^\[(.*)\]$/$1/;
287	$upath .= "?$query" if length $query;	447	$upath .= $query if length $query;
288		448
289	$upath =~ s%^/?%/%;	449	$upath =~ s%^/?%/%;
290		450
291	# cookie processing	451	# cookie processing
292	if (my $jar = $arg{cookie_jar}) {	452	if (my $jar = $arg{cookie_jar}) {
…		…
322	my ($rhost, $rport, $rscheme, $rpath); # request host, port, path	482	my ($rhost, $rport, $rscheme, $rpath); # request host, port, path
323		483
324	if ($proxy) {	484	if ($proxy) {
325	($rpath, $rhost, $rport, $rscheme) = ($url, @$proxy);	485	($rpath, $rhost, $rport, $rscheme) = ($url, @$proxy);
326		486
		487	$rscheme = "http" unless defined $rscheme;
		488
327	# don't support https requests over https-proxy transport,	489	# don't support https requests over https-proxy transport,
328	# can't be done with tls as spec'ed, unless you double-encrypt.	490	# can't be done with tls as spec'ed, unless you double-encrypt.
329	$rscheme = "http" if $uscheme eq "https" && $rscheme eq "https";	491	$rscheme = "http" if $uscheme eq "https" && $rscheme eq "https";
330	} else {	492	} else {
331	($rhost, $rport, $rscheme, $rpath) = ($uhost, $uport, $uscheme, $upath);	493	($rhost, $rport, $rscheme, $rpath) = ($uhost, $uport, $uscheme, $upath);
332	}	494	}
333		495
334	$hdr{host} = $uhost;	496	# leave out fragment and query string, just a heuristic
		497	$hdr{referer} = "$uscheme://$uauthority$upath" unless exists $hdr{referer};
		498	$hdr{"user-agent"} = $USERAGENT unless exists $hdr{"user-agent"};
		499
335	$hdr{"content-length"} = length $arg{body};	500	$hdr{"content-length"} = length $arg{body}
		501	if length $arg{body} \|\| $method ne "GET";
		502
		503	$hdr{connection} = "close TE"; #1.1
		504	$hdr{te} = "trailers" unless exists $hdr{te}; #1.1
336		505
337	my %state = (connect_guard => 1);	506	my %state = (connect_guard => 1);
338		507
339	_get_slot $uhost, sub {	508	_get_slot $uhost, sub {
340	$state{slot_guard} = shift;	509	$state{slot_guard} = shift;
341		510
342	return unless $state{connect_guard};	511	return unless $state{connect_guard};
343		512
344	$state{connect_guard} = AnyEvent::Socket::tcp_connect $rhost, $rport, sub {	513	my $connect_cb = sub {
345	$state{fh} = shift	514	$state{fh} = shift
		515	or do {
		516	my $err = "$!";
		517	%state = ();
346	or return $cb->(undef, { Status => 599, Reason => "$!", URL => $url });	518	return $cb->(undef, { @pseudo, Status => 599, Reason => $err });
		519	};
		520
347	pop; # free memory, save a tree	521	pop; # free memory, save a tree
348		522
349	return unless delete $state{connect_guard};	523	return unless delete $state{connect_guard};
350		524
351	# get handle	525	# get handle
352	$state{handle} = new AnyEvent::Handle	526	$state{handle} = new AnyEvent::Handle
353	fh => $state{fh},	527	fh => $state{fh},
		528	peername => $rhost,
		529	tls_ctx => $arg{tls_ctx},
		530	# these need to be reconfigured on keepalive handles
354	timeout => $timeout;	531	timeout => $timeout,
		532	on_error => sub {
		533	%state = ();
		534	$cb->(undef, { @pseudo, Status => 599, Reason => $_[2] });
		535	},
		536	on_eof => sub {
		537	%state = ();
		538	$cb->(undef, { @pseudo, Status => 599, Reason => "Unexpected end-of-file" });
		539	},
		540	;
355		541
356	# limit the number of persistent connections	542	# limit the number of persistent connections
357	# keepalive not yet supported	543	# keepalive not yet supported
358	if ($KA_COUNT{$_[1]} < $MAX_PERSISTENT_PER_HOST) {	544	# if ($KA_COUNT{$_[1]} < $MAX_PERSISTENT_PER_HOST) {
