ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/AnyEvent-HTTP/HTTP.pm
(Generate patch)

Comparing AnyEvent-HTTP/HTTP.pm (file contents):
Revision 1.13 by root, Thu Jun 5 16:43:45 2008 UTC vs.
Revision 1.62 by root, Thu Dec 30 04:30:24 2010 UTC

3AnyEvent::HTTP - simple but non-blocking HTTP/HTTPS client 3AnyEvent::HTTP - simple but non-blocking HTTP/HTTPS client
4 4
5=head1 SYNOPSIS 5=head1 SYNOPSIS
6 6
7 use AnyEvent::HTTP; 7 use AnyEvent::HTTP;
8
9 http_get "http://www.nethype.de/", sub { print $_[1] };
10
11 # ... do something else here
8 12
9=head1 DESCRIPTION 13=head1 DESCRIPTION
10 14
11This module is an L<AnyEvent> user, you need to make sure that you use and 15This module is an L<AnyEvent> user, you need to make sure that you use and
12run a supported event loop. 16run a supported event loop.
35package AnyEvent::HTTP; 39package AnyEvent::HTTP;
36 40
37use strict; 41use strict;
38no warnings; 42no warnings;
39 43
40use Carp; 44use Errno ();
41 45
42use AnyEvent (); 46use AnyEvent 5.0 ();
43use AnyEvent::Util (); 47use AnyEvent::Util ();
44use AnyEvent::Socket ();
45use AnyEvent::Handle (); 48use AnyEvent::Handle ();
46 49
47use base Exporter::; 50use base Exporter::;
48 51
49our $VERSION = '1.0'; 52our $VERSION = '1.46';
50 53
51our @EXPORT = qw(http_get http_request); 54our @EXPORT = qw(http_get http_post http_head http_request);
52 55
53our $USERAGENT = "Mozilla/5.0 (compatible; AnyEvent::HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)"; 56our $USERAGENT = "Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)";
54our $MAX_RECURSE = 10; 57our $MAX_RECURSE = 10;
55our $MAX_PERSISTENT = 8; 58our $MAX_PERSISTENT = 8;
56our $PERSISTENT_TIMEOUT = 2; 59our $PERSISTENT_TIMEOUT = 2;
57our $TIMEOUT = 300; 60our $TIMEOUT = 300;
58 61
59# changing these is evil 62# changing these is evil
60our $MAX_PERSISTENT_PER_HOST = 2; 63our $MAX_PERSISTENT_PER_HOST = 0;
61our $MAX_PER_HOST = 4; 64our $MAX_PER_HOST = 4;
62 65
63our $PROXY; 66our $PROXY;
67our $ACTIVE = 0;
64 68
65my %KA_COUNT; # number of open keep-alive connections per host 69my %KA_COUNT; # number of open keep-alive connections per host
66my %CO_SLOT; # number of open connections, and wait queue, per host 70my %CO_SLOT; # number of open connections, and wait queue, per host
67 71
68=item http_get $url, key => value..., $cb->($data, $headers) 72=item http_get $url, key => value..., $cb->($data, $headers)
69 73
70Executes an HTTP-GET request. See the http_request function for details on 74Executes an HTTP-GET request. See the http_request function for details on
71additional parameters. 75additional parameters and the return value.
72 76
73=item http_head $url, key => value..., $cb->($data, $headers) 77=item http_head $url, key => value..., $cb->($data, $headers)
74 78
75Executes an HTTP-HEAD request. See the http_request function for details on 79Executes an HTTP-HEAD request. See the http_request function for details
76additional parameters. 80on additional parameters and the return value.
77 81
78=item http_post $url, $body, key => value..., $cb->($data, $headers) 82=item http_post $url, $body, key => value..., $cb->($data, $headers)
79 83
80Executes an HTTP-POST request with a request body of C<$bod>. See the 84Executes an HTTP-POST request with a request body of C<$body>. See the
81http_request function for details on additional parameters. 85http_request function for details on additional parameters and the return
86value.
82 87
83=item http_request $method => $url, key => value..., $cb->($data, $headers) 88=item http_request $method => $url, key => value..., $cb->($data, $headers)
84 89
85Executes a HTTP request of type C<$method> (e.g. C<GET>, C<POST>). The URL 90Executes a HTTP request of type C<$method> (e.g. C<GET>, C<POST>). The URL
86must be an absolute http or https URL. 91must be an absolute http or https URL.
87 92
93When called in void context, nothing is returned. In other contexts,
94C<http_request> returns a "cancellation guard" - you have to keep the
95object at least alive until the callback get called. If the object gets
96destroyed before the callback is called, the request will be cancelled.
97
88The callback will be called with the response data as first argument 98The callback will be called with the response body data as first argument
89(or C<undef> if it wasn't available due to errors), and a hash-ref with 99(or C<undef> if an error occured), and a hash-ref with response headers as
90response headers as second argument. 100second argument.
91 101
92All the headers in that hash are lowercased. In addition to the response 102All the headers in that hash are lowercased. In addition to the response
93headers, the three "pseudo-headers" C<HTTPVersion>, C<Status> and 103headers, the "pseudo-headers" (uppercase to avoid clashing with possible
104response headers) C<HTTPVersion>, C<Status> and C<Reason> contain the
94C<Reason> contain the three parts of the HTTP Status-Line of the same 105three parts of the HTTP Status-Line of the same name.
106
107The pseudo-header C<URL> contains the actual URL (which can differ from
108the requested URL when following redirects - for example, you might get
109an error that your URL scheme is not supported even though your URL is a
110valid http URL because it redirected to an ftp URL, in which case you can
111look at the URL pseudo header).
112
113The pseudo-header C<Redirect> only exists when the request was a result
114of an internal redirect. In that case it is an array reference with
115the C<($data, $headers)> from the redirect response. Note that this
116response could in turn be the result of a redirect itself, and C<<
117$headers->{Redirect}[1]{Redirect} >> will then contain the original
118response, and so on.
