ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/AnyEvent-HTTP/HTTP.pm

Comparing AnyEvent-HTTP/HTTP.pm (file contents):
Revision 1.56 by root, Mon Sep 6 05:30:54 2010 UTC vs.
Revision 1.133 by root, Sat Sep 1 03:16:43 2018 UTC

…		…
15	This module is an L<AnyEvent> user, you need to make sure that you use and	15	This module is an L<AnyEvent> user, you need to make sure that you use and
16	run a supported event loop.	16	run a supported event loop.
17		17
18	This module implements a simple, stateless and non-blocking HTTP	18	This module implements a simple, stateless and non-blocking HTTP
19	client. It supports GET, POST and other request methods, cookies and more,	19	client. It supports GET, POST and other request methods, cookies and more,
20	all on a very low level. It can follow redirects supports proxies and	20	all on a very low level. It can follow redirects, supports proxies, and
21	automatically limits the number of connections to the values specified in	21	automatically limits the number of connections to the values specified in
22	the RFC.	22	the RFC.
23		23
24	It should generally be a "good client" that is enough for most HTTP	24	It should generally be a "good client" that is enough for most HTTP
25	tasks. Simple tasks should be simple, but complex tasks should still be	25	tasks. Simple tasks should be simple, but complex tasks should still be
…		…
36		36
37	=cut	37	=cut
38		38
39	package AnyEvent::HTTP;	39	package AnyEvent::HTTP;
40		40
41	use strict;	41	use common::sense;
42	no warnings;
43		42
44	use Errno ();	43	use Errno ();
45		44
46	use AnyEvent 5.0 ();	45	use AnyEvent 5.0 ();
47	use AnyEvent::Util ();	46	use AnyEvent::Util ();
48	use AnyEvent::Socket ();
49	use AnyEvent::Handle ();	47	use AnyEvent::Handle ();
50		48
51	use base Exporter::;	49	use base Exporter::;
52		50
53	our $VERSION = '1.45';	51	our $VERSION = 2.24;
54		52
55	our @EXPORT = qw(http_get http_post http_head http_request);	53	our @EXPORT = qw(http_get http_post http_head http_request);
56		54
57	our $USERAGENT = "Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)";	55	our $USERAGENT = "Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)";
58	our $MAX_RECURSE = 10;	56	our $MAX_RECURSE = 10;
59	our $MAX_PERSISTENT = 8;
60	our $PERSISTENT_TIMEOUT = 2;	57	our $PERSISTENT_TIMEOUT = 3;
61	our $TIMEOUT = 300;	58	our $TIMEOUT = 300;
62		59	our $MAX_PER_HOST = 4; # changing this is evil
63	# changing these is evil
64	our $MAX_PERSISTENT_PER_HOST = 0;
65	our $MAX_PER_HOST = 4;
66		60
67	our $PROXY;	61	our $PROXY;
68	our $ACTIVE = 0;	62	our $ACTIVE = 0;
69		63
70	my %KA_COUNT; # number of open keep-alive connections per host	64	my %KA_CACHE; # indexed by uhost currently, points to [$handle...] array
71	my %CO_SLOT; # number of open connections, and wait queue, per host	65	my %CO_SLOT; # number of open connections, and wait queue, per host
72		66
73	=item http_get $url, key => value..., $cb->($data, $headers)	67	=item http_get $url, key => value..., $cb->($data, $headers)
74		68
75	Executes an HTTP-GET request. See the http_request function for details on	69	Executes an HTTP-GET request. See the http_request function for details on
…		…
92	must be an absolute http or https URL.	86	must be an absolute http or https URL.
93		87
94	When called in void context, nothing is returned. In other contexts,	88	When called in void context, nothing is returned. In other contexts,
95	C<http_request> returns a "cancellation guard" - you have to keep the	89	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	90	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.	91	destroyed before the callback is called, the request will be cancelled.
98		92
99	The callback will be called with the response body data as first argument	93	The callback will be called with the response body data as first argument
100	(or C<undef> if an error occured), and a hash-ref with response headers as	94	(or C<undef> if an error occurred), and a hash-ref with response headers
101	second argument.	95	(and trailers) as second argument.
102		96
103	All the headers in that hash are lowercased. In addition to the response	97	All the headers in that hash are lowercased. In addition to the response
104	headers, the "pseudo-headers" (uppercase to avoid clashing with possible	98	headers, the "pseudo-headers" (uppercase to avoid clashing with possible
105	response headers) C<HTTPVersion>, C<Status> and C<Reason> contain the	99	response headers) C<HTTPVersion>, C<Status> and C<Reason> contain the
106	three parts of the HTTP Status-Line of the same name.	100	three parts of the HTTP Status-Line of the same name. If an error occurs
		101	during the body phase of a request, then the original C<Status> and
		102	C<Reason> values from the header are available as C<OrigStatus> and
		103	C<OrigReason>.
107		104
108	The pseudo-header C<URL> contains the actual URL (which can differ from	105	The pseudo-header C<URL> contains the actual URL (which can differ from
109	the requested URL when following redirects - for example, you might get	106	the requested URL when following redirects - for example, you might get
110	an error that your URL scheme is not supported even though your URL is a	107	an error that your URL scheme is not supported even though your URL is a
111	valid http URL because it redirected to an ftp URL, in which case you can	108	valid http URL because it redirected to an ftp URL, in which case you can
…		…
120		117
121	If the server sends a header multiple times, then their contents will be	118	If the server sends a header multiple times, then their contents will be
122	joined together with a comma (C<,>), as per the HTTP spec.	119	joined together with a comma (C<,>), as per the HTTP spec.
123		120
124	If an internal error occurs, such as not being able to resolve a hostname,	121	If an internal error occurs, such as not being able to resolve a hostname,
125	then C<$data> will be C<undef>, C<< $headers->{Status} >> will be C<59x>	122	then C<$data> will be C<undef>, C<< $headers->{Status} >> will be
126	(usually C<599>) and the C<Reason> pseudo-header will contain an error	123	C<590>-C<599> and the C<Reason> pseudo-header will contain an error
127	message.	124	message. Currently the following status codes are used:
		125
		126	=over 4
		127
		128	=item 595 - errors during connection establishment, proxy handshake.
		129
		130	=item 596 - errors during TLS negotiation, request sending and header processing.
		131
		132	=item 597 - errors during body receiving or processing.
		133
		134	=item 598 - user aborted request via C<on_header> or C<on_body>.
		135
		136	=item 599 - other, usually nonretryable, errors (garbled URL etc.).
		137
		138	=back
128		139
129	A typical callback might look like this:	140	A typical callback might look like this:
130		141
131	sub {	142	sub {
132	my ($body, $hdr) = @_;	143	my ($body, $hdr) = @_;
…		…
143		154
144	=over 4	155	=over 4
145		156
146	=item recurse => $count (default: $MAX_RECURSE)	157	=item recurse => $count (default: $MAX_RECURSE)
147		158
148	Whether to recurse requests or not, e.g. on redirects, authentication	159	Whether to recurse requests or not, e.g. on redirects, authentication and
149	retries and so on, and how often to do so.	160	other retries and so on, and how often to do so.
		161
		162	Only redirects to http and https URLs are supported. While most common
		163	redirection forms are handled entirely within this module, some require
		164	the use of the optional L<URI> module. If it is required but missing, then
		165	the request will fail with an error.
150		166
151	=item headers => hashref	167	=item headers => hashref
152		168
153	The request headers to use. Currently, C<http_request> may provide its	169	The request headers to use. Currently, C<http_request> may provide its own
154	own C<Host:>, C<Content-Length:>, C<Connection:> and C<Cookie:> headers	170	C<Host:>, C<Content-Length:>, C<Connection:> and C<Cookie:> headers and
155	and will provide defaults for C<User-Agent:> and C<Referer:> (this can be	171	will provide defaults at least for C<TE:>, C<Referer:> and C<User-Agent:>
156	suppressed by using C<undef> for these headers in which case they won't be	172	(this can be suppressed by using C<undef> for these headers in which case
157	sent at all).	173	they won't be sent at all).
		174
		175	You really should provide your own C<User-Agent:> header value that is
		176	appropriate for your program - I wouldn't be surprised if the default
		177	AnyEvent string gets blocked by webservers sooner or later.
		178
		179	Also, make sure that your headers names and values do not contain any
		180	embedded newlines.
158		181
159	=item timeout => $seconds	182	=item timeout => $seconds
160		183
161	The time-out to use for various stages - each connect attempt will reset	184	The time-out to use for various stages - each connect attempt will reset
162	the timeout, as will read or write activity, i.e. this is not an overall	185	the timeout, as will read or write activity, i.e. this is not an overall
…		…
164		187
165	Default timeout is 5 minutes.	188	Default timeout is 5 minutes.
166		189
167	=item proxy => [$host, $port[, $scheme]] or undef	190	=item proxy => [$host, $port[, $scheme]] or undef
168		191
169	Use the given http proxy for all requests. If not specified, then the	192	Use the given http proxy for all requests, or no proxy if C<undef> is
170	default proxy (as specified by C<$ENV{http_proxy}>) is used.	193	used.
171		194
172	C<$scheme> must be either missing, C<http> for HTTP or C<https> for	195	C<$scheme> must be either missing or must be C<http> for HTTP.
173	HTTPS.	196
		197	If not specified, then the default proxy is used (see
		198	C<AnyEvent::HTTP::set_proxy>).
		199
		200	Currently, if your proxy requires authorization, you have to specify an
		201	appropriate "Proxy-Authorization" header in every request.
174		202
175	=item body => $string	203	=item body => $string
176		204
177	The request body, usually empty. Will be-sent as-is (future versions of	205	The request body, usually empty. Will be sent as-is (future versions of
178	this module might offer more options).	206	this module might offer more options).
179		207
180	=item cookie_jar => $hash_ref	208	=item cookie_jar => $hash_ref
181		209
182	Passing this parameter enables (simplified) cookie-processing, loosely	210	Passing this parameter enables (simplified) cookie-processing, loosely
183	based on the original netscape specification.	211	based on the original netscape specification.