359	++$KA_COUNT{$_[1]};	545	# ++$KA_COUNT{$_[1]};
360	$state{handle}{ka_count_guard} = AnyEvent::Util::guard {	546	# $state{handle}{ka_count_guard} = AnyEvent::Util::guard {
361	--$KA_COUNT{$_[1]}	547	# --$KA_COUNT{$_[1]}
362	};	548	# };
363	$hdr{connection} = "keep-alive";	549	# $hdr{connection} = "keep-alive";
364	} else {
365	delete $hdr{connection};
366	}	550	# }
367
368	# (re-)configure handle
369	$state{handle}->on_error (sub {
370	my $errno = "$!";
371	%state = ();
372	$cb->(undef, { Status => 599, Reason => $errno, URL => $url });
373	});
374	$state{handle}->on_eof (sub {
375	%state = ();
376	$cb->(undef, { Status => 599, Reason => "unexpected end-of-file", URL => $url });
377	});
378		551
379	$state{handle}->starttls ("connect") if $rscheme eq "https";	552	$state{handle}->starttls ("connect") if $rscheme eq "https";
380		553
381	# handle actual, non-tunneled, request	554	# handle actual, non-tunneled, request
382	my $handle_actual_request = sub {	555	my $handle_actual_request = sub {
383	$state{handle}->starttls ("connect") if $uscheme eq "https" && !exists $state{handle}{tls};	556	$state{handle}->starttls ("connect") if $uscheme eq "https" && !exists $state{handle}{tls};
384		557
385	# send request	558	# send request
386	$state{handle}->push_write (	559	$state{handle}->push_write (
387	"$method $rpath HTTP/1.0\015\012"	560	"$method $rpath HTTP/1.1\015\012"
388	. (join "", map "\u$_: $hdr{$_}\015\012", keys %hdr)	561	. (join "", map "\u$_: $hdr{$_}\015\012", grep defined $hdr{$_}, keys %hdr)
389	. "\015\012"	562	. "\015\012"
390	. (delete $arg{body})	563	. (delete $arg{body})
391	);	564	);
392		565
		566	# return if error occured during push_write()
		567	return unless %state;
		568
393	%hdr = (); # reduce memory usage, save a kitten	569	%hdr = (); # reduce memory usage, save a kitten, also make it possible to re-use
394		570
395	# status line	571	# status line and headers
396	$state{handle}->push_read (line => $qr_nl, sub {	572	$state{read_response} = sub {
397	$_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix
398	or return (%state = (), $cb->(undef, { Status => 599, Reason => "invalid server response ($_[1])", URL => $url }));
399
400	my %hdr = ( # response headers
401	HTTPVersion => ",$1",
402	Status => ",$2",
403	Reason => ",$3",
404	URL => ",$url"
405	);
406
407	# headers, could be optimized a bit
408	$state{handle}->unshift_read (line => $qr_nlnl, sub {
409	for ("$_[1]\012") {	573	for ("$_[1]") {
410	y/\015//d; # weed out any \015, as they show up in the weirdest of places.	574	y/\015//d; # weed out any \015, as they show up in the weirdest of places.
411		575
412	# we support spaces in field names, as lotus domino	576	/^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\012]*) )? \012/igxc
413	# creates them (actually spaces around seperators	577	or return (%state = (), $cb->(undef, { @pseudo, Status => 599, Reason => "Invalid server response" }));
414	# are strictly allowed in http, they are a security issue).
415	$hdr{lc $1} .= ",$2"
416	while /\G
417	([^:\000-\037]+):
418	[\011\040]*
419	((?: [^\012]+ \| \012[\011\040] )*)
420	\012
421	/gxc;
422		578
		579	# 100 Continue handling
		580	# should not happen as we don't send expect: 100-continue,
		581	# but we handle it just in case.
		582	# since we send the request body regardless, if we get an error
		583	# we are out of-sync, which we currently do NOT handle correctly.