119
95name. If the server sends a header multiple lines, then their contents 120If the server sends a header multiple times, then their contents will be
96will be joined together with C<\x00>. 121joined together with a comma (C<,>), as per the HTTP spec.
97 122
98If an internal error occurs, such as not being able to resolve a hostname, 123If an internal error occurs, such as not being able to resolve a hostname,
99then C<$data> will be C<undef>, C<< $headers->{Status} >> will be C<599> 124then C<$data> will be C<undef>, C<< $headers->{Status} >> will be C<59x>
100and the C<Reason> pseudo-header will contain an error message. 125(usually C<599>) and the C<Reason> pseudo-header will contain an error
126message.
101 127
102A typical callback might look like this: 128A typical callback might look like this:
103 129
104 sub { 130 sub {
105 my ($body, $hdr) = @_; 131 my ($body, $hdr) = @_;
121Whether to recurse requests or not, e.g. on redirects, authentication 147Whether to recurse requests or not, e.g. on redirects, authentication
122retries and so on, and how often to do so. 148retries and so on, and how often to do so.
123 149
124=item headers => hashref 150=item headers => hashref
125 151
126The request headers to use. Currently, C<http_request> may provide its 152The request headers to use, with the header name (I<MUST be in lowercase>)
127own C<Host:>, C<Content-Length:>, C<Connection:> and C<Cookie:> headers 153as key and header value as hash value.
128and will provide defaults for C<User-Agent:> and C<Referer:>. 154
155Currently, http_request> may provide its own C<host>, C<content-length>,
156C<connection> and C<cookie> headers and will provide defaults for
157C<user-agent> and C<referer> (this can be suppressed by using a value of
158C<undef> for these headers in which case they won't be sent at all).
129 159
130=item timeout => $seconds 160=item timeout => $seconds
131 161
132The time-out to use for various stages - each connect attempt will reset 162The time-out to use for various stages - each connect attempt will reset
133the timeout, as will read or write activity. Default timeout is 5 minutes. 163the timeout, as will read or write activity, i.e. this is not an overall
164timeout.
165
166Default timeout is 5 minutes.
134 167
135=item proxy => [$host, $port[, $scheme]] or undef 168=item proxy => [$host, $port[, $scheme]] or undef
136 169
137Use the given http proxy for all requests. If not specified, then the 170Use the given http proxy for all requests. If not specified, then the
138default proxy (as specified by C<$ENV{http_proxy}>) is used. 171default proxy (as specified by C<$ENV{http_proxy}>) is used.
139 172
140C<$scheme> must be either missing or C<http> for HTTP, or C<https> for 173C<$scheme> must be either missing, C<http> for HTTP or C<https> for
141HTTPS. 174HTTPS.
142 175
143=item body => $string 176=item body => $string
144 177
145The request body, usually empty. Will be-sent as-is (future versions of 178The request body, usually empty. Will be-sent as-is (future versions of
151based on the original netscape specification. 184based on the original netscape specification.
152 185
153The C<$hash_ref> must be an (initially empty) hash reference which will 186The C<$hash_ref> must be an (initially empty) hash reference which will
154get updated automatically. It is possible to save the cookie_jar to 187get updated automatically. It is possible to save the cookie_jar to
155persistent storage with something like JSON or Storable, but this is not 188persistent storage with something like JSON or Storable, but this is not
156recommended, as expire times are currently being ignored. 189recommended, as expiry times are currently being ignored.
157 190
158Note that this cookie implementation is not of very high quality, nor 191Note that this cookie implementation is not of very high quality, nor
159meant to be complete. If you want complete cookie management you have to 192meant to be complete. If you want complete cookie management you have to
160do that on your own. C<cookie_jar> is meant as a quick fix to get some 193do that on your own. C<cookie_jar> is meant as a quick fix to get some
161cookie-using sites working. Cookies are a privacy disaster, do not use 194cookie-using sites working. Cookies are a privacy disaster, do not use
162them unless required to. 195them unless required to.
196
197=item tls_ctx => $scheme | $tls_ctx
198
199Specifies the AnyEvent::TLS context to be used for https connections. This
200parameter follows the same rules as the C<tls_ctx> parameter to
201L<AnyEvent::Handle>, but additionally, the two strings C<low> or
202C<high> can be specified, which give you a predefined low-security (no
203verification, highest compatibility) and high-security (CA and common-name
204verification) TLS context.
205
206The default for this option is C<low>, which could be interpreted as "give
207me the page, no matter what".
208
209=item on_prepare => $callback->($fh)
210
211In rare cases you need to "tune" the socket before it is used to
212connect (for exmaple, to bind it on a given IP address). This parameter
213overrides the prepare callback passed to C<AnyEvent::Socket::tcp_connect>
214and behaves exactly the same way (e.g. it has to provide a
215timeout). See the description for the C<$prepare_cb> argument of
216C<AnyEvent::Socket::tcp_connect> for details.
217
218=item tcp_connect => $callback->($host, $service, $connect_cb, $prepare_cb)
219
220In even rarer cases you want total control over how AnyEvent::HTTP
221establishes connections. Normally it uses L<AnyEvent::Socket::tcp_connect>
222to do this, but you can provide your own C<tcp_connect> function -
223obviously, it has to follow the same calling conventions, except that it
224may always return a connection guard object.
225
226There are probably lots of weird uses for this function, starting from
227tracing the hosts C<http_request> actually tries to connect, to (inexact
228but fast) host => IP address caching or even socks protocol support.
229
230=item on_header => $callback->($headers)
231
232When specified, this callback will be called with the header hash as soon
233as headers have been successfully received from the remote server (not on
234locally-generated errors).
235
236It has to return either true (in which case AnyEvent::HTTP will continue),
237or false, in which case AnyEvent::HTTP will cancel the download (and call
238the finish callback with an error code of C<598>).
239
240This callback is useful, among other things, to quickly reject unwanted
241content, which, if it is supposed to be rare, can be faster than first
242doing a C<HEAD> request.
243
244Example: cancel the request unless the content-type is "text/html".