184		212
185	The C<$hash_ref> must be an (initially empty) hash reference which will	213	The C<$hash_ref> must be an (initially empty) hash reference which
186	get updated automatically. It is possible to save the cookie_jar to	214	will get updated automatically. It is possible to save the cookie jar
187	persistent storage with something like JSON or Storable, but this is not	215	to persistent storage with something like JSON or Storable - see the
188	recommended, as expiry times are currently being ignored.	216	C<AnyEvent::HTTP::cookie_jar_expire> function if you wish to remove
		217	expired or session-only cookies, and also for documentation on the format
		218	of the cookie jar.
189		219
190	Note that this cookie implementation is not of very high quality, nor	220	Note that this cookie implementation is not meant to be complete. If
191	meant to be complete. If you want complete cookie management you have to	221	you want complete cookie management you have to do that on your
192	do that on your own. C<cookie_jar> is meant as a quick fix to get some	222	own. C<cookie_jar> is meant as a quick fix to get most cookie-using sites
193	cookie-using sites working. Cookies are a privacy disaster, do not use	223	working. Cookies are a privacy disaster, do not use them unless required
194	them unless required to.	224	to.
		225
		226	When cookie processing is enabled, the C<Cookie:> and C<Set-Cookie:>
		227	headers will be set and handled by this module, otherwise they will be
		228	left untouched.
195		229
196	=item tls_ctx => $scheme | $tls_ctx	230	=item tls_ctx => $scheme | $tls_ctx
197		231
198	Specifies the AnyEvent::TLS context to be used for https connections. This	232	Specifies the AnyEvent::TLS context to be used for https connections. This
199	parameter follows the same rules as the C<tls_ctx> parameter to	233	parameter follows the same rules as the C<tls_ctx> parameter to
…		…
203	verification) TLS context.	237	verification) TLS context.
204		238
205	The default for this option is C<low>, which could be interpreted as "give	239	The default for this option is C<low>, which could be interpreted as "give
206	me the page, no matter what".	240	me the page, no matter what".
207		241
		242	See also the C<sessionid> parameter.
		243
		244	=item session => $string
		245
		246	The module might reuse connections to the same host internally. Sometimes
		247	(e.g. when using TLS), you do not want to reuse connections from other
		248	sessions. This can be achieved by setting this parameter to some unique
		249	ID (such as the address of an object storing your state data, or the TLS
		250	context) - only connections using the same unique ID will be reused.
		251
208	=item on_prepare => $callback->($fh)	252	=item on_prepare => $callback->($fh)
209		253
210	In rare cases you need to "tune" the socket before it is used to	254	In rare cases you need to "tune" the socket before it is used to
211	connect (for exmaple, to bind it on a given IP address). This parameter	255	connect (for example, to bind it on a given IP address). This parameter
212	overrides the prepare callback passed to C<AnyEvent::Socket::tcp_connect>	256	overrides the prepare callback passed to C<AnyEvent::Socket::tcp_connect>
213	and behaves exactly the same way (e.g. it has to provide a	257	and behaves exactly the same way (e.g. it has to provide a
214	timeout). See the description for the C<$prepare_cb> argument of	258	timeout). See the description for the C<$prepare_cb> argument of
215	C<AnyEvent::Socket::tcp_connect> for details.	259	C<AnyEvent::Socket::tcp_connect> for details.
216		260
		261	=item tcp_connect => $callback->($host, $service, $connect_cb, $prepare_cb)
		262
		263	In even rarer cases you want total control over how AnyEvent::HTTP
		264	establishes connections. Normally it uses L<AnyEvent::Socket::tcp_connect>
		265	to do this, but you can provide your own C<tcp_connect> function -
		266	obviously, it has to follow the same calling conventions, except that it
		267	may always return a connection guard object.
		268
		269	There are probably lots of weird uses for this function, starting from
		270	tracing the hosts C<http_request> actually tries to connect, to (inexact
		271	but fast) host => IP address caching or even socks protocol support.
		272
217	=item on_header => $callback->($headers)	273	=item on_header => $callback->($headers)
218		274
219	When specified, this callback will be called with the header hash as soon	275	When specified, this callback will be called with the header hash as soon
220	as headers have been successfully received from the remote server (not on	276	as headers have been successfully received from the remote server (not on
221	locally-generated errors).	277	locally-generated errors).
…		…
226		282
227	This callback is useful, among other things, to quickly reject unwanted	283	This callback is useful, among other things, to quickly reject unwanted
228	content, which, if it is supposed to be rare, can be faster than first	284	content, which, if it is supposed to be rare, can be faster than first
229	doing a C<HEAD> request.	285	doing a C<HEAD> request.
230		286
		287	The downside is that cancelling the request makes it impossible to re-use
		288	the connection. Also, the C<on_header> callback will not receive any
		289	trailer (headers sent after the response body).
		290
231	Example: cancel the request unless the content-type is "text/html".	291	Example: cancel the request unless the content-type is "text/html".
232		292
233	on_header => sub {	293	on_header => sub {
234	$_[0]{"content-type"} =~ /^text\/html\s*(?:;|$)/	294	$_[0]{"content-type"} =~ /^text\/html\s*(?:;|$)/
235	},	295	},
…		…
241	string instead of the body data.	301	string instead of the body data.
242		302
243	It has to return either true (in which case AnyEvent::HTTP will continue),	303	It has to return either true (in which case AnyEvent::HTTP will continue),
244	or false, in which case AnyEvent::HTTP will cancel the download (and call	304	or false, in which case AnyEvent::HTTP will cancel the download (and call
245	the completion callback with an error code of C<598>).	305	the completion callback with an error code of C<598>).
		306
		307	The downside to cancelling the request is that it makes it impossible to
		308	re-use the connection.
246		309
247	This callback is useful when the data is too large to be held in memory	310	This callback is useful when the data is too large to be held in memory
248	(so the callback writes it to a file) or when only some information should	311	(so the callback writes it to a file) or when only some information should
249	be extracted, or when the body should be processed incrementally.	312	be extracted, or when the body should be processed incrementally.
250		313
…		…
262	called. Instead of the C<$body> argument containing the body data, the	325	called. Instead of the C<$body> argument containing the body data, the
263	callback will receive the L<AnyEvent::Handle> object associated with the	326	callback will receive the L<AnyEvent::Handle> object associated with the
264	connection. In error cases, C<undef> will be passed. When there is no body	327	connection. In error cases, C<undef> will be passed. When there is no body
265	(e.g. status C<304>), the empty string will be passed.	328	(e.g. status C<304>), the empty string will be passed.
266		329
267	The handle object might or might not be in TLS mode, might be connected to	330	The handle object might or might not be in TLS mode, might be connected
268	a proxy, be a persistent connection etc., and configured in unspecified	331	to a proxy, be a persistent connection, use chunked transfer encoding
269	ways. The user is responsible for this handle (it will not be used by this	332	etc., and configured in unspecified ways. The user is responsible for this
270	module anymore).	333	handle (it will not be used by this module anymore).
271		334
272	This is useful with some push-type services, where, after the initial	335	This is useful with some push-type services, where, after the initial
273	headers, an interactive protocol is used (typical example would be the	336	headers, an interactive protocol is used (typical example would be the
274	push-style twitter API which starts a JSON/XML stream).	337	push-style twitter API which starts a JSON/XML stream).
275		338
276	If you think you need this, first have a look at C<on_body>, to see if	339	If you think you need this, first have a look at C<on_body>, to see if
277	that doesn't solve your problem in a better way.	340	that doesn't solve your problem in a better way.
278		341
		342	=item persistent => $boolean
		343
		344	Try to create/reuse a persistent connection. When this flag is set
		345	(default: true for idempotent requests, false for all others), then
		346	C<http_request> tries to re-use an existing (previously-created)
		347	persistent connection to same host (i.e. identical URL scheme, hostname,
		348	port and session) and, failing that, tries to create a new one.
		349
		350	Requests failing in certain ways will be automatically retried once, which
		351	is dangerous for non-idempotent requests, which is why it defaults to off
		352	for them. The reason for this is because the bozos who designed HTTP/1.1
		353	made it impossible to distinguish between a fatal error and a normal
		354	connection timeout, so you never know whether there was a problem with
		355	your request or not.
		356
		357	When reusing an existent connection, many parameters (such as TLS context)
		358	will be ignored. See the C<session> parameter for a workaround.
		359
		360	=item keepalive => $boolean
		361
		362	Only used when C<persistent> is also true. This parameter decides whether
		363	C<http_request> tries to handshake a HTTP/1.0-style keep-alive connection
		364	(as opposed to only a HTTP/1.1 persistent connection).
		365
		366	The default is true, except when using a proxy, in which case it defaults
		367	to false, as HTTP/1.0 proxies cannot support this in a meaningful way.
		368
		369	=item handle_params => { key => value ... }
		370
		371	The key-value pairs in this hash will be passed to any L<AnyEvent::Handle>
		372	constructor that is called - not all requests will create a handle, and
		373	sometimes more than one is created, so this parameter is only good for
		374	setting hints.
		375
		376	Example: set the maximum read size to 4096, to potentially conserve memory
		377	at the cost of speed.
		378
		379	handle_params => {
		380	max_read_size => 4096,
		381	},
		382
279	=back	383	=back
280		384
281	Example: make a simple HTTP GET request for http://www.nethype.de/	385	Example: do a simple HTTP GET request for http://www.nethype.de/ and print
		386	the response body.
282		387
283	http_request GET => "http://www.nethype.de/", sub {	388	http_request GET => "http://www.nethype.de/", sub {
284	my ($body, $hdr) = @_;	389	my ($body, $hdr) = @_;
285	print "$body\n";	390	print "$body\n";
286	};	391	};
287		392
288	Example: make a HTTP HEAD request on https://www.google.com/, use a	393	Example: do a HTTP HEAD request on https://www.google.com/, use a
289	timeout of 30 seconds.	394	timeout of 30 seconds.