		584	return $state{handle}->push_read (line => $qr_nlnl, $state{read_response})
		585	if $2 eq 100;
		586
		587	push @pseudo,
		588	HTTPVersion => $1,
		589	Status => $2,
		590	Reason => $3,
423	/\G$/	591	;
		592
		593	my $hdr = parse_hdr
424	or return (%state = (), $cb->(undef, { Status => 599, Reason => "garbled response headers", URL => $url }));	594	or return (%state = (), $cb->(undef, { @pseudo, Status => 599, Reason => "Garbled response headers" }));
		595
		596	%hdr = (%$hdr, @pseudo);
		597	}
		598
		599	# redirect handling
		600	# microsoft and other shitheads don't give a shit for following standards,
		601	# try to support some common forms of broken Location headers.
		602	if ($hdr{location} !~ /^(?: $ \| [^:\/?\#]+ : )/x) {
		603	$hdr{location} =~ s/^\.\/+//;
		604
		605	my $url = "$rscheme://$uhost:$uport";
		606
		607	unless ($hdr{location} =~ s/^\///) {
		608	$url .= $upath;
		609	$url =~ s/\/[^\/]*$//;
425	}	610	}
426		611
427	substr $_, 0, 1, ""	612	$hdr{location} = "$url/$hdr{location}";
428	for values %hdr;	613	}
429		614
430	my $finish = sub {	615	my $redirect;
		616
		617	if ($recurse) {
		618	my $status = $hdr{Status};
		619
		620	# industry standard is to redirect POST as GET for
		621	# 301, 302 and 303, in contrast to http/1.0 and 1.1.
		622	# also, the UA should ask the user for 301 and 307 and POST,
		623	# industry standard seems to be to simply follow.
		624	# we go with the industry standard.
		625	if ($status == 301 or $status == 302 or $status == 303) {
		626	# HTTP/1.1 is unclear on how to mutate the method
		627	$method = "GET" unless $method eq "HEAD";
		628	$redirect = 1;
		629	} elsif ($status == 307) {
		630	$redirect = 1;
		631	}
		632	}
		633
		634	my $finish = sub { # ($data, $err_status, $err_reason[, $keepalive])
		635	my $keepalive = pop;
		636
431	$state{handle}->destroy;	637	$state{handle}->destroy if $state{handle};
432	%state = ();	638	%state = ();
433		639
		640	if (defined $_[1]) {
		641	$hdr{OrigStatus} = $hdr{Status}; $hdr{Status} = $_[1];
		642	$hdr{OrigReason} = $hdr{Reason}; $hdr{Reason} = $_[2];
		643	}
		644
434	# set-cookie processing	645	# set-cookie processing
435	if ($arg{cookie_jar}) {	646	if ($arg{cookie_jar}) {
436	for ($hdr{"set-cookie"}) {	647	for ($hdr{"set-cookie"}) {
437	# parse NAME=VALUE	648	# parse NAME=VALUE
438	my @kv;	649	my @kv;
439		650
440	while (/\G\s* ([^=;,[:space:]]+) \s=\s (?: "((?:[^\\"]+\|\\.))" \| ([^=;,[:space:]]) )/gcxs) {	651	while (/\G\s* ([^=;,[:space:]]+) \s=\s (?: "((?:[^\\"]+\|\\.))" \| ([^=;,[:space:]]) )/gcxs) {
441	my $name = $1;	652	my $name = $1;
442	my $value = $3;	653	my $value = $3;
443		654
444	unless ($value) {	655	unless ($value) {
445	$value = $2;	656	$value = $2;
446	$value =~ s/\\(.)/$1/gs;	657	$value =~ s/\\(.)/$1/gs;
		658	}
		659
		660	push @kv, $name => $value;
		661
		662	last unless /\G\s*;/gc;
		663	}
		664
		665	last unless @kv;
		666
		667	my $name = shift @kv;
		668	my %kv = (value => shift @kv, @kv);
		669
		670	my $cdom;
		671	my $cpath = (delete $kv{path}) \|\| "/";
		672
		673	if (exists $kv{domain}) {
		674	$cdom = delete $kv{domain};
		675
		676	$cdom =~ s/^\.?/./; # make sure it starts with a "."
		677
		678	next if $cdom =~ /\.$/;
		679
		680	# this is not rfc-like and not netscape-like. go figure.