245
246 on_header => sub {
247 $_[0]{"content-type"} =~ /^text\/html\s*(?:;|$)/
248 },
249
250=item on_body => $callback->($partial_body, $headers)
251
252When specified, all body data will be passed to this callback instead of
253to the completion callback. The completion callback will get the empty
254string instead of the body data.
255
256It has to return either true (in which case AnyEvent::HTTP will continue),
257or false, in which case AnyEvent::HTTP will cancel the download (and call
258the completion callback with an error code of C<598>).
259
260This callback is useful when the data is too large to be held in memory
261(so the callback writes it to a file) or when only some information should
262be extracted, or when the body should be processed incrementally.
263
264It is usually preferred over doing your own body handling via
265C<want_body_handle>, but in case of streaming APIs, where HTTP is
266only used to create a connection, C<want_body_handle> is the better
267alternative, as it allows you to install your own event handler, reducing
268resource usage.
269
270=item want_body_handle => $enable
271
272When enabled (default is disabled), the behaviour of AnyEvent::HTTP
273changes considerably: after parsing the headers, and instead of
274downloading the body (if any), the completion callback will be
275called. Instead of the C<$body> argument containing the body data, the
276callback will receive the L<AnyEvent::Handle> object associated with the
277connection. In error cases, C<undef> will be passed. When there is no body
278(e.g. status C<304>), the empty string will be passed.
279
280The handle object might or might not be in TLS mode, might be connected to
281a proxy, be a persistent connection etc., and configured in unspecified
282ways. The user is responsible for this handle (it will not be used by this
283module anymore).
284
285This is useful with some push-type services, where, after the initial
286headers, an interactive protocol is used (typical example would be the
287push-style twitter API which starts a JSON/XML stream).
288
289If you think you need this, first have a look at C<on_body>, to see if
290that doesn't solve your problem in a better way.
163 291
164=back 292=back
165 293
166Example: make a simple HTTP GET request for http://www.nethype.de/ 294Example: make a simple HTTP GET request for http://www.nethype.de/
167 295
181 use Data::Dumper; 309 use Data::Dumper;
182 print Dumper $hdr; 310 print Dumper $hdr;
183 } 311 }
184 ; 312 ;
185 313
314Example: make another simple HTTP GET request, but immediately try to
315cancel it.
316
317 my $request = http_request GET => "http://www.nethype.de/", sub {
318 my ($body, $hdr) = @_;
319 print "$body\n";
320 };
321
322 undef $request;
323
186=cut 324=cut
187 325
188sub _slot_schedule; 326sub _slot_schedule;
189sub _slot_schedule($) { 327sub _slot_schedule($) {
190 my $host = shift; 328 my $host = shift;
191 329
192 while ($CO_SLOT{$host}[0] < $MAX_PER_HOST) { 330 while ($CO_SLOT{$host}[0] < $MAX_PER_HOST) {
193 if (my $cb = shift @{ $CO_SLOT{$host}[1] }) { 331 if (my $cb = shift @{ $CO_SLOT{$host}[1] }) {
194 # somebody wants that slot 332 # somebody wants that slot
195 ++$CO_SLOT{$host}[0]; 333 ++$CO_SLOT{$host}[0];
334 ++$ACTIVE;
196 335
197 $cb->(AnyEvent::Util::guard { 336 $cb->(AnyEvent::Util::guard {
337 --$ACTIVE;
198 --$CO_SLOT{$host}[0]; 338 --$CO_SLOT{$host}[0];
199 _slot_schedule $host; 339 _slot_schedule $host;
200 }); 340 });
201 } else { 341 } else {
202 # nobody wants the slot, maybe we can forget about it 342 # nobody wants the slot, maybe we can forget about it
211 push @{ $CO_SLOT{$_[0]}[1] }, $_[1]; 351 push @{ $CO_SLOT{$_[0]}[1] }, $_[1];
212 352
213 _slot_schedule $_[0]; 353 _slot_schedule $_[0];
214} 354}
215 355
356our $qr_nlnl = qr{(?<![^\012])\015?\012};
357
358our $TLS_CTX_LOW = { cache => 1, sslv2 => 1 };
359our $TLS_CTX_HIGH = { cache => 1, verify => 1, verify_peername => "https" };
360
216sub http_request($$$;@) { 361sub http_request($$@) {
217 my $cb = pop; 362 my $cb = pop;
218 my ($method, $url, %arg) = @_; 363 my ($method, $url, %arg) = @_;
219 364
220 my %hdr; 365 my %hdr;
366
367 $arg{tls_ctx} = $TLS_CTX_LOW if $arg{tls_ctx} eq "low" || !exists $arg{tls_ctx};
368 $arg{tls_ctx} = $TLS_CTX_HIGH if $arg{tls_ctx} eq "high";
221 369