290		395
291	http_request	396	http_request
292	GET => "https://www.google.com",	397	HEAD => "https://www.google.com",
		398	headers => { "user-agent" => "MySearchClient 1.0" },
293	timeout => 30,	399	timeout => 30,
294	sub {	400	sub {
295	my ($body, $hdr) = @_;	401	my ($body, $hdr) = @_;
296	use Data::Dumper;	402	use Data::Dumper;
297	print Dumper $hdr;	403	print Dumper $hdr;
298	}	404	}
299	;	405	;
300		406
301	Example: make another simple HTTP GET request, but immediately try to	407	Example: do another simple HTTP GET request, but immediately try to
302	cancel it.	408	cancel it.
303		409
304	my $request = http_request GET => "http://www.nethype.de/", sub {	410	my $request = http_request GET => "http://www.nethype.de/", sub {
305	my ($body, $hdr) = @_;	411	my ($body, $hdr) = @_;
306	print "$body\n";	412	print "$body\n";
307	};	413	};
308		414
309	undef $request;	415	undef $request;
310		416
311	=cut	417	=cut
		418
		419	#############################################################################
		420	# wait queue/slots
312		421
313	sub _slot_schedule;	422	sub _slot_schedule;
314	sub _slot_schedule($) {	423	sub _slot_schedule($) {
315	my $host = shift;	424	my $host = shift;
316		425
…		…
338	push @{ $CO_SLOT{$_[0]}[1] }, $_[1];	447	push @{ $CO_SLOT{$_[0]}[1] }, $_[1];
339		448
340	_slot_schedule $_[0];	449	_slot_schedule $_[0];
341	}	450	}
342		451
343	our $qr_nl = qr{\015?\012};	452	#############################################################################
		453	# cookie handling
		454
		455	# expire cookies
		456	sub cookie_jar_expire($;$) {
		457	my ($jar, $session_end) = @_;
		458
		459	%$jar = () if $jar->{version} != 2;
		460
		461	my $anow = AE::now;
		462
		463	while (my ($chost, $paths) = each %$jar) {
		464	next unless ref $paths;
		465
		466	while (my ($cpath, $cookies) = each %$paths) {
		467	while (my ($cookie, $kv) = each %$cookies) {
		468	if (exists $kv->{_expires}) {
		469	delete $cookies->{$cookie}
		470	if $anow > $kv->{_expires};
		471	} elsif ($session_end) {
		472	delete $cookies->{$cookie};
		473	}
		474	}
		475
		476	delete $paths->{$cpath}
		477	unless %$cookies;
		478	}
		479
		480	delete $jar->{$chost}
		481	unless %$paths;
		482	}
		483	}
		484
		485	# extract cookies from jar
		486	sub cookie_jar_extract($$$$) {
		487	my ($jar, $scheme, $host, $path) = @_;
		488
		489	%$jar = () if $jar->{version} != 2;
		490
		491	$host = AnyEvent::Util::idn_to_ascii $host
		492	if $host =~ /[^\x00-\x7f]/;
		493
		494	my @cookies;
		495
		496	while (my ($chost, $paths) = each %$jar) {
		497	next unless ref $paths;
		498
		499	# exact match or suffix including . match
		500	$chost eq $host or ".$chost" eq substr $host, -1 - length $chost
		501	or next;
		502
		503	while (my ($cpath, $cookies) = each %$paths) {
		504	next unless $cpath eq substr $path, 0, length $cpath;
		505
		506	while (my ($cookie, $kv) = each %$cookies) {
		507	next if $scheme ne "https" && exists $kv->{secure};
		508
		509	if (exists $kv->{_expires} and AE::now > $kv->{_expires}) {
		510	delete $cookies->{$cookie};
		511	next;
		512	}
		513
		514	my $value = $kv->{value};
		515
		516	if ($value =~ /[=;,[:space:]]/) {
		517	$value =~ s/([\\"])/\\$1/g;
		518	$value = "\"$value\"";
		519	}
		520
		521	push @cookies, "$cookie=$value";
		522	}
		523	}
		524	}
		525
		526	\@cookies
		527	}
		528
		529	# parse set_cookie header into jar
		530	sub cookie_jar_set_cookie($$$$) {
		531	my ($jar, $set_cookie, $host, $date) = @_;
		532
		533	%$jar = () if $jar->{version} != 2;
		534
		535	my $anow = int AE::now;
		536	my $snow; # server-now
		537
		538	for ($set_cookie) {
		539	# parse NAME=VALUE
		540	my @kv;
		541
		542	# expires is not http-compliant in the original cookie-spec,
		543	# we support the official date format and some extensions
		544	while (
		545	m{
		546	\G\s*
		547	(?:
		548	expires \s*=\s* ([A-Z][a-z][a-z]+,\ [^,;]+)
		549	| ([^=;,[:space:]]+) (?: \s*=\s* (?: "((?:[^\\"]+|\\.)*)" | ([^;,[:space:]]*) ) )?
		550	)
		551	}gcxsi
		552	) {
		553	my $name = $2;
		554	my $value = $4;
		555
		556	if (defined $1) {
		557	# expires
		558	$name = "expires";
		559	$value = $1;
		560	} elsif (defined $3) {
		561	# quoted
		562	$value = $3;
		563	$value =~ s/\\(.)/$1/gs;
		564	}
		565
		566	push @kv, @kv ? lc $name : $name, $value;
		567
		568	last unless /\G\s*;/gc;
		569	}
		570
		571	last unless @kv;
		572
		573	my $name = shift @kv;
		574	my %kv = (value => shift @kv, @kv);
		575
		576	if (exists $kv{"max-age"}) {
		577	$kv{_expires} = $anow + delete $kv{"max-age"};
		578	} elsif (exists $kv{expires}) {
		579	$snow ||= parse_date ($date) || $anow;
		580	$kv{_expires} = $anow + (parse_date (delete $kv{expires}) - $snow);
		581	} else {
		582	delete $kv{_expires};
		583	}
		584
		585	my $cdom;
		586	my $cpath = (delete $kv{path}) || "/";
		587
		588	if (exists $kv{domain}) {
		589	$cdom = $kv{domain};
		590
		591	$cdom =~ s/^\.?/./; # make sure it starts with a "."
		592
		593	next if $cdom =~ /\.$/;
		594
		595	# this is not rfc-like and not netscape-like. go figure.
		596	my $ndots = $cdom =~ y/.//;
		597	next if $ndots < ($cdom =~ /\.[^.][^.]\.[^.][^.]$/ ? 3 : 2);
		598
		599	$cdom = substr $cdom, 1; # remove initial .
		600	} else {
		601	$cdom = $host;
		602	}
		603
		604	# store it
		605	$jar->{version} = 2;
		606	$jar->{lc $cdom}{$cpath}{$name} = \%kv;
		607
		608	redo if /\G\s*,/gc;
		609	}
		610	}
		611
		612	#############################################################################
		613	# keepalive/persistent connection cache
		614
		615	# fetch a connection from the keepalive cache
		616	sub ka_fetch($) {
		617	my $ka_key = shift;
		618
		619	my $hdl = pop @{ $KA_CACHE{$ka_key} }; # currently we reuse the MOST RECENTLY USED connection
		620	delete $KA_CACHE{$ka_key}
		621	unless @{ $KA_CACHE{$ka_key} };
		622
		623	$hdl
		624	}
		625
		626	sub ka_store($$) {
		627	my ($ka_key, $hdl) = @_;
		628
		629	my $kaa = $KA_CACHE{$ka_key} ||= [];
		630
		631	my $destroy = sub {
		632	my @ka = grep $_ != $hdl, @{ $KA_CACHE{$ka_key} };
		633
		634	$hdl->destroy;
		635
		636	@ka
		637	? $KA_CACHE{$ka_key} = \@ka
		638	: delete $KA_CACHE{$ka_key};
		639	};
		640
		641	# on error etc., destroy
		642	$hdl->on_error ($destroy);
		643	$hdl->on_eof ($destroy);
		644	$hdl->on_read ($destroy);
		645	$hdl->timeout ($PERSISTENT_TIMEOUT);
		646
		647	push @$kaa, $hdl;
		648	shift @$kaa while @$kaa > $MAX_PER_HOST;
		649	}
		650
		651	#############################################################################
		652	# utilities
		653
		654	# continue to parse $_ for headers and place them into the arg
		655	sub _parse_hdr() {
		656	my %hdr;
		657
		658	# things seen, not parsed:
		659	# p3pP="NON CUR OTPi OUR NOR UNI"
		660
		661	$hdr{lc $1} .= ",$2"
		662	while /\G
		663	([^:\000-\037]*):
		664	[\011\040]*
		665	((?: [^\012]+ | \012[\011\040] )*)
		666	\012
		667	/gxc;
		668
		669	/\G$/
		670	or return;
		671
		672	# remove the "," prefix we added to all headers above
		673	substr $_, 0, 1, ""
		674	for values %hdr;
		675
		676	\%hdr
		677	}
		678
		679	#############################################################################
		680	# http_get
		681
344	our $qr_nlnl = qr{(?<![^\012])\015?\012};	682	our $qr_nlnl = qr{(?<![^\012])\015?\012};
345		683
346	our $TLS_CTX_LOW = { cache => 1, sslv2 => 1 };	684	our $TLS_CTX_LOW = { cache => 1, sslv2 => 1 };
347	our $TLS_CTX_HIGH = { cache => 1, verify => 1, verify_peername => "https" };	685	our $TLS_CTX_HIGH = { cache => 1, verify => 1, verify_peername => "https" };
		686
		687	# maybe it should just become a normal object :/
		688
		689	sub _destroy_state(\%) {
		690	my ($state) = @_;
		691
		692	$state->{handle}->destroy if $state->{handle};
		693	%$state = ();
		694	}
		695
		696	sub _error(\%$$) {