		681	my $ndots = $cdom =~ y/.//;
		682	next if $ndots < ($cdom =~ /\.[^.][^.]\.[^.][^.]$/ ? 3 : 2);
		683	} else {
		684	$cdom = $uhost;
		685	}
		686
		687	# store it
		688	$arg{cookie_jar}{version} = 1;
		689	$arg{cookie_jar}{$cdom}{$cpath}{$name} = \%kv;
		690
		691	redo if /\G\s*,/gc;
		692	}
		693	}
		694
		695	if ($redirect && exists $hdr{location}) {
		696	# we ignore any errors, as it is very common to receive
		697	# Content-Length != 0 but no actual body
		698	# we also access %hdr, as $_[1] might be an erro
		699	http_request (
		700	$method => $hdr{location},
		701	%arg,
		702	recurse => $recurse - 1,
		703	Redirect => [$_[0], \%hdr],
		704	$cb);
		705	} else {
		706	$cb->($_[0], \%hdr);
		707	}
		708	};
		709
		710	my $len = $hdr{"content-length"};
		711
		712	if (!$redirect && $arg{on_header} && !$arg{on_header}(\%hdr)) {
		713	$finish->(undef, 598 => "Request cancelled by on_header");
		714	} elsif (
		715	$hdr{Status} =~ /^(?:1..\|204\|205\|304)$/
		716	or $method eq "HEAD"
		717	or (defined $len && !$len)
		718	) {
		719	# no body
		720	$finish->("", undef, undef, 1);
		721	} else {
		722	# body handling, many different code paths
		723	# - no body expected
		724	# - want_body_handle
		725	# - te chunked
		726	# - 2x length known (with or without on_body)
		727	# - 2x length not known (with or without on_body)
		728	if (!$redirect && $arg{want_body_handle}) {
		729	$_[0]->on_eof (undef);
		730	$_[0]->on_error (undef);
		731	$_[0]->on_read (undef);
		732
		733	$finish->(delete $state{handle});
		734
		735	} elsif ($hdr{"transfer-encoding"} =~ /\bchunked\b/i) {
		736	my $cl = 0;
		737	my $body = undef;
		738	my $on_body = $arg{on_body} \|\| sub { $body .= shift; 1 };
		739
		740	$_[0]->on_error (sub { $finish->(undef, 599 => $_[2]) });
		741
		742	my $read_chunk; $read_chunk = sub {
		743	$_[1] =~ /^([0-9a-fA-F]+)/
		744	or $finish->(undef, 599 => "Garbled chunked transfer encoding");
		745
		746	my $len = hex $1;
		747
		748	if ($len) {
		749	$cl += $len;
		750
		751	$_[0]->push_read (chunk => $len, sub {
		752	$on_body->($_[1], \%hdr)
		753	or return $finish->(undef, 598 => "Request cancelled by on_body");
		754
		755	$_[0]->push_read (line => sub {
		756	length $_[1]
		757	and return $finish->(undef, 599 => "Garbled chunked transfer encoding");
		758	$_[0]->push_read (line => $read_chunk);
		759	});
		760	});
		761	} else {
		762	$hdr{"content-length"} \|\|= $cl;
		763
		764	$_[0]->push_read (line => $qr_nlnl, sub {
		765	if (length $_[1]) {
		766	for ("$_[1]") {
		767	y/\015//d; # weed out any \015, as they show up in the weirdest of places.
		768
		769	my $hdr = parse_hdr
		770	or return $finish->(undef, 599 => "Garbled response trailers");
		771
		772	%hdr = (%hdr, %$hdr);
		773	}
447	}	774	}
448		775
449	push @kv, $name => $value;	776	$finish->($body, undef, undef, 1);
450
451	last unless /\G\s*;/gc;
452	}	777	});
453
454	last unless @kv;
455
456	my $name = shift @kv;
457	my %kv = (value => shift @kv, @kv);
458
459	my $cdom;
460	my $cpath = (delete $kv{path}) \|\| "/";
461
462	if (exists $kv{domain}) {
463	$cdom = delete $kv{domain};
464
465	$cdom =~ s/^\.?/./; # make sure it starts with a "."
466
467	next if $cdom =~ /\.$/;
468
469	# this is not rfc-like and not netscape-like. go figure.