222 $method = uc $method; 370 $method = uc $method;
223 371
224 if (my $hdr = $arg{headers}) { 372 if (my $hdr = $arg{headers}) {
225 while (my ($k, $v) = each %$hdr) { 373 while (my ($k, $v) = each %$hdr) {
226 $hdr{lc $k} = $v; 374 $hdr{lc $k} = $v;
227 } 375 }
228 } 376 }
229 377
378 # pseudo headers for all subsequent responses
379 my @pseudo = (URL => $url);
380 push @pseudo, Redirect => delete $arg{Redirect} if exists $arg{Redirect};
381
230 my $recurse = exists $arg{recurse} ? $arg{recurse} : $MAX_RECURSE; 382 my $recurse = exists $arg{recurse} ? delete $arg{recurse} : $MAX_RECURSE;
231 383
232 return $cb->(undef, { Status => 599, Reason => "recursion limit reached" }) 384 return $cb->(undef, { Status => 599, Reason => "Too many redirections", @pseudo })
233 if $recurse < 0; 385 if $recurse < 0;
234 386
235 my $proxy = $arg{proxy} || $PROXY; 387 my $proxy = $arg{proxy} || $PROXY;
236 my $timeout = $arg{timeout} || $TIMEOUT; 388 my $timeout = $arg{timeout} || $TIMEOUT;
237 389
238 $hdr{"user-agent"} ||= $USERAGENT;
239
240 my ($scheme, $authority, $upath, $query, $fragment) = 390 my ($uscheme, $uauthority, $upath, $query, $fragment) =
241 $url =~ m|(?:([^:/?#]+):)?(?://([^/?#]*))?([^?#]*)(?:\?([^#]*))?(?:#(.*))?|; 391 $url =~ m|(?:([^:/?#]+):)?(?://([^/?#]*))?([^?#]*)(?:(\?[^#]*))?(?:#(.*))?|;
242 392
243 $scheme = lc $scheme; 393 $uscheme = lc $uscheme;
244 394
245 my $uport = $scheme eq "http" ? 80 395 my $uport = $uscheme eq "http" ? 80
246 : $scheme eq "https" ? 443 396 : $uscheme eq "https" ? 443
247 : return $cb->(undef, { Status => 599, Reason => "only http and https URL schemes supported" }); 397 : return $cb->(undef, { Status => 599, Reason => "Only http and https URL schemes supported", @pseudo });
248 398
249 $hdr{referer} ||= "$scheme://$authority$upath"; # leave out fragment and query string, just a heuristic
250
251 $authority =~ /^(?: .*\@ )? ([^\@:]+) (?: : (\d+) )?$/x 399 $uauthority =~ /^(?: .*\@ )? ([^\@:]+) (?: : (\d+) )?$/x
252 or return $cb->(undef, { Status => 599, Reason => "unparsable URL" }); 400 or return $cb->(undef, { Status => 599, Reason => "Unparsable URL", @pseudo });
253 401
254 my $uhost = $1; 402 my $uhost = $1;
255 $uport = $2 if defined $2; 403 $uport = $2 if defined $2;
256 404
405 $hdr{host} = defined $2 ? "$uhost:$2" : "$uhost"
406 unless exists $hdr{host};
407
257 $uhost =~ s/^\[(.*)\]$/$1/; 408 $uhost =~ s/^\[(.*)\]$/$1/;
258 $upath .= "?$query" if length $query; 409 $upath .= $query if length $query;
259 410
260 $upath =~ s%^/?%/%; 411 $upath =~ s%^/?%/%;
261 412
262 # cookie processing 413 # cookie processing
263 if (my $jar = $arg{cookie_jar}) { 414 if (my $jar = $arg{cookie_jar}) {
264 %$jar = () if $jar->{version} < 1; 415 %$jar = () if $jar->{version} != 1;
265 416
266 my @cookie; 417 my @cookie;
267 418
268 while (my ($chost, $v) = each %$jar) { 419 while (my ($chost, $v) = each %$jar) {
420 if ($chost =~ /^\./) {
269 next unless $chost eq substr $uhost, -length $chost; 421 next unless $chost eq substr $uhost, -length $chost;
270 next unless $chost =~ /^\./; 422 } elsif ($chost =~ /\./) {
423 next unless $chost eq $uhost;
424 } else {
425 next;
426 }
271 427
272 while (my ($cpath, $v) = each %$v) { 428 while (my ($cpath, $v) = each %$v) {
273 next unless $cpath eq substr $upath, 0, length $cpath; 429 next unless $cpath eq substr $upath, 0, length $cpath;
274 430
275 while (my ($k, $v) = each %$v) { 431 while (my ($k, $v) = each %$v) {
276 next if $scheme ne "https" && exists $v->{secure}; 432 next if $uscheme ne "https" && exists $v->{secure};
433 my $value = $v->{value};
434 $value =~ s/([\\"])/\\$1/g;
277 push @cookie, "$k=$v->{value}"; 435 push @cookie, "$k=\"$value\"";
278 } 436 }
279 } 437 }
280 } 438 }
281 439
282 $hdr{cookie} = join "; ", @cookie 440 $hdr{cookie} = join "; ", @cookie
283 if @cookie; 441 if @cookie;
284 } 442 }
285 443
286 my ($rhost, $rport, $rpath); # request host, port, path 444 my ($rhost, $rport, $rscheme, $rpath); # request host, port, path
287 445
288 if ($proxy) { 446 if ($proxy) {
289 ($rhost, $rport, $scheme) = @$proxy; 447 ($rpath, $rhost, $rport, $rscheme) = ($url, @$proxy);
290 $rpath = $url; 448
449 $rscheme = "http" unless defined $rscheme;
450
451 # don't support https requests over https-proxy transport,
452 # can't be done with tls as spec'ed, unless you double-encrypt.