		697	my ($state, $cb, $hdr) = @_;
		698
		699	&_destroy_state ($state);
		700
		701	$cb->(undef, $hdr);
		702	()
		703	}
		704
		705	our %IDEMPOTENT = (
		706	DELETE => 1,
		707	GET => 1,
		708	HEAD => 1,
		709	OPTIONS => 1,
		710	PUT => 1,
		711	TRACE => 1,
		712
		713	ACL => 1,
		714	"BASELINE-CONTROL" => 1,
		715	BIND => 1,
		716	CHECKIN => 1,
		717	CHECKOUT => 1,
		718	COPY => 1,
		719	LABEL => 1,
		720	LINK => 1,
		721	MERGE => 1,
		722	MKACTIVITY => 1,
		723	MKCALENDAR => 1,
		724	MKCOL => 1,
		725	MKREDIRECTREF => 1,
		726	MKWORKSPACE => 1,
		727	MOVE => 1,
		728	ORDERPATCH => 1,
		729	PROPFIND => 1,
		730	PROPPATCH => 1,
		731	REBIND => 1,
		732	REPORT => 1,
		733	SEARCH => 1,
		734	UNBIND => 1,
		735	UNCHECKOUT => 1,
		736	UNLINK => 1,
		737	UNLOCK => 1,
		738	UPDATE => 1,
		739	UPDATEREDIRECTREF => 1,
		740	"VERSION-CONTROL" => 1,
		741	);
348		742
349	sub http_request($$@) {	743	sub http_request($$@) {
350	my $cb = pop;	744	my $cb = pop;
351	my ($method, $url, %arg) = @_;	745	my ($method, $url, %arg) = @_;
352		746
…		…
367	my @pseudo = (URL => $url);	761	my @pseudo = (URL => $url);
368	push @pseudo, Redirect => delete $arg{Redirect} if exists $arg{Redirect};	762	push @pseudo, Redirect => delete $arg{Redirect} if exists $arg{Redirect};
369		763
370	my $recurse = exists $arg{recurse} ? delete $arg{recurse} : $MAX_RECURSE;	764	my $recurse = exists $arg{recurse} ? delete $arg{recurse} : $MAX_RECURSE;
371		765
372	return $cb->(undef, { Status => 599, Reason => "Too many redirections", @pseudo })	766	return $cb->(undef, { @pseudo, Status => 599, Reason => "Too many redirections" })
373	if $recurse < 0;	767	if $recurse < 0;
374		768
375	my $proxy = $arg{proxy} || $PROXY;	769	my $proxy = exists $arg{proxy} ? $arg{proxy} : $PROXY;
376	my $timeout = $arg{timeout} || $TIMEOUT;	770	my $timeout = $arg{timeout} || $TIMEOUT;
377		771
378	my ($uscheme, $uauthority, $upath, $query, $fragment) =	772	my ($uscheme, $uauthority, $upath, $query, undef) = # ignore fragment
379	$url =~ m|(?:([^:/?#]+):)?(?://([^/?#]*))?([^?#]*)(?:(\?[^#]*))?(?:#(.*))?|;	773	$url =~ m|^([^:]+):(?://([^/?#]*))?([^?#]*)(?:(\?[^#]*))?(?:#(.*))?$|;
380		774
381	$uscheme = lc $uscheme;	775	$uscheme = lc $uscheme;
382		776
383	my $uport = $uscheme eq "http" ? 80	777	my $uport = $uscheme eq "http" ? 80
384	: $uscheme eq "https" ? 443	778	: $uscheme eq "https" ? 443
385	: return $cb->(undef, { Status => 599, Reason => "Only http and https URL schemes supported", @pseudo });	779	: return $cb->(undef, { @pseudo, Status => 599, Reason => "Only http and https URL schemes supported" });
386		780
387	$uauthority =~ /^(?: .*\@ )? ([^\@:]+) (?: : (\d+) )?$/x	781	$uauthority =~ /^(?: .*\@ )? ([^\@]+?) (?: : (\d+) )?$/x
388	or return $cb->(undef, { Status => 599, Reason => "Unparsable URL", @pseudo });	782	or return $cb->(undef, { @pseudo, Status => 599, Reason => "Unparsable URL" });
389		783
390	my $uhost = $1;	784	my $uhost = lc $1;
391	$uport = $2 if defined $2;	785	$uport = $2 if defined $2;
392		786
393	$hdr{host} = defined $2 ? "$uhost:$2" : "$uhost"	787	$hdr{host} = defined $2 ? "$uhost:$2" : "$uhost"
394	unless exists $hdr{host};	788	unless exists $hdr{host};
395		789
…		…
398		792
399	$upath =~ s%^/?%/%;	793	$upath =~ s%^/?%/%;
400		794
401	# cookie processing	795	# cookie processing
402	if (my $jar = $arg{cookie_jar}) {	796	if (my $jar = $arg{cookie_jar}) {
403	%$jar = () if $jar->{version} != 1;	797	my $cookies = cookie_jar_extract $jar, $uscheme, $uhost, $upath;
404		798
405	my @cookie;
406
407	while (my ($chost, $v) = each %$jar) {
408	if ($chost =~ /^\./) {
409	next unless $chost eq substr $uhost, -length $chost;
410	} elsif ($chost =~ /\./) {
411	next unless $chost eq $uhost;
412	} else {
413	next;
414	}
415
416	while (my ($cpath, $v) = each %$v) {
417	next unless $cpath eq substr $upath, 0, length $cpath;
418
419	while (my ($k, $v) = each %$v) {
420	next if $uscheme ne "https" && exists $v->{secure};
421	my $value = $v->{value};
422	$value =~ s/([\\"])/\\$1/g;
423	push @cookie, "$k=\"$value\"";
424	}
425	}
426	}
427
428	$hdr{cookie} = join "; ", @cookie	799	$hdr{cookie} = join "; ", @$cookies
429	if @cookie;	800	if @$cookies;
430	}	801	}
431		802
432	my ($rhost, $rport, $rscheme, $rpath); # request host, port, path	803	my ($rhost, $rport, $rscheme, $rpath); # request host, port, path
433		804
434	if ($proxy) {	805	if ($proxy) {
…		…
437	$rscheme = "http" unless defined $rscheme;	808	$rscheme = "http" unless defined $rscheme;
438		809
439	# don't support https requests over https-proxy transport,	810	# don't support https requests over https-proxy transport,
440	# can't be done with tls as spec'ed, unless you double-encrypt.	811	# can't be done with tls as spec'ed, unless you double-encrypt.
441	$rscheme = "http" if $uscheme eq "https" && $rscheme eq "https";	812	$rscheme = "http" if $uscheme eq "https" && $rscheme eq "https";
		813
		814	$rhost = lc $rhost;
		815	$rscheme = lc $rscheme;
442	} else {	816	} else {
443	($rhost, $rport, $rscheme, $rpath) = ($uhost, $uport, $uscheme, $upath);	817	($rhost, $rport, $rscheme, $rpath) = ($uhost, $uport, $uscheme, $upath);
444	}	818	}
445		819
446	# leave out fragment and query string, just a heuristic	820	# leave out fragment and query string, just a heuristic
447	$hdr{referer} ||= "$uscheme://$uauthority$upath" unless exists $hdr{referer};	821	$hdr{referer} = "$uscheme://$uauthority$upath" unless exists $hdr{referer};
448	$hdr{"user-agent"} ||= $USERAGENT unless exists $hdr{"user-agent"};	822	$hdr{"user-agent"} = $USERAGENT unless exists $hdr{"user-agent"};
449		823
450	$hdr{"content-length"} = length $arg{body}	824	$hdr{"content-length"} = length $arg{body}
451	if length $arg{body} || $method ne "GET";	825	if length $arg{body} || $method ne "GET";
452		826
		827	my $idempotent = $IDEMPOTENT{$method};
		828
		829	# default value for keepalive is true iff the request is for an idempotent method
		830	my $persistent = exists $arg{persistent} ? !!$arg{persistent} : $idempotent;
		831	my $keepalive = exists $arg{keepalive} ? !!$arg{keepalive} : !$proxy;
		832	my $was_persistent; # true if this is actually a recycled connection
		833
		834	# the key to use in the keepalive cache
		835	my $ka_key = "$uscheme\x00$uhost\x00$uport\x00$arg{sessionid}";
		836
		837	$hdr{connection} = ($persistent ? $keepalive ? "keep-alive, " : "" : "close, ") . "Te"; #1.1
		838	$hdr{te} = "trailers" unless exists $hdr{te}; #1.1
		839
453	my %state = (connect_guard => 1);	840	my %state = (connect_guard => 1);
		841
		842	my $ae_error = 595; # connecting
		843
		844	# handle actual, non-tunneled, request
		845	my $handle_actual_request = sub {
		846	$ae_error = 596; # request phase
		847
		848	my $hdl = $state{handle};
		849
		850	$hdl->starttls ("connect") if $uscheme eq "https" && !exists $hdl->{tls};
		851
		852	# send request
		853	$hdl->push_write (
		854	"$method $rpath HTTP/1.1\015\012"
		855	. (join "", map "\u$_: $hdr{$_}\015\012", grep defined $hdr{$_}, keys %hdr)
		856	. "\015\012"
		857	. $arg{body}
		858	);
		859
		860	# return if error occurred during push_write()
		861	return unless %state;
		862
		863	# reduce memory usage, save a kitten, also re-use it for the response headers.
		864	%hdr = ();
		865
		866	# status line and headers
		867	$state{read_response} = sub {
		868	return unless %state;
		869
		870	for ("$_[1]") {
		871	y/\015//d; # weed out any \015, as they show up in the weirdest of places.
		872
		873	/^HTTP\/0*([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\012]*) )? \012/gxci
		874	or return _error %state, $cb, { @pseudo, Status => 599, Reason => "Invalid server response" };
		875
		876	# 100 Continue handling
		877	# should not happen as we don't send expect: 100-continue,
		878	# but we handle it just in case.