470	my $ndots = $cdom =~ y/.//;
471	next if $ndots < ($cdom =~ /\.[^.][^.]\.[^.][^.]$/ ? 3 : 2);
472	} else {
473	$cdom = $uhost;
474	}
475
476	# store it
477	$arg{cookie_jar}{version} = 1;
478	$arg{cookie_jar}{$cdom}{$cpath}{$name} = \%kv;
479
480	redo if /\G\s*,/gc;
481	}	778	}
482	}	779	};
483		780
484	# microsoft and other shitheads don't give a shit for following standards,	781	$_[0]->push_read (line => $read_chunk);
485	# try to support some common forms of broken Location headers.
486	if ($_[1]{location} !~ /^(?: $ \| [^:\/?\#]+ : )/x) {
487	$_[1]{location} =~ s/^\.\/+//;
488		782
489	my $url = "$rscheme://$uhost:$uport";	783	} elsif ($arg{on_body}) {
		784	$_[0]->on_error (sub { $finish->(undef, 599 => $_[2]) });
490		785
491	unless ($_[1]{location} =~ s/^\///) {
492	$url .= $upath;
493	$url =~ s/\/[^\/]*$//;
494	}
495
496	$_[1]{location} = "$url/$_[1]{location}";
497	}
498
499	if ($_[1]{Status} =~ /^30[12]$/ && $recurse && $method ne "POST") {
500	# apparently, mozilla et al. just change POST to GET here
501	# more research is needed before we do the same
502	http_request ($method, $_[1]{location}, %arg, recurse => $recurse - 1, $cb);
503	} elsif ($_[1]{Status} == 303 && $recurse) {
504	# even http/1.1 is unclear on how to mutate the method
505	$method = "GET" unless $method eq "HEAD";
506	http_request ($method => $_[1]{location}, %arg, recurse => $recurse - 1, $cb);
507	} elsif ($_[1]{Status} == 307 && $recurse && $method =~ /^(?:GET\|HEAD)$/) {
508	http_request ($method => $_[1]{location}, %arg, recurse => $recurse - 1, $cb);
509	} else {	786	if ($len) {
510	$cb->($_[0], $_[1]);	787	$_[0]->on_read (sub {
511	}	788	$len -= length $_[0]{rbuf};
512	};
513		789
514	if ($hdr{Status} =~ /^(?:1..\|204\|304)$/ or $method eq "HEAD") {	790	$arg{on_body}(delete $_[0]{rbuf}, \%hdr)
515	$finish->(undef, \%hdr);	791	or return $finish->(undef, 598 => "Request cancelled by on_body");
516	} else {	792
517	if (exists $hdr{"content-length"}) {
518	$_[0]->unshift_read (chunk => $hdr{"content-length"}, sub {
519	# could cache persistent connection now
520	if ($hdr{connection} =~ /\bkeep-alive\b/i) {
521	# but we don't, due to misdesigns, this is annoyingly complex
522	};	793	$len > 0
523		794	or $finish->("", undef, undef, 1);
524	$finish->($_[1], \%hdr);
525	});	795	});
526	} else {	796	} else {
527	# too bad, need to read until we get an error or EOF,	797	$_[0]->on_eof (sub {
528	# no way to detect winged data.	798	$finish->("");
		799	});
		800	$_[0]->on_read (sub {
		801	$arg{on_body}(delete $_[0]{rbuf}, \%hdr)
		802	or $finish->(undef, 598 => "Request cancelled by on_body");
		803	});
		804	}
		805	} else {
		806	$_[0]->on_eof (undef);
		807
		808	if ($len) {
		809	$_[0]->on_error (sub { $finish->(undef, 599 => $_[2]) });
		810	$_[0]->on_read (sub {
		811	$finish->((substr delete $_[0]{rbuf}, 0, $len, ""), undef, undef, 1)
		812	if $len <= length $_[0]{rbuf};
		813	});
		814	} else {
529	$_[0]->on_error (sub {	815	$_[0]->on_error (sub {
530	# delete ought to be more efficient, as we would have to make	816	($! == Errno::EPIPE \|\| !$!)
531	# a copy otherwise as $_[0] gets destroyed.