453 $rscheme = "http" if $uscheme eq "https" && $rscheme eq "https";
291 } else { 454 } else {
292 ($rhost, $rport, $rpath) = ($uhost, $uport, $upath); 455 ($rhost, $rport, $rscheme, $rpath) = ($uhost, $uport, $uscheme, $upath);
293 $hdr{host} = $uhost;
294 } 456 }
295 457
458 # leave out fragment and query string, just a heuristic
459 $hdr{referer} ||= "$uscheme://$uauthority$upath" unless exists $hdr{referer};
460 $hdr{"user-agent"} ||= $USERAGENT unless exists $hdr{"user-agent"};
461
296 $hdr{"content-length"} = length $arg{body}; 462 $hdr{"content-length"} = length $arg{body}
463 if length $arg{body} || $method ne "GET";
297 464
298 my %state = (connect_guard => 1); 465 my %state = (connect_guard => 1);
299 466
300 _get_slot $uhost, sub { 467 _get_slot $uhost, sub {
301 $state{slot_guard} = shift; 468 $state{slot_guard} = shift;
302 469
303 return unless $state{connect_guard}; 470 return unless $state{connect_guard};
304 471
305 $state{connect_guard} = AnyEvent::Socket::tcp_connect $rhost, $rport, sub { 472 my $tcp_connect = $arg{tcp_connect}
473 || do { require AnyEvent::Socket; \&AnyEvent::Socket::tcp_connect };
474
475 $state{connect_guard} = $tcp_connect->(
476 $rhost,
477 $rport,
478 sub {
306 $state{fh} = shift 479 $state{fh} = shift
307 or return $cb->(undef, { Status => 599, Reason => "$!" });
308
309 delete $state{connect_guard}; # reduce memory usage, save a tree
310
311 # get handle
312 $state{handle} = new AnyEvent::Handle
313 fh => $state{fh},
314 ($scheme eq "https" ? (tls => "connect") : ());
315
316 # limit the number of persistent connections
317 if ($KA_COUNT{$_[1]} < $MAX_PERSISTENT_PER_HOST) {
318 ++$KA_COUNT{$_[1]};
319 $state{handle}{ka_count_guard} = AnyEvent::Util::guard { --$KA_COUNT{$_[1]} };
320 $hdr{connection} = "keep-alive";
321 delete $hdr{connection}; # keep-alive not yet supported
322 } else {
323 delete $hdr{connection};
324 }
325
326 # (re-)configure handle
327 $state{handle}->timeout ($timeout);
328 $state{handle}->on_error (sub {
329 %state = ();
330 $cb->(undef, { Status => 599, Reason => "$!" });
331 });
332 $state{handle}->on_eof (sub {
333 %state = ();
334 $cb->(undef, { Status => 599, Reason => "unexpected end-of-file" });
335 });
336
337 # send request
338 $state{handle}->push_write (
339 "$method $rpath HTTP/1.0\015\012"
340 . (join "", map "$_: $hdr{$_}\015\012", keys %hdr)
341 . "\015\012"
342 . (delete $arg{body})
343 );
344
345 %hdr = (); # reduce memory usage, save a kitten
346
347 # status line
348 $state{handle}->push_read (line => qr/\015?\012/, sub {
349 $_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) \s+ ([^\015\012]+)/ix
350 or return (%state = (), $cb->(undef, { Status => 599, Reason => "invalid server response ($_[1])" }));
351
352 my %hdr = ( # response headers
353 HTTPVersion => "\x00$1",
354 Status => "\x00$2",
355 Reason => "\x00$3",
356 );
357
358 # headers, could be optimized a bit
359 $state{handle}->unshift_read (line => qr/\015?\012\015?\012/, sub {
360 for ("$_[1]\012") {
361 # we support spaces in field names, as lotus domino
362 # creates them.
363 $hdr{lc $1} .= "\x00$2"
364 while /\G
365 ([^:\000-\037]+):
366 [\011\040]*
367 ((?: [^\015\012]+ | \015?\012[\011\040] )*)
368 \015?\012
369 /gxc;
370
371 /\G$/
372 or return (%state = (), $cb->(undef, { Status => 599, Reason => "garbled response headers" }));
373 } 480 or do {
374 481 my $err = "$!";
375 substr $_, 0, 1, ""
376 for values %hdr;
377
378 my $finish = sub {
379 %state = (); 482 %state = ();
483 return $cb->(undef, { Status => 599, Reason => $err, @pseudo });
484 };
380 485
381 # set-cookie processing 486 pop; # free memory, save a tree
382 if ($arg{cookie_jar} && exists $hdr{"set-cookie"}) {
383 for (split /\x00/, $hdr{"set-cookie"}) {
384 my ($cookie, @arg) = split /;\s*/;
385 my ($name, $value) = split /=/, $cookie, 2;
386 my %kv = (value => $value, map { split /=/, $_, 2 } @arg);
387
388 my $cdom = (delete $kv{domain}) || $uhost;
389 my $cpath = (delete $kv{path}) || "/";
390
391 $cdom =~ s/^.?/./; # make sure it starts with a "."
392 487
393 next if $cdom =~ /\.$/; 488 return unless delete $state{connect_guard};
394 489
395 # this is not rfc-like and not netscape-like. go figure. 490 # get handle
396 my $ndots = $cdom =~ y/.//; 491 $state{handle} = new AnyEvent::Handle
397 next if $ndots < ($cdom =~ /\.[^.][^.]\.[^.][^.]$/ ? 3 : 2); 492 fh => $state{fh},
398 493 peername => $rhost,
494 tls_ctx => $arg{tls_ctx},
495 # these need to be reconfigured on keepalive handles
496 timeout => $timeout,
497 on_error => sub {
498 %state = ();
499 $cb->(undef, { Status => 599, Reason => $_[2], @pseudo });
500 },
501 on_eof => sub {
502 %state = ();
503 $cb->(undef, { Status => 599, Reason => "Unexpected end-of-file", @pseudo });
504 },
505 ;
506
507 # limit the number of persistent connections
508 # keepalive not yet supported
509# if ($KA_COUNT{$_[1]} < $MAX_PERSISTENT_PER_HOST) {
510# ++$KA_COUNT{$_[1]};
511# $state{handle}{ka_count_guard} = AnyEvent::Util::guard {
512# --$KA_COUNT{$_[1]}
513# };
514# $hdr{connection} = "keep-alive";
515# } else {
516 delete $hdr{connection};
517# }
518
519 $state{handle}->starttls ("connect") if $rscheme eq "https";
520
521 # handle actual, non-tunneled, request
522 my $handle_actual_request = sub {
523 $state{handle}->starttls ("connect") if $uscheme eq "https" && !exists $state{handle}{tls};
524
525 # send request
526 $state{handle}->push_write (
527 "$method $rpath HTTP/1.0\015\012"
528 . (join "", map "\u$_: $hdr{$_}\015\012", grep defined $hdr{$_}, keys %hdr)
529 . "\015\012"
530 . (delete $arg{body})
531 );
532
533 # return if error occured during push_write()
534 return unless %state;
535
536 %hdr = (); # reduce memory usage, save a kitten, also make it possible to re-use
537
538 # status line and headers
539 $state{handle}->push_read (line => $qr_nlnl, sub {
540 for ("$_[1]") {
541 y/\015//d; # weed out any \015, as they show up in the weirdest of places.