		879	# since we send the request body regardless, if we get an error
		880	# we are out of-sync, which we currently do NOT handle correctly.
		881	return $state{handle}->push_read (line => $qr_nlnl, $state{read_response})
		882	if $2 eq 100;
		883
		884	push @pseudo,
		885	HTTPVersion => $1,
		886	Status => $2,
		887	Reason => $3,
		888	;
		889
		890	my $hdr = _parse_hdr
		891	or return _error %state, $cb, { @pseudo, Status => 599, Reason => "Garbled response headers" };
		892
		893	%hdr = (%$hdr, @pseudo);
		894	}
		895
		896	# redirect handling
		897	# relative uri handling forced by microsoft and other shitheads.
		898	# we give our best and fall back to URI if available.
		899	if (exists $hdr{location}) {
		900	my $loc = $hdr{location};
		901
		902	if ($loc =~ m%^//%) { # //
		903	$loc = "$uscheme:$loc";
		904
		905	} elsif ($loc eq "") {
		906	$loc = $url;
		907
		908	} elsif ($loc !~ /^(?: $ | [^:\/?\#]+ : )/x) { # anything "simple"
		909	$loc =~ s/^\.\/+//;
		910
		911	if ($loc !~ m%^[.?#]%) {
		912	my $prefix = "$uscheme://$uauthority";
		913
		914	unless ($loc =~ s/^\///) {
		915	$prefix .= $upath;
		916	$prefix =~ s/\/[^\/]*$//;
		917	}
		918
		919	$loc = "$prefix/$loc";
		920
		921	} elsif (eval { require URI }) { # uri
		922	$loc = URI->new_abs ($loc, $url)->as_string;
		923
		924	} else {
		925	return _error %state, $cb, { @pseudo, Status => 599, Reason => "Cannot parse Location (URI module missing)" };
		926	#$hdr{Status} = 599;
		927	#$hdr{Reason} = "Unparsable Redirect (URI module missing)";
		928	#$recurse = 0;
		929	}
		930	}
		931
		932	$hdr{location} = $loc;
		933	}
		934
		935	my $redirect;
		936
		937	if ($recurse) {
		938	my $status = $hdr{Status};
		939
		940	# industry standard is to redirect POST as GET for
		941	# 301, 302 and 303, in contrast to HTTP/1.0 and 1.1.
		942	# also, the UA should ask the user for 301 and 307 and POST,
		943	# industry standard seems to be to simply follow.
		944	# we go with the industry standard. 308 is defined
		945	# by rfc7538
		946	if ($status == 301 or $status == 302 or $status == 303) {
		947	$redirect = 1;
		948	# HTTP/1.1 is unclear on how to mutate the method
		949	unless ($method eq "HEAD") {
		950	$method = "GET";
		951	delete $arg{body};
		952	}
		953	} elsif ($status == 307 or $status == 308) {
		954	$redirect = 1;
		955	}
		956	}
		957
		958	my $finish = sub { # ($data, $err_status, $err_reason[, $persistent])
		959	if ($state{handle}) {
		960	# handle keepalive
		961	if (
		962	$persistent
		963	&& $_[3]
		964	&& ($hdr{HTTPVersion} < 1.1
		965	? $hdr{connection} =~ /\bkeep-?alive\b/i
		966	: $hdr{connection} !~ /\bclose\b/i)
		967	) {
		968	ka_store $ka_key, delete $state{handle};
		969	} else {
		970	# no keepalive, destroy the handle
		971	$state{handle}->destroy;
		972	}
		973	}
		974
		975	%state = ();
		976
		977	if (defined $_[1]) {
		978	$hdr{OrigStatus} = $hdr{Status}; $hdr{Status} = $_[1];
		979	$hdr{OrigReason} = $hdr{Reason}; $hdr{Reason} = $_[2];
		980	}
		981
		982	# set-cookie processing
		983	if ($arg{cookie_jar}) {
		984	cookie_jar_set_cookie $arg{cookie_jar}, $hdr{"set-cookie"}, $uhost, $hdr{date};
		985	}
		986
		987	if ($redirect && exists $hdr{location}) {
		988	# we ignore any errors, as it is very common to receive
		989	# Content-Length != 0 but no actual body
		990	# we also access %hdr, as $_[1] might be an erro
		991	$state{recurse} =
		992	http_request (
		993	$method => $hdr{location},
		994	%arg,
		995	recurse => $recurse - 1,
		996	Redirect => [$_[0], \%hdr],
		997	sub {
		998	%state = ();
		999	&$cb
		1000	},
		1001	);
		1002	} else {
		1003	$cb->($_[0], \%hdr);
		1004	}
		1005	};
		1006
		1007	$ae_error = 597; # body phase
		1008
		1009	my $chunked = $hdr{"transfer-encoding"} =~ /\bchunked\b/i; # not quite correct...
		1010
		1011	my $len = $chunked ? undef : $hdr{"content-length"};
		1012
		1013	# body handling, many different code paths
		1014	# - no body expected
		1015	# - want_body_handle
		1016	# - te chunked
		1017	# - 2x length known (with or without on_body)
		1018	# - 2x length not known (with or without on_body)
		1019	if (!$redirect && $arg{on_header} && !$arg{on_header}(\%hdr)) {
		1020	$finish->(undef, 598 => "Request cancelled by on_header");
		1021	} elsif (
		1022	$hdr{Status} =~ /^(?:1..|204|205|304)$/
		1023	or $method eq "HEAD"
		1024	or (defined $len && $len == 0) # == 0, not !, because "0 " is true
		1025	) {
		1026	# no body
		1027	$finish->("", undef, undef, 1);
		1028
		1029	} elsif (!$redirect && $arg{want_body_handle}) {
		1030	$_[0]->on_eof (undef);
		1031	$_[0]->on_error (undef);
		1032	$_[0]->on_read (undef);
		1033
		1034	$finish->(delete $state{handle});
		1035
		1036	} elsif ($chunked) {
		1037	my $cl = 0;
		1038	my $body = "";
		1039	my $on_body = (!$redirect && $arg{on_body}) || sub { $body .= shift; 1 };
		1040
		1041	$state{read_chunk} = sub {
		1042	$_[1] =~ /^([0-9a-fA-F]+)/
		1043	or return $finish->(undef, $ae_error => "Garbled chunked transfer encoding");
		1044
		1045	my $len = hex $1;
		1046
		1047	if ($len) {
		1048	$cl += $len;
		1049
		1050	$_[0]->push_read (chunk => $len, sub {
		1051	$on_body->($_[1], \%hdr)
		1052	or return $finish->(undef, 598 => "Request cancelled by on_body");
		1053
		1054	$_[0]->push_read (line => sub {
		1055	length $_[1]
		1056	and return $finish->(undef, $ae_error => "Garbled chunked transfer encoding");
		1057	$_[0]->push_read (line => $state{read_chunk});
		1058	});
		1059	});
		1060	} else {
		1061	$hdr{"content-length"} ||= $cl;
		1062
		1063	$_[0]->push_read (line => $qr_nlnl, sub {
		1064	if (length $_[1]) {
		1065	for ("$_[1]") {
		1066	y/\015//d; # weed out any \015, as they show up in the weirdest of places.
		1067
		1068	my $hdr = _parse_hdr
		1069	or return $finish->(undef, $ae_error => "Garbled response trailers");
		1070
		1071	%hdr = (%hdr, %$hdr);
		1072	}
		1073	}
		1074
		1075	$finish->($body, undef, undef, 1);
		1076	});
		1077	}
		1078	};
		1079
		1080	$_[0]->push_read (line => $state{read_chunk});
		1081
		1082	} elsif (!$redirect && $arg{on_body}) {
		1083	if (defined $len) {
		1084	$_[0]->on_read (sub {
		1085	$len -= length $_[0]{rbuf};
		1086
		1087	$arg{on_body}(delete $_[0]{rbuf}, \%hdr)
		1088	or return $finish->(undef, 598 => "Request cancelled by on_body");
		1089
		1090	$len > 0
		1091	or $finish->("", undef, undef, 1);
		1092	});
		1093	} else {
		1094	$_[0]->on_eof (sub {
		1095	$finish->("");
		1096	});
		1097	$_[0]->on_read (sub {
		1098	$arg{on_body}(delete $_[0]{rbuf}, \%hdr)
		1099	or $finish->(undef, 598 => "Request cancelled by on_body");
		1100	});
		1101	}
		1102	} else {
		1103	$_[0]->on_eof (undef);
		1104
		1105	if (defined $len) {
		1106	$_[0]->on_read (sub {
		1107	$finish->((substr delete $_[0]{rbuf}, 0, $len, ""), undef, undef, 1)
		1108	if $len <= length $_[0]{rbuf};
		1109	});
		1110	} else {
		1111	$_[0]->on_error (sub {
		1112	($! == Errno::EPIPE || !$!)
		1113	? $finish->(delete $_[0]{rbuf})
		1114	: $finish->(undef, $ae_error => $_[2]);
		1115	});
		1116	$_[0]->on_read (sub { });
		1117	}
		1118	}
		1119	};
		1120
		1121	# if keepalive is enabled, then the server closing the connection
		1122	# before a response can happen legally - we retry on idempotent methods.
		1123	if ($was_persistent && $idempotent) {
		1124	my $old_eof = $hdl->{on_eof};
		1125	$hdl->{on_eof} = sub {
		1126	_destroy_state %state;
		1127
		1128	%state = ();
		1129	$state{recurse} =
		1130	http_request (
		1131	$method => $url,
		1132	%arg,
		1133	recurse => $recurse - 1,
		1134	persistent => 0,
		1135	sub {
		1136	%state = ();
		1137	&$cb
		1138	}
		1139	);
		1140	};
		1141	$hdl->on_read (sub {
		1142	return unless %state;
		1143
		1144	# as soon as we receive something, a connection close
		1145	# once more becomes a hard error
		1146	$hdl->{on_eof} = $old_eof;
		1147	$hdl->push_read (line => $qr_nlnl, $state{read_response});
		1148	});
		1149	} else {
		1150	$hdl->push_read (line => $qr_nlnl, $state{read_response});
		1151	}
		1152	};
		1153
		1154	my $prepare_handle = sub {
		1155	my ($hdl) = $state{handle};
		1156
		1157	$hdl->on_error (sub {
		1158	_error %state, $cb, { @pseudo, Status => $ae_error, Reason => $_[2] };
		1159	});
		1160	$hdl->on_eof (sub {
		1161	_error %state, $cb, { @pseudo, Status => $ae_error, Reason => "Unexpected end-of-file" };
		1162	});
		1163	$hdl->timeout_reset;
		1164	$hdl->timeout ($timeout);
		1165	};
		1166
		1167	# connected to proxy (or origin server)
		1168	my $connect_cb = sub {
		1169	my $fh = shift
		1170	or return _error %state, $cb, { @pseudo, Status => $ae_error, Reason => "$!" };
		1171
		1172	return unless delete $state{connect_guard};
		1173
		1174	# get handle
		1175	$state{handle} = new AnyEvent::Handle
		1176	%{ $arg{handle_params} },
		1177	fh => $fh,
		1178	peername => $uhost,
		1179	tls_ctx => $arg{tls_ctx},
		1180	;
		1181
		1182	$prepare_handle->();
		1183
		1184	#$state{handle}->starttls ("connect") if $rscheme eq "https";
		1185
		1186	# now handle proxy-CONNECT method
		1187	if ($proxy && $uscheme eq "https") {
		1188	# oh dear, we have to wrap it into a connect request
		1189
		1190	my $auth = exists $hdr{"proxy-authorization"}
		1191	? "proxy-authorization: " . (delete $hdr{"proxy-authorization"}) . "\015\012"
		1192	: "";
		1193
		1194	# maybe re-use $uauthority with patched port?
		1195	$state{handle}->push_write ("CONNECT $uhost:$uport HTTP/1.0\015\012$auth\015\012");
		1196	$state{handle}->push_read (line => $qr_nlnl, sub {
		1197	$_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix
		1198	or return _error %state, $cb, { @pseudo, Status => 599, Reason => "Invalid proxy connect response ($_[1])" };
		1199
		1200	if ($2 == 200) {
		1201	$rpath = $upath;
		1202	$handle_actual_request->();
		1203	} else {
		1204	_error %state, $cb, { @pseudo, Status => $2, Reason => $3 };
		1205	}
		1206	});
		1207	} else {
		1208	delete $hdr{"proxy-authorization"} unless $proxy;
		1209
		1210	$handle_actual_request->();
		1211	}
		1212	};
454		1213
455	_get_slot $uhost, sub {	1214	_get_slot $uhost, sub {
456	$state{slot_guard} = shift;	1215	$state{slot_guard} = shift;
457		1216
458	return unless $state{connect_guard};	1217	return unless $state{connect_guard};
459		1218
460	$state{connect_guard} = AnyEvent::Socket::tcp_connect $rhost, $rport, sub {	1219	# try to use an existing keepalive connection, but only if we, ourselves, plan
461	$state{fh} = shift	1220	# on a keepalive request (in theory, this should be a separate config option).
462	or do {	1221	if ($persistent && $KA_CACHE{$ka_key}) {
463	my $err = "$!";	1222	$was_persistent = 1;
464	%state = ();
465	return $cb->(undef, { Status => 599, Reason => $err, @pseudo });
466	};
467		1223
468	pop; # free memory, save a tree	1224	$state{handle} = ka_fetch $ka_key;
469		1225	# $state{handle}->destroyed
470	return unless delete $state{connect_guard};	1226	# and die "AnyEvent::HTTP: unexpectedly got a destructed handle (1), please report.";#d#
471		1227	$prepare_handle->();
472	# get handle	1228	# $state{handle}->destroyed
473	$state{handle} = new AnyEvent::Handle	1229	# and die "AnyEvent::HTTP: unexpectedly got a destructed handle (2), please report.";#d#
474	fh => $state{fh},
475	peername => $rhost,
476	tls_ctx => $arg{tls_ctx},
477	# these need to be reconfigured on keepalive handles
478	timeout => $timeout,
479	on_error => sub {
480	%state = ();
481	$cb->(undef, { Status => 599, Reason => $_[2], @pseudo });
482	},
483	on_eof => sub {
484	%state = ();
485	$cb->(undef, { Status => 599, Reason => "Unexpected end-of-file", @pseudo });
486	},
487	;
488
489	# limit the number of persistent connections
490	# keepalive not yet supported
491	# if ($KA_COUNT{$_[1]} < $MAX_PERSISTENT_PER_HOST) {
492	# ++$KA_COUNT{$_[1]};
493	# $state{handle}{ka_count_guard} = AnyEvent::Util::guard {
494	# --$KA_COUNT{$_[1]}
495	# };
496	# $hdr{connection} = "keep-alive";
497	# } else {
498	delete $hdr{connection};
499	# }
500
501	$state{handle}->starttls ("connect") if $rscheme eq "https";
502
503	# handle actual, non-tunneled, request
504	my $handle_actual_request = sub {
505	$state{handle}->starttls ("connect") if $uscheme eq "https" && !exists $state{handle}{tls};
506
507	# send request
508	$state{handle}->push_write (
509	"$method $rpath HTTP/1.0\015\012"
510	. (join "", map "\u$_: $hdr{$_}\015\012", grep defined $hdr{$_}, keys %hdr)
511	. "\015\012"
512	. (delete $arg{body})
513	);
514
515	# return if error occured during push_write()
516	return unless %state;
517
518	%hdr = (); # reduce memory usage, save a kitten, also make it possible to re-use
519
520	# status line
521	$state{handle}->push_read (line => $qr_nl, sub {
522	$_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix
523	or return (%state = (), $cb->(undef, { Status => 599, Reason => "Invalid server response ($_[1])", @pseudo }));
524
525	push @pseudo,
526	HTTPVersion => $1,
527	Status => $2,
528	Reason => $3,
529	;
530
531	# headers, could be optimized a bit
532	$state{handle}->unshift_read (line => $qr_nlnl, sub {
533	for ("$_[1]") {
534	y/\015//d; # weed out any \015, as they show up in the weirdest of places.
535
536	# things seen, not parsed:
537	# p3pP="NON CUR OTPi OUR NOR UNI"
538
539	$hdr{lc $1} .= ",$2"
540	while /\G
541	([^:\000-\037]*):
542	[\011\040]*
543	((?: [^\012]+ | \012[\011\040] )*)
544	\012
545	/gxc;
546
547	/\G$/
548	or return (%state = (), $cb->(undef, { Status => 599, Reason => "Garbled response headers", @pseudo }));
549	}
550
551	# remove the "," prefix we added to all headers above
552	substr $_, 0, 1, ""
553	for values %hdr;
554
555	# patch in all pseudo headers
556	%hdr = (%hdr, @pseudo);
557
558	# redirect handling
559	# microsoft and other shitheads don't give a shit for following standards,
560	# try to support some common forms of broken Location headers.
561	if ($hdr{location} !~ /^(?: $ | [^:\/?\#]+ : )/x) {
562	$hdr{location} =~ s/^\.\/+//;
563
564	my $url = "$rscheme://$uhost:$uport";
565
566	unless ($hdr{location} =~ s/^\///) {
567	$url .= $upath;
568	$url =~ s/\/[^\/]*$//;
569	}
570
571	$hdr{location} = "$url/$hdr{location}";
572	}
573
574	my $redirect;
575
576	if ($recurse) {
577	my $status = $hdr{Status};
578
579	if (($status == 301 || $status == 302) && $method ne "POST") {
580	# apparently, mozilla et al. just change POST to GET here
581	# more research is needed before we do the same
582	$redirect = 1;
583	} elsif ($status == 303) {
584	# even http/1.1 is unclear on how to mutate the method
585	$method = "GET" unless $method eq "HEAD";
586	$redirect = 1;
587	} elsif ($status == 307 && $method =~ /^(?:GET|HEAD)$/) {
588	$redirect = 1;
589	}
590	}
591
592	my $finish = sub {
593	$state{handle}->destroy if $state{handle};
594	%state = ();
595
596	# set-cookie processing
597	if ($arg{cookie_jar}) {
598	for ($_[1]{"set-cookie"}) {
599	# parse NAME=VALUE
600	my @kv;
601
602	while (/\G\s* ([^=;,[:space:]]+) \s*=\s* (?: "((?:[^\\"]+|\\.)*)" | ([^=;,[:space:]]*) )/gcxs) {
603	my $name = $1;
604	my $value = $3;
605
606	unless ($value) {
607	$value = $2;
608	$value =~ s/\\(.)/$1/gs;
609	}
610
611	push @kv, $name => $value;
612
613	last unless /\G\s*;/gc;
614	}
615
616	last unless @kv;
617
618	my $name = shift @kv;
619	my %kv = (value => shift @kv, @kv);
620
621	my $cdom;
622	my $cpath = (delete $kv{path}) || "/";
623
624	if (exists $kv{domain}) {
625	$cdom = delete $kv{domain};
626
627	$cdom =~ s/^\.?/./; # make sure it starts with a "."
628
629	next if $cdom =~ /\.$/;
630
631	# this is not rfc-like and not netscape-like. go figure.
632	my $ndots = $cdom =~ y/.//;
633	next if $ndots < ($cdom =~ /\.[^.][^.]\.[^.][^.]$/ ? 3 : 2);
634	} else {
635	$cdom = $uhost;
636	}
637
638	# store it
639	$arg{cookie_jar}{version} = 1;
640	$arg{cookie_jar}{$cdom}{$cpath}{$name} = \%kv;
641
642	redo if /\G\s*,/gc;
643	}
644	}
645
646	if ($redirect && exists $hdr{location}) {
647	# we ignore any errors, as it is very common to receive
648	# Content-Length != 0 but no actual body
649	# we also access %hdr, as $_[1] might be an erro
650	http_request (
651	$method => $hdr{location},
652	%arg,
653	recurse => $recurse - 1,
654	Redirect => \@_,
655	$cb);
656	} else {
657	$cb->($_[0], $_[1]);
658	}
659	};
660
661	my $len = $hdr{"content-length"};
662
663	if (!$redirect && $arg{on_header} && !$arg{on_header}(\%hdr)) {
664	$finish->(undef, { Status => 598, Reason => "Request cancelled by on_header", @pseudo });
665	} elsif (
666	$hdr{Status} =~ /^(?:1..|[23]04)$/
667	or $method eq "HEAD"
668	or (defined $len && !$len)
669	) {
670	# no body
671	$finish->("", \%hdr);
672	} else {
673	# body handling, four different code paths
674	# for want_body_handle, on_body (2x), normal (2x)
675	# we might read too much here, but it does not matter yet (no pers. connections)
676	if (!$redirect && $arg{want_body_handle}) {
677	$_[0]->on_eof (undef);
678	$_[0]->on_error (undef);
679	$_[0]->on_read (undef);
680
681	$finish->(delete $state{handle}, \%hdr);
682
683	} elsif ($arg{on_body}) {
684	$_[0]->on_error (sub { $finish->(undef, { Status => 599, Reason => $_[2], @pseudo }) });
685	if ($len) {
686	$_[0]->on_eof (undef);
687	$_[0]->on_read (sub {
688	$len -= length $_[0]{rbuf};
689
690	$arg{on_body}(delete $_[0]{rbuf}, \%hdr)
691	or $finish->(undef, { Status => 598, Reason => "Request cancelled by on_body", @pseudo });
692
693	$len > 0
694	or $finish->("", \%hdr);
695	});
696	} else {
697	$_[0]->on_eof (sub {
698	$finish->("", \%hdr);
699	});
700	$_[0]->on_read (sub {
701	$arg{on_body}(delete $_[0]{rbuf}, \%hdr)
702	or $finish->(undef, { Status => 598, Reason => "Request cancelled by on_body", @pseudo });
703	});
704	}
705	} else {
706	$_[0]->on_eof (undef);
707
708	if ($len) {
709	$_[0]->on_error (sub { $finish->(undef, { Status => 599, Reason => $_[2], @pseudo }) });
710	$_[0]->on_read (sub {
711	$finish->((substr delete $_[0]{rbuf}, 0, $len, ""), \%hdr)
712	if $len <= length $_[0]{rbuf};
713	});
714	} else {
715	$_[0]->on_error (sub {
716	$! == Errno::EPIPE || !$!
717	? $finish->(delete $_[0]{rbuf}, \%hdr)
718	: $finish->(undef, { Status => 599, Reason => $_[2], @pseudo });
719	});
720	$_[0]->on_read (sub { });
721	}
722	}
723	}
724	});
725	});
726	};
727
728	# now handle proxy-CONNECT method
729	if ($proxy && $uscheme eq "https") {
730	# oh dear, we have to wrap it into a connect request
731
732	# maybe re-use $uauthority with patched port?
733	$state{handle}->push_write ("CONNECT $uhost:$uport HTTP/1.0\015\012Host: $uhost\015\012\015\012");
734	$state{handle}->push_read (line => $qr_nlnl, sub {
735	$_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix
736	or return (%state = (), $cb->(undef, { Status => 599, Reason => "Invalid proxy connect response ($_[1])", @pseudo }));
737
738	if ($2 == 200) {
739	$rpath = $upath;
740	&$handle_actual_request;	1230	$handle_actual_request->();
741	} else {	1231
742	%state = ();
743	$cb->(undef, { Status => $2, Reason => $3, @pseudo });
744	}
745	});
746	} else {	1232	} else {
747	&$handle_actual_request;	1233	my $tcp_connect = $arg{tcp_connect}
		1234	|| do { require AnyEvent::Socket; \&AnyEvent::Socket::tcp_connect };
		1235
		1236	$state{connect_guard} = $tcp_connect->($rhost, $rport, $connect_cb, $arg{on_prepare} || sub { $timeout });
748	}	1237	}
749
750	}, $arg{on_prepare} || sub { $timeout };
751	};	1238	};
752		1239
753	defined wantarray && AnyEvent::Util::guard { %state = () }	1240	defined wantarray && AnyEvent::Util::guard { _destroy_state %state }
754	}	1241	}
755		1242
756	sub http_get($@) {	1243	sub http_get($@) {
757	unshift @_, "GET";	1244	unshift @_, "GET";
758	&http_request	1245	&http_request
…		…
776	AnyEvent::HTTP uses the AnyEvent::Socket::tcp_connect function for	1263	AnyEvent::HTTP uses the AnyEvent::Socket::tcp_connect function for
777	the actual connection, which in turn uses AnyEvent::DNS to resolve	1264	the actual connection, which in turn uses AnyEvent::DNS to resolve
778	hostnames. The latter is a simple stub resolver and does no caching	1265	hostnames. The latter is a simple stub resolver and does no caching
779	on its own. If you want DNS caching, you currently have to provide	1266	on its own. If you want DNS caching, you currently have to provide
780	your own default resolver (by storing a suitable resolver object in	1267	your own default resolver (by storing a suitable resolver object in
781	C<$AnyEvent::DNS::RESOLVER>).	1268	C<$AnyEvent::DNS::RESOLVER>) or your own C<tcp_connect> callback.
782		1269
783	=head2 GLOBAL FUNCTIONS AND VARIABLES	1270	=head2 GLOBAL FUNCTIONS AND VARIABLES
784		1271
785	=over 4	1272	=over 4
786		1273
787	=item AnyEvent::HTTP::set_proxy "proxy-url"	1274	=item AnyEvent::HTTP::set_proxy "proxy-url"
788		1275
789	Sets the default proxy server to use. The proxy-url must begin with a	1276	Sets the default proxy server to use. The proxy-url must begin with a
790	string of the form C<http://host:port> (optionally C<https:...>), croaks	1277	string of the form C<http://host:port>, croaks otherwise.
791	otherwise.
792		1278
793	To clear an already-set proxy, use C<undef>.	1279	To clear an already-set proxy, use C<undef>.
794		1280
		1281	When AnyEvent::HTTP is loaded for the first time it will query the
		1282	default proxy from the operating system, currently by looking at
		1283	C<$ENV{http_proxy>}.
		1284
		1285	=item AnyEvent::HTTP::cookie_jar_expire $jar[, $session_end]
		1286
		1287	Remove all cookies from the cookie jar that have been expired. If
		1288	C<$session_end> is given and true, then additionally remove all session
		1289	cookies.
		1290
		1291	You should call this function (with a true C<$session_end>) before you
		1292	save cookies to disk, and you should call this function after loading them
		1293	again. If you have a long-running program you can additionally call this
		1294	function from time to time.
		1295
		1296	A cookie jar is initially an empty hash-reference that is managed by this
		1297	module. Its format is subject to change, but currently it is as follows:
		1298
		1299	The key C<version> has to contain C<2>, otherwise the hash gets
		1300	cleared. All other keys are hostnames or IP addresses pointing to
		1301	hash-references. The key for these inner hash references is the
		1302	server path for which this cookie is meant, and the values are again
		1303	hash-references. Each key of those hash-references is a cookie name, and
		1304	the value, you guessed it, is another hash-reference, this time with the
		1305	key-value pairs from the cookie, except for C<expires> and C<max-age>,
		1306	which have been replaced by a C<_expires> key that contains the cookie
		1307	expiry timestamp. Session cookies are indicated by not having an
		1308	C<_expires> key.
		1309
		1310	Here is an example of a cookie jar with a single cookie, so you have a
		1311	chance of understanding the above paragraph:
		1312
		1313	{
		1314	version => 2,
		1315	"10.0.0.1" => {
		1316	"/" => {
		1317	"mythweb_id" => {
		1318	_expires => 1293917923,
		1319	value => "ooRung9dThee3ooyXooM1Ohm",
		1320	},
		1321	},
		1322	},
		1323	}
		1324
		1325	=item $date = AnyEvent::HTTP::format_date $timestamp
		1326
		1327	Takes a POSIX timestamp (seconds since the epoch) and formats it as a HTTP
		1328	Date (RFC 2616).
		1329
		1330	=item $timestamp = AnyEvent::HTTP::parse_date $date
		1331
		1332	Takes a HTTP Date (RFC 2616) or a Cookie date (netscape cookie spec) or a
		1333	bunch of minor variations of those, and returns the corresponding POSIX
		1334	timestamp, or C<undef> if the date cannot be parsed.
		1335
795	=item $AnyEvent::HTTP::MAX_RECURSE	1336	=item $AnyEvent::HTTP::MAX_RECURSE
796		1337
797	The default value for the C<recurse> request parameter (default: C<10>).	1338	The default value for the C<recurse> request parameter (default: C<10>).
		1339
		1340	=item $AnyEvent::HTTP::TIMEOUT
		1341
		1342	The default timeout for connection operations (default: C<300>).
798		1343
799	=item $AnyEvent::HTTP::USERAGENT	1344	=item $AnyEvent::HTTP::USERAGENT
800		1345
801	The default value for the C<User-Agent> header (the default is	1346	The default value for the C<User-Agent> header (the default is
802	C<Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)>).	1347	C<Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)>).
803		1348
804	=item $AnyEvent::HTTP::MAX_PER_HOST	1349	=item $AnyEvent::HTTP::MAX_PER_HOST
805		1350
806	The maximum number of concurrent connections to the same host (identified	1351	The maximum number of concurrent connections to the same host (identified
807	by the hostname). If the limit is exceeded, then the additional requests	1352	by the hostname). If the limit is exceeded, then additional requests
808	are queued until previous connections are closed.	1353	are queued until previous connections are closed. Both persistent and
		1354	non-persistent connections are counted in this limit.
809		1355
810	The default value for this is C<4>, and it is highly advisable to not	1356	The default value for this is C<4>, and it is highly advisable to not
811	increase it.	1357	increase it much.
		1358
		1359	For comparison: the RFC's recommend 4 non-persistent or 2 persistent
		1360	connections, older browsers used 2, newer ones (such as firefox 3)
		1361	typically use 6, and Opera uses 8 because like, they have the fastest
		1362	browser and give a shit for everybody else on the planet.
		1363
		1364	=item $AnyEvent::HTTP::PERSISTENT_TIMEOUT
		1365
		1366	The time after which idle persistent connections get closed by
		1367	AnyEvent::HTTP (default: C<3>).
812		1368
813	=item $AnyEvent::HTTP::ACTIVE	1369	=item $AnyEvent::HTTP::ACTIVE
814		1370
815	The number of active connections. This is not the number of currently	1371	The number of active connections. This is not the number of currently
816	running requests, but the number of currently open and non-idle TCP	1372	running requests, but the number of currently open and non-idle TCP
817	connections. This number of can be useful for load-leveling.	1373	connections. This number can be useful for load-leveling.
818		1374
819	=back	1375	=back
820		1376
821	=cut	1377	=cut
		1378
		1379	our @month = qw(Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec);
		1380	our @weekday = qw(Sun Mon Tue Wed Thu Fri Sat);
		1381
		1382	sub format_date($) {
		1383	my ($time) = @_;
		1384
		1385	# RFC 822/1123 format
		1386	my ($S, $M, $H, $mday, $mon, $year, $wday, $yday, undef) = gmtime $time;
		1387
		1388	sprintf "%s, %02d %s %04d %02d:%02d:%02d GMT",
		1389	$weekday[$wday], $mday, $month[$mon], $year + 1900,
		1390	$H, $M, $S;
		1391	}
		1392
		1393	sub parse_date($) {
		1394	my ($date) = @_;
		1395
		1396	my ($d, $m, $y, $H, $M, $S);
		1397
		1398	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$/) {
		1399	# RFC 822/1123, required by RFC 2616 (with " ")
		1400	# cookie dates (with "-")
		1401
		1402	($d, $m, $y, $H, $M, $S) = ($1, $2, $3, $4, $5, $6);
		1403
		1404	} elsif ($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]?) GMT$/) {
		1405	# RFC 850
		1406	($d, $m, $y, $H, $M, $S) = ($1, $2, $3 < 69 ? $3 + 2000 : $3 + 1900, $4, $5, $6);
		1407
		1408	} 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])$/) {
		1409	# ISO C's asctime
		1410	($d, $m, $y, $H, $M, $S) = ($2, $1, $6, $3, $4, $5);
		1411	}
		1412	# other formats fail in the loop below
		1413
		1414	for (0..11) {
		1415	if ($m eq $month[$_]) {
		1416	require Time::Local;
		1417	return eval { Time::Local::timegm ($S, $M, $H, $d, $_, $y) };
		1418	}
		1419	}
		1420
		1421	undef
		1422	}
822		1423
823	sub set_proxy($) {	1424	sub set_proxy($) {
824	if (length $_[0]) {	1425	if (length $_[0]) {
825	$_[0] =~ m%^(https?):// ([^:/]+) (?: : (\d*) )?%ix	1426	$_[0] =~ m%^(http):// ([^:/]+) (?: : (\d*) )?%ix
826	or Carp::croak "$_[0]: invalid proxy URL";	1427	or Carp::croak "$_[0]: invalid proxy URL";
827	$PROXY = [$2, $3 || 3128, $1]	1428	$PROXY = [$2, $3 || 3128, $1]
828	} else {	1429	} else {
829	undef $PROXY;	1430	undef $PROXY;
830	}	1431	}
…		…
833	# initialise proxy from environment	1434	# initialise proxy from environment
834	eval {	1435	eval {
835	set_proxy $ENV{http_proxy};	1436	set_proxy $ENV{http_proxy};
836	};	1437	};
837		1438
		1439	=head2 SHOWCASE
		1440
		1441	This section contains some more elaborate "real-world" examples or code
		1442	snippets.
		1443
		1444	=head2 HTTP/1.1 FILE DOWNLOAD
		1445
		1446	Downloading files with HTTP can be quite tricky, especially when something
		1447	goes wrong and you want to resume.
		1448
		1449	Here is a function that initiates and resumes a download. It uses the
		1450	last modified time to check for file content changes, and works with many
		1451	HTTP/1.0 servers as well, and usually falls back to a complete re-download
		1452	on older servers.
		1453
		1454	It calls the completion callback with either C<undef>, which means a
		1455	nonretryable error occurred, C<0> when the download was partial and should
		1456	be retried, and C<1> if it was successful.
		1457
		1458	use AnyEvent::HTTP;
		1459
		1460	sub download($$$) {
		1461	my ($url, $file, $cb) = @_;
		1462
		1463	open my $fh, "+<", $file
		1464	or die "$file: $!";
		1465
		1466	my %hdr;
		1467	my $ofs = 0;
		1468
		1469	if (stat $fh and -s _) {
		1470	$ofs = -s _;
		1471	warn "-s is ", $ofs;
		1472	$hdr{"if-unmodified-since"} = AnyEvent::HTTP::format_date +(stat _)[9];
		1473	$hdr{"range"} = "bytes=$ofs-";
		1474	}
		1475
		1476	http_get $url,
		1477	headers => \%hdr,
		1478	on_header => sub {
		1479	my ($hdr) = @_;
		1480
		1481	if ($hdr->{Status} == 200 && $ofs) {
		1482	# resume failed
		1483	truncate $fh, $ofs = 0;
		1484	}
		1485
		1486	sysseek $fh, $ofs, 0;
		1487
		1488	1
		1489	},
		1490	on_body => sub {
		1491	my ($data, $hdr) = @_;
		1492
		1493	if ($hdr->{Status} =~ /^2/) {
		1494	length $data == syswrite $fh, $data
		1495	or return; # abort on write errors
		1496	}
		1497
		1498	1
		1499	},
		1500	sub {
		1501	my (undef, $hdr) = @_;
		1502
		1503	my $status = $hdr->{Status};
		1504
		1505	if (my $time = AnyEvent::HTTP::parse_date $hdr->{"last-modified"}) {
		1506	utime $time, $time, $fh;
		1507	}
		1508
		1509	if ($status == 200 || $status == 206 || $status == 416) {
		1510	# download ok || resume ok || file already fully downloaded
		1511	$cb->(1, $hdr);
		1512
		1513	} elsif ($status == 412) {
		1514	# file has changed while resuming, delete and retry
		1515	unlink $file;
		1516	$cb->(0, $hdr);
		1517
		1518	} elsif ($status == 500 or $status == 503 or $status =~ /^59/) {
		1519	# retry later
		1520	$cb->(0, $hdr);
		1521
		1522	} else {
		1523	$cb->(undef, $hdr);
		1524	}
		1525	}
		1526	;
		1527	}
		1528
		1529	download "http://server/somelargefile", "/tmp/somelargefile", sub {
		1530	if ($_[0]) {
		1531	print "OK!\n";
		1532	} elsif (defined $_[0]) {
		1533	print "please retry later\n";
		1534	} else {
		1535	print "ERROR\n";
		1536	}
		1537	};
		1538
		1539	=head3 SOCKS PROXIES
		1540
		1541	Socks proxies are not directly supported by AnyEvent::HTTP. You can
		1542	compile your perl to support socks, or use an external program such as
		1543	F<socksify> (dante) or F<tsocks> to make your program use a socks proxy
		1544	transparently.
		1545
		1546	Alternatively, for AnyEvent::HTTP only, you can use your own
		1547	C<tcp_connect> function that does the proxy handshake - here is an example
		1548	that works with socks4a proxies:
		1549
		1550	use Errno;
		1551	use AnyEvent::Util;
		1552	use AnyEvent::Socket;
		1553	use AnyEvent::Handle;
		1554
		1555	# host, port and username of/for your socks4a proxy
		1556	my $socks_host = "10.0.0.23";
		1557	my $socks_port = 9050;
		1558	my $socks_user = "";
		1559
		1560	sub socks4a_connect {
		1561	my ($host, $port, $connect_cb, $prepare_cb) = @_;
		1562
		1563	my $hdl = new AnyEvent::Handle
		1564	connect => [$socks_host, $socks_port],
		1565	on_prepare => sub { $prepare_cb->($_[0]{fh}) },
		1566	on_error => sub { $connect_cb->() },
		1567	;
		1568
		1569	$hdl->push_write (pack "CCnNZ*Z*", 4, 1, $port, 1, $socks_user, $host);
		1570
		1571	$hdl->push_read (chunk => 8, sub {
		1572	my ($hdl, $chunk) = @_;
		1573	my ($status, $port, $ipn) = unpack "xCna4", $chunk;
		1574
		1575	if ($status == 0x5a) {
		1576	$connect_cb->($hdl->{fh}, (format_address $ipn) . ":$port");
		1577	} else {
		1578	$! = Errno::ENXIO; $connect_cb->();
		1579	}
		1580	});
		1581
		1582	$hdl
		1583	}
		1584
		1585	Use C<socks4a_connect> instead of C<tcp_connect> when doing C<http_request>s,
		1586	possibly after switching off other proxy types:
		1587
		1588	AnyEvent::HTTP::set_proxy undef; # usually you do not want other proxies
		1589
		1590	http_get 'http://www.google.com', tcp_connect => \&socks4a_connect, sub {
		1591	my ($data, $headers) = @_;
		1592	...
		1593	};
		1594
838	=head1 SEE ALSO	1595	=head1 SEE ALSO
839		1596
840	L<AnyEvent>.	1597	L<AnyEvent>.
841		1598
842	=head1 AUTHOR	1599	=head1 AUTHOR

Diff Legend

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