532	$finish->(delete $_[0]{rbuf}, \%hdr);	817	? $finish->(delete $_[0]{rbuf})
		818	: $finish->(undef, 599 => $_[2]);
533	});	819	});
534	$_[0]->on_eof (undef);
535	$_[0]->on_read (sub { });	820	$_[0]->on_read (sub { });
536	}	821	}
537	}	822	}
538	});	823	}
539	});	824	};
		825
		826	$state{handle}->push_read (line => $qr_nlnl, $state{read_response});
540	};	827	};
541		828
542	# now handle proxy-CONNECT method	829	# now handle proxy-CONNECT method
543	if ($proxy && $uscheme eq "https") {	830	if ($proxy && $uscheme eq "https") {
544	# oh dear, we have to wrap it into a connect request	831	# oh dear, we have to wrap it into a connect request
545		832
546	# maybe re-use $uauthority with patched port?	833	# maybe re-use $uauthority with patched port?
547	$state{handle}->push_write ("CONNECT $uhost:$uport HTTP/1.0\015\012Host: $uhost\015\012\015\012");	834	$state{handle}->push_write ("CONNECT $uhost:$uport HTTP/1.0\015\012Host: $uhost\015\012\015\012");
548	$state{handle}->push_read (line => $qr_nlnl, sub {	835	$state{handle}->push_read (line => $qr_nlnl, sub {
549	$_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix	836	$_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix
550	or return (%state = (), $cb->(undef, { Status => 599, Reason => "invalid proxy connect response ($_[1])", URL => $url }));	837	or return (%state = (), $cb->(undef, { @pseudo, Status => 599, Reason => "Invalid proxy connect response ($_[1])" }));
551		838
552	if ($2 == 200) {	839	if ($2 == 200) {
553	$rpath = $upath;	840	$rpath = $upath;
554	&$handle_actual_request;	841	&$handle_actual_request;
555	} else {	842	} else {
556	%state = ();	843	%state = ();
557	$cb->(undef, { Status => $2, Reason => $3, URL => $url });	844	$cb->(undef, { @pseudo, Status => $2, Reason => $3 });
558	}	845	}
559	});	846	});
560	} else {	847	} else {
561	&$handle_actual_request;	848	&$handle_actual_request;
562	}	849	}
563
564	}, sub {
565	$timeout
566	};	850	};
		851
		852	my $tcp_connect = $arg{tcp_connect}
		853	\|\| do { require AnyEvent::Socket; \&AnyEvent::Socket::tcp_connect };
		854
		855	$state{connect_guard} = $tcp_connect->($rhost, $rport, $connect_cb, $arg{on_prepare} \|\| sub { $timeout });
		856
567	};	857	};
568		858
569	defined wantarray && AnyEvent::Util::guard { %state = () }	859	defined wantarray && AnyEvent::Util::guard { %state = () }
570	}	860	}
571		861
…		…
585	&http_request	875	&http_request
586	}	876	}
587		877
588	=back	878	=back
589		879
		880	=head2 DNS CACHING
		881
		882	AnyEvent::HTTP uses the AnyEvent::Socket::tcp_connect function for
		883	the actual connection, which in turn uses AnyEvent::DNS to resolve
		884	hostnames. The latter is a simple stub resolver and does no caching
		885	on its own. If you want DNS caching, you currently have to provide
		886	your own default resolver (by storing a suitable resolver object in
		887	C<$AnyEvent::DNS::RESOLVER>).
		888
590	=head2 GLOBAL FUNCTIONS AND VARIABLES	889	=head2 GLOBAL FUNCTIONS AND VARIABLES
591		890
592	=over 4	891	=over 4
593		892
594	=item AnyEvent::HTTP::set_proxy "proxy-url"	893	=item AnyEvent::HTTP::set_proxy "proxy-url"
595		894
596	Sets the default proxy server to use. The proxy-url must begin with a	895	Sets the default proxy server to use. The proxy-url must begin with a
597	string of the form C<http://host:port> (optionally C<https:...>).	896	string of the form C<http://host:port> (optionally C<https:...>), croaks
		897	otherwise.