542
543 /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )? \015?\012/igxc
544 or return (%state = (), $cb->(undef, { Status => 599, Reason => "Invalid server response", @pseudo }));
545
546 push @pseudo,
547 HTTPVersion => $1,
548 Status => $2,
549 Reason => $3,
550 ;
551
552 # things seen, not parsed:
553 # p3pP="NON CUR OTPi OUR NOR UNI"
554
555 $hdr{lc $1} .= ",$2"
556 while /\G
557 ([^:\000-\037]*):
558 [\011\040]*
559 ((?: [^\012]+ | \012[\011\040] )*)
560 \012
561 /gxc;
562
563 /\G$/
564 or return (%state = (), $cb->(undef, { Status => 599, Reason => "Garbled response headers", @pseudo }));
565 }
566
567 # remove the "," prefix we added to all headers above
568 substr $_, 0, 1, ""
569 for values %hdr;
570
571 # patch in all pseudo headers
572 %hdr = (%hdr, @pseudo);
573
574 # redirect handling
575 # microsoft and other shitheads don't give a shit for following standards,
576 # try to support some common forms of broken Location headers.
577 if ($hdr{location} !~ /^(?: $ | [^:\/?\#]+ : )/x) {
578 $hdr{location} =~ s/^\.\/+//;
579
580 my $url = "$rscheme://$uhost:$uport";
581
582 unless ($hdr{location} =~ s/^\///) {
583 $url .= $upath;
584 $url =~ s/\/[^\/]*$//;
585 }
586
587 $hdr{location} = "$url/$hdr{location}";
588 }
589
590 my $redirect;
591
592 if ($recurse) {
593 my $status = $hdr{Status};
594
595 # industry standard is to redirect POST as GET for
596 # 301, 302 and 303, in contrast to http/1.0 and 1.1.
597 # also, the UA should ask the user for 301 and 307 and POST,
598 # industry standard seems to be to simply follow.
599 # we go with the industry standard.
600 if ($status == 301 or $status == 302 or $status == 303) {
601 # HTTP/1.1 is unclear on how to mutate the method
602 $method = "GET" unless $method eq "HEAD";
399 # store it 603 $redirect = 1;
400 $arg{cookie_jar}{version} = 1; 604 } elsif ($status == 307) {
401 $arg{cookie_jar}{$cdom}{$cpath}{$name} = \%kv; 605 $redirect = 1;
402 } 606 }
403 } 607 }
404 608
405 if ($_[1]{Status} =~ /^30[12]$/ && $recurse) { 609 my $finish = sub {
406 # microsoft and other assholes don't give a shit for following standards, 610 $state{handle}->destroy if $state{handle};
407 # try to support a common form of broken Location header. 611 %state = ();
408 $_[1]{location} =~ s%^/%$scheme://$uhost:$uport/%;
409 612
410 http_request ($method, $_[1]{location}, %arg, recurse => $recurse - 1, $cb); 613 # set-cookie processing
614 if ($arg{cookie_jar}) {
615 for ($_[1]{"set-cookie"}) {
616 # parse NAME=VALUE
617 my @kv;
618
619 while (/\G\s* ([^=;,[:space:]]+) \s*=\s* (?: "((?:[^\\"]+|\\.)*)" | ([^=;,[:space:]]*) )/gcxs) {
620 my $name = $1;
621 my $value = $3;
622
623 unless ($value) {
624 $value = $2;
625 $value =~ s/\\(.)/$1/gs;
626 }
627
628 push @kv, $name => $value;
629
630 last unless /\G\s*;/gc;
631 }
632
633 last unless @kv;
634
635 my $name = shift @kv;
636 my %kv = (value => shift @kv, @kv);
637
638 my $cdom;
639 my $cpath = (delete $kv{path}) || "/";
640
641 if (exists $kv{domain}) {
642 $cdom = delete $kv{domain};
643
644 $cdom =~ s/^\.?/./; # make sure it starts with a "."
645
646 next if $cdom =~ /\.$/;
647
648 # this is not rfc-like and not netscape-like. go figure.
649 my $ndots = $cdom =~ y/.//;
650 next if $ndots < ($cdom =~ /\.[^.][^.]\.[^.][^.]$/ ? 3 : 2);
651 } else {
652 $cdom = $uhost;
653 }
654
655 # store it
656 $arg{cookie_jar}{version} = 1;
657 $arg{cookie_jar}{$cdom}{$cpath}{$name} = \%kv;
658
659 redo if /\G\s*,/gc;
660 }
661 }
662
663 if ($redirect && exists $hdr{location}) {
664 # we ignore any errors, as it is very common to receive
665 # Content-Length != 0 but no actual body
666 # we also access %hdr, as $_[1] might be an erro
667 http_request (
668 $method => $hdr{location},
669 %arg,
670 recurse => $recurse - 1,
671 Redirect => \@_,
672 $cb);
673 } else {
674 $cb->($_[0], $_[1]);
675 }
676 };
677
678 my $len = $hdr{"content-length"};
679
680 if (!$redirect && $arg{on_header} && !$arg{on_header}(\%hdr)) {
681 $finish->(undef, { Status => 598, Reason => "Request cancelled by on_header", @pseudo });
682 } elsif (
683 $hdr{Status} =~ /^(?:1..|[23]04)$/
684 or $method eq "HEAD"
685 or (defined $len && !$len)
686 ) {
687 # no body
688 $finish->("", \%hdr);
411 } else { 689 } else {
412 $cb->($_[0], $_[1]); 690 # body handling, four different code paths
691 # for want_body_handle, on_body (2x), normal (2x)
692 # we might read too much here, but it does not matter yet (no pers. connections)
693 if (!$redirect && $arg{want_body_handle}) {
694 $_[0]->on_eof (undef);
695 $_[0]->on_error (undef);
696 $_[0]->on_read (undef);
697
698 $finish->(delete $state{handle}, \%hdr);
699
700 } elsif ($arg{on_body}) {
701 $_[0]->on_error (sub { $finish->(undef, { Status => 599, Reason => $_[2], @pseudo }) });
702 if ($len) {
703 $_[0]->on_eof (undef);
704 $_[0]->on_read (sub {
705 $len -= length $_[0]{rbuf};
706
707 $arg{on_body}(delete $_[0]{rbuf}, \%hdr)
708 or $finish->(undef, { Status => 598, Reason => "Request cancelled by on_body", @pseudo });
709
710 $len > 0
711 or $finish->("", \%hdr);
712 });
713 } else {
714 $_[0]->on_eof (sub {
715 $finish->("", \%hdr);
716 });
717 $_[0]->on_read (sub {
718 $arg{on_body}(delete $_[0]{rbuf}, \%hdr)
719 or $finish->(undef, { Status => 598, Reason => "Request cancelled by on_body", @pseudo });
720 });
721 }
722 } else {
723 $_[0]->on_eof (undef);
724
725 if ($len) {
726 $_[0]->on_error (sub { $finish->(undef, { Status => 599, Reason => $_[2], @pseudo }) });
727 $_[0]->on_read (sub {
728 $finish->((substr delete $_[0]{rbuf}, 0, $len, ""), \%hdr)
729 if $len <= length $_[0]{rbuf};
730 });
731 } else {
732 $_[0]->on_error (sub {
733 ($! == Errno::EPIPE || !$!)