		898
		899	To clear an already-set proxy, use C<undef>.
		900
		901	=item $date = AnyEvent::HTTP::format_date $timestamp
		902
		903	Takes a POSIX timestamp (seconds since the epoch) and formats it as a HTTP
		904	Date (RFC 2616).
		905
		906	=item $timestamp = AnyEvent::HTTP::parse_date $date
		907
		908	Takes a HTTP Date (RFC 2616) and returns the corresponding POSIX
		909	timestamp, or C<undef> if the date cannot be parsed.
598		910
599	=item $AnyEvent::HTTP::MAX_RECURSE	911	=item $AnyEvent::HTTP::MAX_RECURSE
600		912
601	The default value for the C<recurse> request parameter (default: C<10>).	913	The default value for the C<recurse> request parameter (default: C<10>).
602		914
603	=item $AnyEvent::HTTP::USERAGENT	915	=item $AnyEvent::HTTP::USERAGENT
604		916
605	The default value for the C<User-Agent> header (the default is	917	The default value for the C<User-Agent> header (the default is
606	C<Mozilla/5.0 (compatible; AnyEvent::HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)>).	918	C<Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)>).
607		919
608	=item $AnyEvent::HTTP::MAX_PERSISTENT	920	=item $AnyEvent::HTTP::MAX_PER_HOST
609		921
610	The maximum number of persistent connections to keep open (default: 8).	922	The maximum number of concurrent connections to the same host (identified
		923	by the hostname). If the limit is exceeded, then the additional requests
		924	are queued until previous connections are closed.
611		925
612	Not implemented currently.	926	The default value for this is C<4>, and it is highly advisable to not
613		927	increase it.
614	=item $AnyEvent::HTTP::PERSISTENT_TIMEOUT
615
616	The maximum time to cache a persistent connection, in seconds (default: 2).
617
618	Not implemented currently.
619		928
620	=item $AnyEvent::HTTP::ACTIVE	929	=item $AnyEvent::HTTP::ACTIVE
621		930
622	The number of active connections. This is not the number of currently	931	The number of active connections. This is not the number of currently
623	running requests, but the number of currently open and non-idle TCP	932	running requests, but the number of currently open and non-idle TCP
…		…
625		934
626	=back	935	=back
627		936
628	=cut	937	=cut
629		938
		939	our @month = qw(Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec);
		940	our @weekday = qw(Sun Mon Tue Wed Thu Fri Sat);
		941
		942	sub format_date($) {
		943	my ($time) = @_;
		944
		945	# RFC 822/1123 format
		946	my ($S, $M, $H, $mday, $mon, $year, $wday, $yday, undef) = gmtime $time;
		947
		948	sprintf "%s, %02d %s %04d %02d:%02d:%02d GMT",
		949	$weekday[$wday], $mday, $month[$mon], $year + 1900,
		950	$H, $M, $S;
		951	}
		952
		953	sub parse_date($) {
		954	my ($date) = @_;
		955
		956	my ($d, $m, $y, $H, $M, $S);
		957
		958	if ($date =~ /^[A-Z][a-z][a-z], ([0-9][0-9]) ([A-Z][a-z][a-z]) ([0-9][0-9][0-9][0-9]) ([0-9][0-9]):([0-9][0-9]):([0-9][0-9]) GMT$/) {
		959	# RFC 822/1123, required by RFC 2616
		960	($d, $m, $y, $H, $M, $S) = ($1, $2, $3, $4, $5, $6);
		961
		962	} elsif ($date =~ /^[A-Z][a-z]+, ([0-9][0-9])-([A-Z][a-z][a-z])-([0-9][0-9]) ([0-9][0-9]):([0-9][0-9]):([0-9][0-9]) GMT$/) {
		963	# RFC 850
		964	($d, $m, $y, $H, $M, $S) = ($1, $2, $3 < 69 ? $3 + 2000 : $3 + 1900, $4, $5, $6);
		965
		966	} elsif ($date =~ /^[A-Z][a-z][a-z] ([A-Z][a-z][a-z]) ([0-9 ][0-9]) ([0-9][0-9]):([0-9][0-9]):([0-9][0-9]) ([0-9][0-9][0-9][0-9])$/) {
		967	# ISO C's asctime
		968	($d, $m, $y, $H, $M, $S) = ($2, $1, $6, $3, $4, $5);
		969	}
		970	# other formats fail in the loop below
		971
		972	for (0..11) {
		973	if ($m eq $month[$_]) {
		974	require Time::Local;
		975	return Time::Local::timegm ($S, $M, $H, $d, $_, $y);
		976	}
		977	}
		978
		979	undef
		980	}
		981
630	sub set_proxy($) {	982	sub set_proxy($) {
		983	if (length $_[0]) {
631	$PROXY = [$2, $3 \|\| 3128, $1] if $_[0] =~ m%^(https?):// ([^:/]+) (?: : (\d*) )?%ix;	984	$_[0] =~ m%^(https?):// ([^:/]+) (?: : (\d*) )?%ix
		985	or Carp::croak "$_[0]: invalid proxy URL";
		986	$PROXY = [$2, $3 \|\| 3128, $1]
		987	} else {
		988	undef $PROXY;
		989	}
632	}	990	}
633		991
634	# initialise proxy from environment	992	# initialise proxy from environment
		993	eval {
635	set_proxy $ENV{http_proxy};	994	set_proxy $ENV{http_proxy};
		995	};
		996
		997	=head2 SOCKS PROXIES
		998
		999	Socks proxies are not directly supported by AnyEvent::HTTP. You can
		1000	compile your perl to support socks, or use an external program such as
		1001	F<socksify> (dante) or F<tsocks> to make your program use a socks proxy
		1002	transparently.
		1003
		1004	Alternatively, for AnyEvent::HTTP only, you can use your own
		1005	C<tcp_connect> function that does the proxy handshake - here is an example
		1006	that works with socks4a proxies:
		1007
		1008	use Errno;
		1009	use AnyEvent::Util;
		1010	use AnyEvent::Socket;
		1011	use AnyEvent::Handle;
		1012
		1013	# host, port and username of/for your socks4a proxy
		1014	my $socks_host = "10.0.0.23";
		1015	my $socks_port = 9050;
		1016	my $socks_user = "";
		1017
		1018	sub socks4a_connect {
		1019	my ($host, $port, $connect_cb, $prepare_cb) = @_;
		1020
		1021	my $hdl = new AnyEvent::Handle
		1022	connect => [$socks_host, $socks_port],
		1023	on_prepare => sub { $prepare_cb->($_[0]{fh}) },
		1024	on_error => sub { $connect_cb->() },
		1025	;
		1026
		1027	$hdl->push_write (pack "CCnNZZ", 4, 1, $port, 1, $socks_user, $host);
		1028
		1029	$hdl->push_read (chunk => 8, sub {
		1030	my ($hdl, $chunk) = @_;
		1031	my ($status, $port, $ipn) = unpack "xCna4", $chunk;
		1032
		1033	if ($status == 0x5a) {
		1034	$connect_cb->($hdl->{fh}, (format_address $ipn) . ":$port");
		1035	} else {
		1036	$! = Errno::ENXIO; $connect_cb->();
		1037	}
		1038	});
		1039
		1040	$hdl
		1041	}
		1042
		1043	Use C<socks4a_connect> instead of C<tcp_connect> when doing C<http_request>s,
		1044	possibly after switching off other proxy types:
		1045
		1046	AnyEvent::HTTP::set_proxy undef; # usually you do not want other proxies
		1047
		1048	http_get 'http://www.google.com', tcp_connect => \&socks4a_connect, sub {
		1049	my ($data, $headers) = @_;
		1050	...
		1051	};
636		1052
637	=head1 SEE ALSO	1053	=head1 SEE ALSO
638		1054
639	L<AnyEvent>.	1055	L<AnyEvent>.
640		1056

Diff Legend

-–
+Removed lines
-+
+Added lines
-<
+Changed lines
->
+Changed lines

Comparing AnyEvent-HTTP/HTTP.pm (file contents): Revision 1.39 by root, Thu Jun 11 12:46:01 2009 UTC vs. Revision 1.69 by root, Fri Dec 31 19:32:47 2010 UTC

Diff Legend

Comparing AnyEvent-HTTP/HTTP.pm (file contents):
Revision 1.39 by root, Thu Jun 11 12:46:01 2009 UTC vs.
Revision 1.69 by root, Fri Dec 31 19:32:47 2010 UTC