734 ? $finish->(delete $_[0]{rbuf}, \%hdr)
735 : $finish->(undef, { Status => 599, Reason => $_[2], @pseudo });
736 });
737 $_[0]->on_read (sub { });
738 }
739 }
413 } 740 }
414 }; 741 });
742 };
415 743
416 if ($hdr{Status} =~ /^(?:1..|204|304)$/ or $method eq "HEAD") { 744 # now handle proxy-CONNECT method
417 $finish->(undef, \%hdr); 745 if ($proxy && $uscheme eq "https") {
418 } else { 746 # oh dear, we have to wrap it into a connect request
419 if (exists $hdr{"content-length"}) {
420 $_[0]->unshift_read (chunk => $hdr{"content-length"}, sub {
421 # could cache persistent connection now
422 if ($hdr{connection} =~ /\bkeep-alive\b/i) {
423 # but we don't, due to misdesigns, this is annoyingly complex
424 };
425 747
426 $finish->($_[1], \%hdr); 748 # maybe re-use $uauthority with patched port?
427 }); 749 $state{handle}->push_write ("CONNECT $uhost:$uport HTTP/1.0\015\012Host: $uhost\015\012\015\012");
750 $state{handle}->push_read (line => $qr_nlnl, sub {
751 $_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix
752 or return (%state = (), $cb->(undef, { Status => 599, Reason => "Invalid proxy connect response ($_[1])", @pseudo }));
753
754 if ($2 == 200) {
755 $rpath = $upath;
756 &$handle_actual_request;
428 } else { 757 } else {
429 # too bad, need to read until we get an error or EOF,
430 # no way to detect winged data.
431 $_[0]->on_error (sub {
432 $finish->($_[0]{rbuf}, \%hdr);
433 }); 758 %state = ();
434 $_[0]->on_eof (undef); 759 $cb->(undef, { Status => $2, Reason => $3, @pseudo });
435 $_[0]->on_read (sub { });
436 } 760 }
437 } 761 });
762 } else {
763 &$handle_actual_request;
438 }); 764 }
765
439 }); 766 },
440 }, sub { 767 $arg{on_prepare} || sub { $timeout }
441 $timeout
442 }; 768 );
443 }; 769 };
444 770
445 defined wantarray && AnyEvent::Util::guard { %state = () } 771 defined wantarray && AnyEvent::Util::guard { %state = () }
446} 772}
447 773
448sub http_get($$;@) { 774sub http_get($@) {
449 unshift @_, "GET"; 775 unshift @_, "GET";
450 &http_request 776 &http_request
451} 777}
452 778
453sub http_head($$;@) { 779sub http_head($@) {
454 unshift @_, "HEAD"; 780 unshift @_, "HEAD";
455 &http_request 781 &http_request
456} 782}
457 783
458sub http_post($$$;@) { 784sub http_post($$@) {
785 my $url = shift;
459 unshift @_, "POST", "body"; 786 unshift @_, "POST", $url, "body";
460 &http_request 787 &http_request
461} 788}
462 789
463=back 790=back
464 791
792=head2 DNS CACHING
793
794AnyEvent::HTTP uses the AnyEvent::Socket::tcp_connect function for
795the actual connection, which in turn uses AnyEvent::DNS to resolve
796hostnames. The latter is a simple stub resolver and does no caching
797on its own. If you want DNS caching, you currently have to provide
798your own default resolver (by storing a suitable resolver object in
799C<$AnyEvent::DNS::RESOLVER>).
800
465=head2 GLOBAL FUNCTIONS AND VARIABLES 801=head2 GLOBAL FUNCTIONS AND VARIABLES
466 802
467=over 4 803=over 4
468 804
469=item AnyEvent::HTTP::set_proxy "proxy-url" 805=item AnyEvent::HTTP::set_proxy "proxy-url"
470 806
471Sets the default proxy server to use. The proxy-url must begin with a 807Sets the default proxy server to use. The proxy-url must begin with a
472string of the form C<http://host:port> (optionally C<https:...>). 808string of the form C<http://host:port> (optionally C<https:...>), croaks
809otherwise.
810
811To clear an already-set proxy, use C<undef>.
812
813=item $date = AnyEvent::HTTP::format_date $timestamp
814
815Takes a POSIX timestamp (seconds since the epoch) and formats it as a HTTP
816Date (RFC 2616).
817
818=item $timestamp = AnyEvent::HTTP::parse_date $date
819
820Takes a HTTP Date (RFC 2616) and returns the corresponding POSIX
821timestamp, or C<undef> if the date cannot be parsed.
473 822
474=item $AnyEvent::HTTP::MAX_RECURSE 823=item $AnyEvent::HTTP::MAX_RECURSE
475 824
476The default value for the C<recurse> request parameter (default: C<10>). 825The default value for the C<recurse> request parameter (default: C<10>).
477 826
478=item $AnyEvent::HTTP::USERAGENT 827=item $AnyEvent::HTTP::USERAGENT
479 828
480The default value for the C<User-Agent> header (the default is 829The default value for the C<User-Agent> header (the default is
481C<Mozilla/5.0 (compatible; AnyEvent::HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)>). 830C<Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)>).
482 831
483=item $AnyEvent::HTTP::MAX_PERSISTENT 832=item $AnyEvent::HTTP::MAX_PER_HOST
484 833
485The maximum number of persistent connections to keep open (default: 8). 834The maximum number of concurrent connections to the same host (identified
835by the hostname). If the limit is exceeded, then the additional requests
836are queued until previous connections are closed.
486 837
487Not implemented currently. 838The default value for this is C<4>, and it is highly advisable to not
839increase it.
488 840
489=item $AnyEvent::HTTP::PERSISTENT_TIMEOUT 841=item $AnyEvent::HTTP::ACTIVE
490 842
491The maximum time to cache a persistent connection, in seconds (default: 2). 843The number of active connections. This is not the number of currently
492 844running requests, but the number of currently open and non-idle TCP
493Not implemented currently. 845connections. This number of can be useful for load-leveling.
494 846
495=back 847=back
496 848
497=cut 849=cut
498 850
851our @month = qw(Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec);
852our @weekday = qw(Sun Mon Tue Wed Thu Fri Sat);
853
854sub format_date($) {
855 my ($time) = @_;
856
857 # RFC 822/1123 format
858 my ($S, $M, $H, $mday, $mon, $year, $wday, $yday, undef) = gmtime $time;
859
860 sprintf "%s, %02d %s %04d %02d:%02d:%02d GMT",
861 $weekday[$wday], $mday, $month[$mon], $year + 1900,
862 $H, $M, $S;
863}
864
865sub parse_date($) {
866 my ($date) = @_;
867
868 my ($d, $m, $y, $H, $M, $S);
869
870 if ($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$/) {
871 # RFC 822/1123, required by RFC 2616
872 ($d, $m, $y, $H, $M, $S) = ($1, $2, $3, $4, $5, $6);
873
874 } elsif ($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$/) {
875 # RFC 850
876 ($d, $m, $y, $H, $M, $S) = ($1, $2, $3 < 69 ? $3 + 2000 : $3 + 1900, $4, $5, $6);
877
878 } elsif ($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])$/) {
879 # ISO C's asctime
880 ($d, $m, $y, $H, $M, $S) = ($2, $1, $6, $3, $4, $5);
881 }
882 # other formats fail in the loop below
883
884 for (0..11) {
885 if ($m eq $month[$_]) {
886 require Time::Local;
887 return Time::Local::timegm ($S, $M, $H, $d, $_, $y);
888 }
889 }
890
891 undef
892}
893
499sub set_proxy($) { 894sub set_proxy($) {
895 if (length $_[0]) {
500 $PROXY = [$2, $3 || 3128, $1] if $_[0] =~ m%^(https?):// ([^:/]+) (?: : (\d*) )?%ix; 896 $_[0] =~ m%^(https?):// ([^:/]+) (?: : (\d*) )?%ix
897 or Carp::croak "$_[0]: invalid proxy URL";
898 $PROXY = [$2, $3 || 3128, $1]
899 } else {
900 undef $PROXY;
901 }
501} 902}
502 903
503# initialise proxy from environment 904# initialise proxy from environment
905eval {
504set_proxy $ENV{http_proxy}; 906 set_proxy $ENV{http_proxy};
907};
908
909=head2 SOCKS PROXIES
910
911Socks proxies are not directly supported by AnyEvent::HTTP. You can
912compile your perl to support socks, or use an external program such as
913F<socksify> (dante) or F<tsocks> to make your program use a socks proxy
914transparently.
915
916Alternatively, for AnyEvent::HTTP only, you can use your own
917C<tcp_connect> function that does the proxy handshake - here is an example
918that works with socks4a proxies:
919
920 use Errno;
921 use AnyEvent::Util;
922 use AnyEvent::Socket;
923 use AnyEvent::Handle;
924
925 # host, port and username of/for your socks4a proxy
926 my $socks_host = "10.0.0.23";
927 my $socks_port = 9050;
928 my $socks_user = "";
929
930 sub socks4a_connect {
931 my ($host, $port, $connect_cb, $prepare_cb) = @_;
932
933 my $hdl = new AnyEvent::Handle
934 connect => [$socks_host, $socks_port],
935 on_prepare => sub { $prepare_cb->($_[0]{fh}) },
936 on_error => sub { $connect_cb->() },
937 ;
938
939 $hdl->push_write (pack "CCnNZ*Z*", 4, 1, $port, 1, $socks_user, $host);
940
941 $hdl->push_read (chunk => 8, sub {
942 my ($hdl, $chunk) = @_;
943 my ($status, $port, $ipn) = unpack "xCna4", $chunk;
944
945 if ($status == 0x5a) {
946 $connect_cb->($hdl->{fh}, (format_address $ipn) . ":$port");
947 } else {
948 $! = Errno::ENXIO; $connect_cb->();
949 }
950 });
951
952 $hdl
953 }
954
955Use C<socks4a_connect> instead of C<tcp_connect> when doing C<http_request>s,
956possibly after switching off other proxy types:
957
958 AnyEvent::HTTP::set_proxy undef; # usually you do not want other proxies
959
960 http_get 'http://www.google.com', tcp_connect => \&socks4a_connect, sub {
961 my ($data, $headers) = @_;
962 ...
963 };
505 964
506=head1 SEE ALSO 965=head1 SEE ALSO
507 966
508L<AnyEvent>. 967L<AnyEvent>.
509 968
510=head1 AUTHOR 969=head1 AUTHOR
511 970
512 Marc Lehmann <schmorp@schmorp.de> 971 Marc Lehmann <schmorp@schmorp.de>
513 http://home.schmorp.de/ 972 http://home.schmorp.de/
973
974With many thanks to Дмитрий Шалашов, who provided countless
975testcases and bugreports.
514 976
515=cut 977=cut
516 978
5171 9791
518 980

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines