… | |
… | |
41 | use strict; |
41 | use strict; |
42 | no warnings; |
42 | no warnings; |
43 | |
43 | |
44 | use Errno (); |
44 | use Errno (); |
45 | |
45 | |
46 | use AnyEvent 4.452 (); |
46 | use AnyEvent 5.0 (); |
47 | use AnyEvent::Util (); |
47 | use AnyEvent::Util (); |
48 | use AnyEvent::Socket (); |
|
|
49 | use AnyEvent::Handle (); |
48 | use AnyEvent::Handle (); |
50 | |
49 | |
51 | use base Exporter::; |
50 | use base Exporter::; |
52 | |
51 | |
53 | our $VERSION = '1.12'; |
52 | our $VERSION = '1.46'; |
54 | |
53 | |
55 | our @EXPORT = qw(http_get http_post http_head http_request); |
54 | our @EXPORT = qw(http_get http_post http_head http_request); |
56 | |
55 | |
57 | our $USERAGENT = "Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)"; |
56 | our $USERAGENT = "Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)"; |
58 | our $MAX_RECURSE = 10; |
57 | our $MAX_RECURSE = 10; |
59 | our $MAX_PERSISTENT = 8; |
58 | our $MAX_PERSISTENT = 8; |
60 | our $PERSISTENT_TIMEOUT = 2; |
59 | our $PERSISTENT_TIMEOUT = 2; |
61 | our $TIMEOUT = 300; |
60 | our $TIMEOUT = 300; |
62 | |
61 | |
63 | # changing these is evil |
62 | # changing these is evil |
64 | our $MAX_PERSISTENT_PER_HOST = 2; |
63 | our $MAX_PERSISTENT_PER_HOST = 0; |
65 | our $MAX_PER_HOST = 4; |
64 | our $MAX_PER_HOST = 4; |
66 | |
65 | |
67 | our $PROXY; |
66 | our $PROXY; |
68 | our $ACTIVE = 0; |
67 | our $ACTIVE = 0; |
69 | |
68 | |
… | |
… | |
92 | must be an absolute http or https URL. |
91 | must be an absolute http or https URL. |
93 | |
92 | |
94 | When called in void context, nothing is returned. In other contexts, |
93 | When called in void context, nothing is returned. In other contexts, |
95 | C<http_request> returns a "cancellation guard" - you have to keep the |
94 | C<http_request> returns a "cancellation guard" - you have to keep the |
96 | object at least alive until the callback get called. If the object gets |
95 | object at least alive until the callback get called. If the object gets |
97 | destroyed before the callbakc is called, the request will be cancelled. |
96 | destroyed before the callback is called, the request will be cancelled. |
98 | |
97 | |
99 | The callback will be called with the response body data as first argument |
98 | 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 |
99 | (or C<undef> if an error occured), and a hash-ref with response headers as |
101 | second argument. |
100 | second argument. |
102 | |
101 | |
103 | All the headers in that hash are lowercased. In addition to the response |
102 | All the headers in that hash are lowercased. In addition to the response |
104 | headers, the "pseudo-headers" C<HTTPVersion>, C<Status> and C<Reason> |
103 | headers, the "pseudo-headers" (uppercase to avoid clashing with possible |
|
|
104 | response headers) C<HTTPVersion>, C<Status> and C<Reason> contain the |
105 | contain the three parts of the HTTP Status-Line of the same name. The |
105 | three parts of the HTTP Status-Line of the same name. |
|
|
106 | |
106 | pseudo-header C<URL> contains the original URL (which can differ from the |
107 | The pseudo-header C<URL> contains the actual URL (which can differ from |
107 | requested URL when following redirects). |
108 | the requested URL when following redirects - for example, you might get |
|
|
109 | an error that your URL scheme is not supported even though your URL is a |
|
|
110 | valid http URL because it redirected to an ftp URL, in which case you can |
|
|
111 | look at the URL pseudo header). |
|
|
112 | |
|
|
113 | The pseudo-header C<Redirect> only exists when the request was a result |
|
|
114 | of an internal redirect. In that case it is an array reference with |
|
|
115 | the C<($data, $headers)> from the redirect response. Note that this |
|
|
116 | response could in turn be the result of a redirect itself, and C<< |
|
|
117 | $headers->{Redirect}[1]{Redirect} >> will then contain the original |
|
|
118 | response, and so on. |
108 | |
119 | |
109 | If the server sends a header multiple times, then their contents will be |
120 | If the server sends a header multiple times, then their contents will be |
110 | joined together with a comma (C<,>), as per the HTTP spec. |
121 | joined together with a comma (C<,>), as per the HTTP spec. |
111 | |
122 | |
112 | If an internal error occurs, such as not being able to resolve a hostname, |
123 | If an internal error occurs, such as not being able to resolve a hostname, |
… | |
… | |
138 | |
149 | |
139 | =item headers => hashref |
150 | =item headers => hashref |
140 | |
151 | |
141 | The request headers to use. Currently, C<http_request> may provide its |
152 | The request headers to use. Currently, C<http_request> may provide its |
142 | own C<Host:>, C<Content-Length:>, C<Connection:> and C<Cookie:> headers |
153 | own C<Host:>, C<Content-Length:>, C<Connection:> and C<Cookie:> headers |
143 | and will provide defaults for C<User-Agent:> and C<Referer:>. |
154 | and will provide defaults for C<User-Agent:> and C<Referer:> (this can be |
|
|
155 | suppressed by using C<undef> for these headers in which case they won't be |
|
|
156 | sent at all). |
144 | |
157 | |
145 | =item timeout => $seconds |
158 | =item timeout => $seconds |
146 | |
159 | |
147 | The time-out to use for various stages - each connect attempt will reset |
160 | The time-out to use for various stages - each connect attempt will reset |
148 | the timeout, as will read or write activity. Default timeout is 5 minutes. |
161 | the timeout, as will read or write activity, i.e. this is not an overall |
|
|
162 | timeout. |
|
|
163 | |
|
|
164 | Default timeout is 5 minutes. |
149 | |
165 | |
150 | =item proxy => [$host, $port[, $scheme]] or undef |
166 | =item proxy => [$host, $port[, $scheme]] or undef |
151 | |
167 | |
152 | Use the given http proxy for all requests. If not specified, then the |
168 | Use the given http proxy for all requests. If not specified, then the |
153 | default proxy (as specified by C<$ENV{http_proxy}>) is used. |
169 | default proxy (as specified by C<$ENV{http_proxy}>) is used. |
154 | |
170 | |
155 | C<$scheme> must be either missing or C<http> for HTTP, or C<https> for |
171 | C<$scheme> must be either missing, C<http> for HTTP or C<https> for |
156 | HTTPS. |
172 | HTTPS. |
157 | |
173 | |
158 | =item body => $string |
174 | =item body => $string |
159 | |
175 | |
160 | The request body, usually empty. Will be-sent as-is (future versions of |
176 | The request body, usually empty. Will be-sent as-is (future versions of |
… | |
… | |
186 | verification) TLS context. |
202 | verification) TLS context. |
187 | |
203 | |
188 | The default for this option is C<low>, which could be interpreted as "give |
204 | The default for this option is C<low>, which could be interpreted as "give |
189 | me the page, no matter what". |
205 | me the page, no matter what". |
190 | |
206 | |
|
|
207 | =item on_prepare => $callback->($fh) |
|
|
208 | |
|
|
209 | In rare cases you need to "tune" the socket before it is used to |
|
|
210 | connect (for exmaple, to bind it on a given IP address). This parameter |
|
|
211 | overrides the prepare callback passed to C<AnyEvent::Socket::tcp_connect> |
|
|
212 | and behaves exactly the same way (e.g. it has to provide a |
|
|
213 | timeout). See the description for the C<$prepare_cb> argument of |
|
|
214 | C<AnyEvent::Socket::tcp_connect> for details. |
|
|
215 | |
|
|
216 | =item tcp_connect => $callback->($host, $service, $connect_cb, $prepare_cb) |
|
|
217 | |
|
|
218 | In even rarer cases you want total control over how AnyEvent::HTTP |
|
|
219 | establishes connections. Normally it uses L<AnyEvent::Socket::tcp_connect> |
|
|
220 | to do this, but you can provide your own C<tcp_connect> function - |
|
|
221 | obviously, it has to follow the same calling conventions, except that it |
|
|
222 | may always return a connection guard object. |
|
|
223 | |
|
|
224 | There are probably lots of weird uses for this function, starting from |
|
|
225 | tracing the hosts C<http_request> actually tries to connect, to (inexact |
|
|
226 | but fast) host => IP address caching or even socks protocol support. |
|
|
227 | |
191 | =item on_header => $callback->($headers) |
228 | =item on_header => $callback->($headers) |
192 | |
229 | |
193 | When specified, this callback will be called with the header hash as soon |
230 | When specified, this callback will be called with the header hash as soon |
194 | as headers have been successfully received from the remote server (not on |
231 | as headers have been successfully received from the remote server (not on |
195 | locally-generated errors). |
232 | locally-generated errors). |
… | |
… | |
221 | This callback is useful when the data is too large to be held in memory |
258 | This callback is useful when the data is too large to be held in memory |
222 | (so the callback writes it to a file) or when only some information should |
259 | (so the callback writes it to a file) or when only some information should |
223 | be extracted, or when the body should be processed incrementally. |
260 | be extracted, or when the body should be processed incrementally. |
224 | |
261 | |
225 | It is usually preferred over doing your own body handling via |
262 | It is usually preferred over doing your own body handling via |
226 | C<want_body_handle>. |
263 | C<want_body_handle>, but in case of streaming APIs, where HTTP is |
|
|
264 | only used to create a connection, C<want_body_handle> is the better |
|
|
265 | alternative, as it allows you to install your own event handler, reducing |
|
|
266 | resource usage. |
227 | |
267 | |
228 | =item want_body_handle => $enable |
268 | =item want_body_handle => $enable |
229 | |
269 | |
230 | When enabled (default is disabled), the behaviour of AnyEvent::HTTP |
270 | When enabled (default is disabled), the behaviour of AnyEvent::HTTP |
231 | changes considerably: after parsing the headers, and instead of |
271 | changes considerably: after parsing the headers, and instead of |
… | |
… | |
243 | This is useful with some push-type services, where, after the initial |
283 | This is useful with some push-type services, where, after the initial |
244 | headers, an interactive protocol is used (typical example would be the |
284 | headers, an interactive protocol is used (typical example would be the |
245 | push-style twitter API which starts a JSON/XML stream). |
285 | push-style twitter API which starts a JSON/XML stream). |
246 | |
286 | |
247 | If you think you need this, first have a look at C<on_body>, to see if |
287 | If you think you need this, first have a look at C<on_body>, to see if |
248 | that doesn'T solve your problem in a better way. |
288 | that doesn't solve your problem in a better way. |
249 | |
289 | |
250 | =back |
290 | =back |
251 | |
291 | |
252 | Example: make a simple HTTP GET request for http://www.nethype.de/ |
292 | Example: make a simple HTTP GET request for http://www.nethype.de/ |
253 | |
293 | |
… | |
… | |
309 | push @{ $CO_SLOT{$_[0]}[1] }, $_[1]; |
349 | push @{ $CO_SLOT{$_[0]}[1] }, $_[1]; |
310 | |
350 | |
311 | _slot_schedule $_[0]; |
351 | _slot_schedule $_[0]; |
312 | } |
352 | } |
313 | |
353 | |
314 | our $qr_nl = qr<\015?\012>; |
354 | our $qr_nlnl = qr{(?<![^\012])\015?\012}; |
315 | our $qr_nlnl = qr<\015?\012\015?\012>; |
|
|
316 | |
355 | |
317 | our $TLS_CTX_LOW = { cache => 1, sslv2 => 1 }; |
356 | our $TLS_CTX_LOW = { cache => 1, sslv2 => 1 }; |
318 | our $TLS_CTX_HIGH = { cache => 1, verify => 1, verify_peername => "https" }; |
357 | our $TLS_CTX_HIGH = { cache => 1, verify => 1, verify_peername => "https" }; |
319 | |
358 | |
320 | sub http_request($$@) { |
359 | sub http_request($$@) { |
… | |
… | |
332 | while (my ($k, $v) = each %$hdr) { |
371 | while (my ($k, $v) = each %$hdr) { |
333 | $hdr{lc $k} = $v; |
372 | $hdr{lc $k} = $v; |
334 | } |
373 | } |
335 | } |
374 | } |
336 | |
375 | |
|
|
376 | # pseudo headers for all subsequent responses |
|
|
377 | my @pseudo = (URL => $url); |
|
|
378 | push @pseudo, Redirect => delete $arg{Redirect} if exists $arg{Redirect}; |
|
|
379 | |
337 | my $recurse = exists $arg{recurse} ? delete $arg{recurse} : $MAX_RECURSE; |
380 | my $recurse = exists $arg{recurse} ? delete $arg{recurse} : $MAX_RECURSE; |
338 | |
381 | |
339 | return $cb->(undef, { Status => 599, Reason => "Too many redirections", URL => $url }) |
382 | return $cb->(undef, { Status => 599, Reason => "Too many redirections", @pseudo }) |
340 | if $recurse < 0; |
383 | if $recurse < 0; |
341 | |
384 | |
342 | my $proxy = $arg{proxy} || $PROXY; |
385 | my $proxy = $arg{proxy} || $PROXY; |
343 | my $timeout = $arg{timeout} || $TIMEOUT; |
386 | my $timeout = $arg{timeout} || $TIMEOUT; |
344 | |
387 | |
345 | my ($uscheme, $uauthority, $upath, $query, $fragment) = |
388 | my ($uscheme, $uauthority, $upath, $query, $fragment) = |
346 | $url =~ m|(?:([^:/?#]+):)?(?://([^/?#]*))?([^?#]*)(?:\?([^#]*))?(?:#(.*))?|; |
389 | $url =~ m|(?:([^:/?#]+):)?(?://([^/?#]*))?([^?#]*)(?:(\?[^#]*))?(?:#(.*))?|; |
347 | |
390 | |
348 | $uscheme = lc $uscheme; |
391 | $uscheme = lc $uscheme; |
349 | |
392 | |
350 | my $uport = $uscheme eq "http" ? 80 |
393 | my $uport = $uscheme eq "http" ? 80 |
351 | : $uscheme eq "https" ? 443 |
394 | : $uscheme eq "https" ? 443 |
352 | : return $cb->(undef, { Status => 599, Reason => "Only http and https URL schemes supported", URL => $url }); |
395 | : return $cb->(undef, { Status => 599, Reason => "Only http and https URL schemes supported", @pseudo }); |
353 | |
396 | |
354 | $uauthority =~ /^(?: .*\@ )? ([^\@:]+) (?: : (\d+) )?$/x |
397 | $uauthority =~ /^(?: .*\@ )? ([^\@:]+) (?: : (\d+) )?$/x |
355 | or return $cb->(undef, { Status => 599, Reason => "Unparsable URL", URL => $url }); |
398 | or return $cb->(undef, { Status => 599, Reason => "Unparsable URL", @pseudo }); |
356 | |
399 | |
357 | my $uhost = $1; |
400 | my $uhost = $1; |
358 | $uport = $2 if defined $2; |
401 | $uport = $2 if defined $2; |
359 | |
402 | |
|
|
403 | $hdr{host} = defined $2 ? "$uhost:$2" : "$uhost" |
|
|
404 | unless exists $hdr{host}; |
|
|
405 | |
360 | $uhost =~ s/^\[(.*)\]$/$1/; |
406 | $uhost =~ s/^\[(.*)\]$/$1/; |
361 | $upath .= "?$query" if length $query; |
407 | $upath .= $query if length $query; |
362 | |
408 | |
363 | $upath =~ s%^/?%/%; |
409 | $upath =~ s%^/?%/%; |
364 | |
410 | |
365 | # cookie processing |
411 | # cookie processing |
366 | if (my $jar = $arg{cookie_jar}) { |
412 | if (my $jar = $arg{cookie_jar}) { |
… | |
… | |
396 | my ($rhost, $rport, $rscheme, $rpath); # request host, port, path |
442 | my ($rhost, $rport, $rscheme, $rpath); # request host, port, path |
397 | |
443 | |
398 | if ($proxy) { |
444 | if ($proxy) { |
399 | ($rpath, $rhost, $rport, $rscheme) = ($url, @$proxy); |
445 | ($rpath, $rhost, $rport, $rscheme) = ($url, @$proxy); |
400 | |
446 | |
|
|
447 | $rscheme = "http" unless defined $rscheme; |
|
|
448 | |
401 | # don't support https requests over https-proxy transport, |
449 | # don't support https requests over https-proxy transport, |
402 | # can't be done with tls as spec'ed, unless you double-encrypt. |
450 | # can't be done with tls as spec'ed, unless you double-encrypt. |
403 | $rscheme = "http" if $uscheme eq "https" && $rscheme eq "https"; |
451 | $rscheme = "http" if $uscheme eq "https" && $rscheme eq "https"; |
404 | } else { |
452 | } else { |
405 | ($rhost, $rport, $rscheme, $rpath) = ($uhost, $uport, $uscheme, $upath); |
453 | ($rhost, $rport, $rscheme, $rpath) = ($uhost, $uport, $uscheme, $upath); |
406 | } |
454 | } |
407 | |
455 | |
408 | $hdr{"user-agent"} ||= $USERAGENT; |
456 | # leave out fragment and query string, just a heuristic |
409 | $hdr{referer} ||= "$uscheme://$uauthority$upath"; # leave out fragment and query string, just a heuristic |
457 | $hdr{referer} ||= "$uscheme://$uauthority$upath" unless exists $hdr{referer}; |
|
|
458 | $hdr{"user-agent"} ||= $USERAGENT unless exists $hdr{"user-agent"}; |
410 | |
459 | |
411 | $hdr{host} = "$uhost:$uport"; |
|
|
412 | $hdr{"content-length"} = length $arg{body}; |
460 | $hdr{"content-length"} = length $arg{body} |
|
|
461 | if length $arg{body} || $method ne "GET"; |
413 | |
462 | |
414 | my %state = (connect_guard => 1); |
463 | my %state = (connect_guard => 1); |
415 | |
464 | |
416 | _get_slot $uhost, sub { |
465 | _get_slot $uhost, sub { |
417 | $state{slot_guard} = shift; |
466 | $state{slot_guard} = shift; |
418 | |
467 | |
419 | return unless $state{connect_guard}; |
468 | return unless $state{connect_guard}; |
420 | |
469 | |
421 | $state{connect_guard} = AnyEvent::Socket::tcp_connect $rhost, $rport, sub { |
470 | my $tcp_connect = $arg{tcp_connect} |
|
|
471 | || do { require AnyEvent::Socket; \&AnyEvent::Socket::tcp_connect }; |
|
|
472 | |
|
|
473 | $state{connect_guard} = $tcp_connect->( |
|
|
474 | $rhost, |
|
|
475 | $rport, |
|
|
476 | sub { |
422 | $state{fh} = shift |
477 | $state{fh} = shift |
|
|
478 | or do { |
|
|
479 | my $err = "$!"; |
|
|
480 | %state = (); |
423 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "$!", URL => $url })); |
481 | return $cb->(undef, { Status => 599, Reason => $err, @pseudo }); |
|
|
482 | }; |
|
|
483 | |
424 | pop; # free memory, save a tree |
484 | pop; # free memory, save a tree |
425 | |
485 | |
426 | return unless delete $state{connect_guard}; |
486 | return unless delete $state{connect_guard}; |
427 | |
487 | |
428 | # get handle |
488 | # get handle |
429 | $state{handle} = new AnyEvent::Handle |
489 | $state{handle} = new AnyEvent::Handle |
430 | fh => $state{fh}, |
490 | fh => $state{fh}, |
|
|
491 | peername => $rhost, |
|
|
492 | tls_ctx => $arg{tls_ctx}, |
|
|
493 | # these need to be reconfigured on keepalive handles |
431 | timeout => $timeout, |
494 | timeout => $timeout, |
432 | peername => $rhost, |
495 | on_error => sub { |
433 | tls_ctx => $arg{tls_ctx}; |
496 | %state = (); |
|
|
497 | $cb->(undef, { Status => 599, Reason => $_[2], @pseudo }); |
|
|
498 | }, |
|
|
499 | on_eof => sub { |
|
|
500 | %state = (); |
|
|
501 | $cb->(undef, { Status => 599, Reason => "Unexpected end-of-file", @pseudo }); |
|
|
502 | }, |
|
|
503 | ; |
434 | |
504 | |
435 | # limit the number of persistent connections |
505 | # limit the number of persistent connections |
436 | # keepalive not yet supported |
506 | # keepalive not yet supported |
437 | if ($KA_COUNT{$_[1]} < $MAX_PERSISTENT_PER_HOST) { |
507 | # if ($KA_COUNT{$_[1]} < $MAX_PERSISTENT_PER_HOST) { |
438 | ++$KA_COUNT{$_[1]}; |
508 | # ++$KA_COUNT{$_[1]}; |
439 | $state{handle}{ka_count_guard} = AnyEvent::Util::guard { |
509 | # $state{handle}{ka_count_guard} = AnyEvent::Util::guard { |
440 | --$KA_COUNT{$_[1]} |
510 | # --$KA_COUNT{$_[1]} |
441 | }; |
511 | # }; |
442 | $hdr{connection} = "keep-alive"; |
512 | # $hdr{connection} = "keep-alive"; |
443 | } else { |
513 | # } else { |
444 | delete $hdr{connection}; |
514 | delete $hdr{connection}; |
445 | } |
515 | # } |
446 | |
516 | |
447 | # (re-)configure handle |
|
|
448 | $state{handle}->on_error (sub { |
|
|
449 | %state = (); |
|
|
450 | $cb->(undef, { Status => 599, Reason => $_[2], URL => $url }); |
|
|
451 | }); |
|
|
452 | $state{handle}->on_eof (sub { |
|
|
453 | %state = (); |
|
|
454 | $cb->(undef, { Status => 599, Reason => "Unexpected end-of-file", URL => $url }); |
|
|
455 | }); |
|
|
456 | |
|
|
457 | $state{handle}->starttls ("connect") if $rscheme eq "https"; |
517 | $state{handle}->starttls ("connect") if $rscheme eq "https"; |
458 | |
518 | |
459 | # handle actual, non-tunneled, request |
519 | # handle actual, non-tunneled, request |
460 | my $handle_actual_request = sub { |
520 | my $handle_actual_request = sub { |
461 | $state{handle}->starttls ("connect") if $uscheme eq "https" && !exists $state{handle}{tls}; |
521 | $state{handle}->starttls ("connect") if $uscheme eq "https" && !exists $state{handle}{tls}; |
462 | |
522 | |
463 | # send request |
523 | # send request |
464 | $state{handle}->push_write ( |
524 | $state{handle}->push_write ( |
465 | "$method $rpath HTTP/1.0\015\012" |
525 | "$method $rpath HTTP/1.0\015\012" |
466 | . (join "", map "\u$_: $hdr{$_}\015\012", keys %hdr) |
526 | . (join "", map "\u$_: $hdr{$_}\015\012", grep defined $hdr{$_}, keys %hdr) |
467 | . "\015\012" |
527 | . "\015\012" |
468 | . (delete $arg{body}) |
528 | . (delete $arg{body}) |
469 | ); |
|
|
470 | |
|
|
471 | %hdr = (); # reduce memory usage, save a kitten |
|
|
472 | |
|
|
473 | # status line |
|
|
474 | $state{handle}->push_read (line => $qr_nl, sub { |
|
|
475 | $_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix |
|
|
476 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "Invalid server response ($_[1])", URL => $url })); |
|
|
477 | |
|
|
478 | my %hdr = ( # response headers |
|
|
479 | HTTPVersion => ",$1", |
|
|
480 | Status => ",$2", |
|
|
481 | Reason => ",$3", |
|
|
482 | URL => ",$url" |
|
|
483 | ); |
529 | ); |
484 | |
530 | |
485 | # headers, could be optimized a bit |
531 | # return if error occured during push_write() |
|
|
532 | return unless %state; |
|
|
533 | |
|
|
534 | %hdr = (); # reduce memory usage, save a kitten, also make it possible to re-use |
|
|
535 | |
|
|
536 | # status line and headers |
486 | $state{handle}->unshift_read (line => $qr_nlnl, sub { |
537 | $state{handle}->push_read (line => $qr_nlnl, sub { |
487 | for ("$_[1]\012") { |
538 | for ("$_[1]") { |
488 | y/\015//d; # weed out any \015, as they show up in the weirdest of places. |
539 | y/\015//d; # weed out any \015, as they show up in the weirdest of places. |
|
|
540 | |
|
|
541 | /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )? \015?\012/igxc |
|
|
542 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "Invalid server response", @pseudo })); |
|
|
543 | |
|
|
544 | push @pseudo, |
|
|
545 | HTTPVersion => $1, |
|
|
546 | Status => $2, |
|
|
547 | Reason => $3, |
|
|
548 | ; |
489 | |
549 | |
490 | # things seen, not parsed: |
550 | # things seen, not parsed: |
491 | # p3pP="NON CUR OTPi OUR NOR UNI" |
551 | # p3pP="NON CUR OTPi OUR NOR UNI" |
492 | |
552 | |
493 | $hdr{lc $1} .= ",$2" |
553 | $hdr{lc $1} .= ",$2" |
494 | while /\G |
554 | while /\G |
495 | ([^:\000-\037]+): |
555 | ([^:\000-\037]*): |
496 | [\011\040]* |
556 | [\011\040]* |
497 | ((?: [^\012]+ | \012[\011\040] )*) |
557 | ((?: [^\012]+ | \012[\011\040] )*) |
498 | \012 |
558 | \012 |
499 | /gxc; |
559 | /gxc; |
500 | |
560 | |
501 | /\G$/ |
561 | /\G$/ |
502 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "Garbled response headers", URL => $url })); |
562 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "Garbled response headers", @pseudo })); |
503 | } |
563 | } |
504 | |
564 | |
|
|
565 | # remove the "," prefix we added to all headers above |
505 | substr $_, 0, 1, "" |
566 | substr $_, 0, 1, "" |
506 | for values %hdr; |
567 | for values %hdr; |
|
|
568 | |
|
|
569 | # patch in all pseudo headers |
|
|
570 | %hdr = (%hdr, @pseudo); |
507 | |
571 | |
508 | # redirect handling |
572 | # redirect handling |
509 | # microsoft and other shitheads don't give a shit for following standards, |
573 | # microsoft and other shitheads don't give a shit for following standards, |
510 | # try to support some common forms of broken Location headers. |
574 | # try to support some common forms of broken Location headers. |
511 | if ($hdr{location} !~ /^(?: $ | [^:\/?\#]+ : )/x) { |
575 | if ($hdr{location} !~ /^(?: $ | [^:\/?\#]+ : )/x) { |
… | |
… | |
522 | } |
586 | } |
523 | |
587 | |
524 | my $redirect; |
588 | my $redirect; |
525 | |
589 | |
526 | if ($recurse) { |
590 | if ($recurse) { |
527 | if ($hdr{Status} =~ /^30[12]$/ && $method ne "POST") { |
591 | my $status = $hdr{Status}; |
528 | # apparently, mozilla et al. just change POST to GET here |
592 | |
529 | # more research is needed before we do the same |
593 | # industry standard is to redirect POST as GET for |
530 | $redirect = 1; |
594 | # 301, 302 and 303, in contrast to http/1.0 and 1.1. |
531 | } elsif ($hdr{Status} == 303) { |
595 | # also, the UA should ask the user for 301 and 307 and POST, |
|
|
596 | # industry standard seems to be to simply follow. |
|
|
597 | # we go with the industry standard. |
|
|
598 | if ($status == 301 or $status == 302 or $status == 303) { |
532 | # even http/1.1 is unclear on how to mutate the method |
599 | # HTTP/1.1 is unclear on how to mutate the method |
533 | $method = "GET" unless $method eq "HEAD"; |
600 | $method = "GET" unless $method eq "HEAD"; |
534 | $redirect = 1; |
601 | $redirect = 1; |
535 | } elsif ($hdr{Status} == 307 && $method =~ /^(?:GET|HEAD)$/) { |
602 | } elsif ($status == 307) { |
536 | $redirect = 1; |
603 | $redirect = 1; |
537 | } |
604 | } |
538 | } |
605 | } |
539 | |
606 | |
540 | my $finish = sub { |
607 | my $finish = sub { |
… | |
… | |
589 | |
656 | |
590 | redo if /\G\s*,/gc; |
657 | redo if /\G\s*,/gc; |
591 | } |
658 | } |
592 | } |
659 | } |
593 | |
660 | |
594 | if ($redirect) { |
661 | if ($redirect && exists $hdr{location}) { |
595 | # we ignore any errors, as it is very common to receive |
662 | # we ignore any errors, as it is very common to receive |
596 | # Content-Length != 0 but no actual body |
663 | # Content-Length != 0 but no actual body |
597 | # we also access %hdr, as $_[1] might be an erro |
664 | # we also access %hdr, as $_[1] might be an erro |
598 | http_request ($method => $hdr{location}, %arg, recurse => $recurse - 1, $cb); |
665 | http_request ( |
|
|
666 | $method => $hdr{location}, |
|
|
667 | %arg, |
|
|
668 | recurse => $recurse - 1, |
|
|
669 | Redirect => \@_, |
|
|
670 | $cb); |
599 | } else { |
671 | } else { |
600 | $cb->($_[0], $_[1]); |
672 | $cb->($_[0], $_[1]); |
601 | } |
673 | } |
602 | }; |
674 | }; |
603 | |
675 | |
604 | my $len = $hdr{"content-length"}; |
676 | my $len = $hdr{"content-length"}; |
605 | |
677 | |
606 | if (!$redirect && $arg{on_header} && !$arg{on_header}(\%hdr)) { |
678 | if (!$redirect && $arg{on_header} && !$arg{on_header}(\%hdr)) { |
607 | $finish->(undef, { Status => 598, Reason => "Request cancelled by on_header", URL => $url }); |
679 | $finish->(undef, { Status => 598, Reason => "Request cancelled by on_header", @pseudo }); |
608 | } elsif ( |
680 | } elsif ( |
609 | $hdr{Status} =~ /^(?:1..|[23]04)$/ |
681 | $hdr{Status} =~ /^(?:1..|[23]04)$/ |
610 | or $method eq "HEAD" |
682 | or $method eq "HEAD" |
611 | or (defined $len && !$len) |
683 | or (defined $len && !$len) |
612 | ) { |
684 | ) { |
… | |
… | |
615 | } else { |
687 | } else { |
616 | # body handling, four different code paths |
688 | # body handling, four different code paths |
617 | # for want_body_handle, on_body (2x), normal (2x) |
689 | # for want_body_handle, on_body (2x), normal (2x) |
618 | # we might read too much here, but it does not matter yet (no pers. connections) |
690 | # we might read too much here, but it does not matter yet (no pers. connections) |
619 | if (!$redirect && $arg{want_body_handle}) { |
691 | if (!$redirect && $arg{want_body_handle}) { |
620 | $_[0]->on_eof (undef); |
692 | $_[0]->on_eof (undef); |
621 | $_[0]->on_error (undef); |
693 | $_[0]->on_error (undef); |
622 | $_[0]->on_read (undef); |
694 | $_[0]->on_read (undef); |
623 | |
695 | |
624 | $finish->(delete $state{handle}, \%hdr); |
696 | $finish->(delete $state{handle}, \%hdr); |
625 | |
697 | |
626 | } elsif ($arg{on_body}) { |
698 | } elsif ($arg{on_body}) { |
627 | $_[0]->on_error (sub { $finish->(undef, { Status => 599, Reason => $_[2], URL => $url }) }); |
699 | $_[0]->on_error (sub { $finish->(undef, { Status => 599, Reason => $_[2], @pseudo }) }); |
628 | if ($len) { |
700 | if ($len) { |
629 | $_[0]->on_eof (undef); |
701 | $_[0]->on_eof (undef); |
630 | $_[0]->on_read (sub { |
702 | $_[0]->on_read (sub { |
631 | $len -= length $_[0]{rbuf}; |
703 | $len -= length $_[0]{rbuf}; |
632 | |
704 | |
633 | $arg{on_body}(delete $_[0]{rbuf}, \%hdr) |
705 | $arg{on_body}(delete $_[0]{rbuf}, \%hdr) |
634 | or $finish->(undef, { Status => 598, Reason => "Request cancelled by on_body", URL => $url }); |
706 | or $finish->(undef, { Status => 598, Reason => "Request cancelled by on_body", @pseudo }); |
635 | |
707 | |
636 | $len > 0 |
708 | $len > 0 |
637 | or $finish->("", \%hdr); |
709 | or $finish->("", \%hdr); |
638 | }); |
710 | }); |
639 | } else { |
711 | } else { |
640 | $_[0]->on_eof (sub { |
712 | $_[0]->on_eof (sub { |
641 | $finish->("", \%hdr); |
713 | $finish->("", \%hdr); |
642 | }); |
714 | }); |
643 | $_[0]->on_read (sub { |
715 | $_[0]->on_read (sub { |
644 | $arg{on_body}(delete $_[0]{rbuf}, \%hdr) |
716 | $arg{on_body}(delete $_[0]{rbuf}, \%hdr) |
645 | or $finish->(undef, { Status => 598, Reason => "Request cancelled by on_body", URL => $url }); |
717 | or $finish->(undef, { Status => 598, Reason => "Request cancelled by on_body", @pseudo }); |
646 | }); |
718 | }); |
647 | } |
719 | } |
648 | } else { |
720 | } else { |
649 | $_[0]->on_eof (undef); |
721 | $_[0]->on_eof (undef); |
650 | |
722 | |
651 | if ($len) { |
723 | if ($len) { |
652 | $_[0]->on_error (sub { $finish->(undef, { Status => 599, Reason => $_[2], URL => $url }) }); |
724 | $_[0]->on_error (sub { $finish->(undef, { Status => 599, Reason => $_[2], @pseudo }) }); |
653 | $_[0]->on_read (sub { |
725 | $_[0]->on_read (sub { |
654 | $finish->((substr delete $_[0]{rbuf}, 0, $len, ""), \%hdr) |
726 | $finish->((substr delete $_[0]{rbuf}, 0, $len, ""), \%hdr) |
655 | if $len <= length $_[0]{rbuf}; |
727 | if $len <= length $_[0]{rbuf}; |
656 | }); |
728 | }); |
657 | } else { |
729 | } else { |
658 | $_[0]->on_error (sub { |
730 | $_[0]->on_error (sub { |
659 | $! == Errno::EPIPE |
731 | ($! == Errno::EPIPE || !$!) |
660 | ? $finish->(delete $_[0]{rbuf}, \%hdr) |
732 | ? $finish->(delete $_[0]{rbuf}, \%hdr) |
661 | : $finish->(undef, { Status => 599, Reason => $_[2], URL => $url }); |
733 | : $finish->(undef, { Status => 599, Reason => $_[2], @pseudo }); |
662 | }); |
734 | }); |
663 | $_[0]->on_read (sub { }); |
735 | $_[0]->on_read (sub { }); |
664 | } |
736 | } |
665 | } |
737 | } |
666 | } |
738 | } |
667 | }); |
739 | }); |
668 | }); |
740 | }; |
669 | }; |
|
|
670 | |
741 | |
671 | # now handle proxy-CONNECT method |
742 | # now handle proxy-CONNECT method |
672 | if ($proxy && $uscheme eq "https") { |
743 | if ($proxy && $uscheme eq "https") { |
673 | # oh dear, we have to wrap it into a connect request |
744 | # oh dear, we have to wrap it into a connect request |
674 | |
745 | |
675 | # maybe re-use $uauthority with patched port? |
746 | # maybe re-use $uauthority with patched port? |
676 | $state{handle}->push_write ("CONNECT $uhost:$uport HTTP/1.0\015\012Host: $uhost\015\012\015\012"); |
747 | $state{handle}->push_write ("CONNECT $uhost:$uport HTTP/1.0\015\012Host: $uhost\015\012\015\012"); |
677 | $state{handle}->push_read (line => $qr_nlnl, sub { |
748 | $state{handle}->push_read (line => $qr_nlnl, sub { |
678 | $_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix |
749 | $_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix |
679 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "Invalid proxy connect response ($_[1])", URL => $url })); |
750 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "Invalid proxy connect response ($_[1])", @pseudo })); |
680 | |
751 | |
681 | if ($2 == 200) { |
752 | if ($2 == 200) { |
682 | $rpath = $upath; |
753 | $rpath = $upath; |
683 | &$handle_actual_request; |
754 | &$handle_actual_request; |
684 | } else { |
755 | } else { |
685 | %state = (); |
756 | %state = (); |
686 | $cb->(undef, { Status => $2, Reason => $3, URL => $url }); |
757 | $cb->(undef, { Status => $2, Reason => $3, @pseudo }); |
|
|
758 | } |
687 | } |
759 | }); |
688 | }); |
|
|
689 | } else { |
760 | } else { |
690 | &$handle_actual_request; |
761 | &$handle_actual_request; |
|
|
762 | } |
|
|
763 | |
691 | } |
764 | }, |
692 | |
765 | $arg{on_prepare} || sub { $timeout } |
693 | }, sub { |
|
|
694 | $timeout |
|
|
695 | }; |
766 | ); |
696 | }; |
767 | }; |
697 | |
768 | |
698 | defined wantarray && AnyEvent::Util::guard { %state = () } |
769 | defined wantarray && AnyEvent::Util::guard { %state = () } |
699 | } |
770 | } |
700 | |
771 | |
… | |
… | |
714 | &http_request |
785 | &http_request |
715 | } |
786 | } |
716 | |
787 | |
717 | =back |
788 | =back |
718 | |
789 | |
|
|
790 | =head2 DNS CACHING |
|
|
791 | |
|
|
792 | AnyEvent::HTTP uses the AnyEvent::Socket::tcp_connect function for |
|
|
793 | the actual connection, which in turn uses AnyEvent::DNS to resolve |
|
|
794 | hostnames. The latter is a simple stub resolver and does no caching |
|
|
795 | on its own. If you want DNS caching, you currently have to provide |
|
|
796 | your own default resolver (by storing a suitable resolver object in |
|
|
797 | C<$AnyEvent::DNS::RESOLVER>). |
|
|
798 | |
719 | =head2 GLOBAL FUNCTIONS AND VARIABLES |
799 | =head2 GLOBAL FUNCTIONS AND VARIABLES |
720 | |
800 | |
721 | =over 4 |
801 | =over 4 |
722 | |
802 | |
723 | =item AnyEvent::HTTP::set_proxy "proxy-url" |
803 | =item AnyEvent::HTTP::set_proxy "proxy-url" |
724 | |
804 | |
725 | Sets the default proxy server to use. The proxy-url must begin with a |
805 | Sets the default proxy server to use. The proxy-url must begin with a |
726 | string of the form C<http://host:port> (optionally C<https:...>). |
806 | string of the form C<http://host:port> (optionally C<https:...>), croaks |
|
|
807 | otherwise. |
|
|
808 | |
|
|
809 | To clear an already-set proxy, use C<undef>. |
727 | |
810 | |
728 | =item $AnyEvent::HTTP::MAX_RECURSE |
811 | =item $AnyEvent::HTTP::MAX_RECURSE |
729 | |
812 | |
730 | The default value for the C<recurse> request parameter (default: C<10>). |
813 | The default value for the C<recurse> request parameter (default: C<10>). |
731 | |
814 | |
732 | =item $AnyEvent::HTTP::USERAGENT |
815 | =item $AnyEvent::HTTP::USERAGENT |
733 | |
816 | |
734 | The default value for the C<User-Agent> header (the default is |
817 | The default value for the C<User-Agent> header (the default is |
735 | C<Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)>). |
818 | C<Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)>). |
736 | |
819 | |
737 | =item $AnyEvent::HTTP::MAX_PERSISTENT |
820 | =item $AnyEvent::HTTP::MAX_PER_HOST |
738 | |
821 | |
739 | The maximum number of persistent connections to keep open (default: 8). |
822 | The maximum number of concurrent connections to the same host (identified |
|
|
823 | by the hostname). If the limit is exceeded, then the additional requests |
|
|
824 | are queued until previous connections are closed. |
740 | |
825 | |
741 | Not implemented currently. |
826 | The default value for this is C<4>, and it is highly advisable to not |
742 | |
827 | increase it. |
743 | =item $AnyEvent::HTTP::PERSISTENT_TIMEOUT |
|
|
744 | |
|
|
745 | The maximum time to cache a persistent connection, in seconds (default: 2). |
|
|
746 | |
|
|
747 | Not implemented currently. |
|
|
748 | |
828 | |
749 | =item $AnyEvent::HTTP::ACTIVE |
829 | =item $AnyEvent::HTTP::ACTIVE |
750 | |
830 | |
751 | The number of active connections. This is not the number of currently |
831 | The number of active connections. This is not the number of currently |
752 | running requests, but the number of currently open and non-idle TCP |
832 | running requests, but the number of currently open and non-idle TCP |
… | |
… | |
755 | =back |
835 | =back |
756 | |
836 | |
757 | =cut |
837 | =cut |
758 | |
838 | |
759 | sub set_proxy($) { |
839 | sub set_proxy($) { |
|
|
840 | if (length $_[0]) { |
760 | $PROXY = [$2, $3 || 3128, $1] if $_[0] =~ m%^(https?):// ([^:/]+) (?: : (\d*) )?%ix; |
841 | $_[0] =~ m%^(https?):// ([^:/]+) (?: : (\d*) )?%ix |
|
|
842 | or Carp::croak "$_[0]: invalid proxy URL"; |
|
|
843 | $PROXY = [$2, $3 || 3128, $1] |
|
|
844 | } else { |
|
|
845 | undef $PROXY; |
|
|
846 | } |
761 | } |
847 | } |
762 | |
848 | |
763 | # initialise proxy from environment |
849 | # initialise proxy from environment |
|
|
850 | eval { |
764 | set_proxy $ENV{http_proxy}; |
851 | set_proxy $ENV{http_proxy}; |
|
|
852 | }; |
|
|
853 | |
|
|
854 | =head2 SOCKS PROXIES |
|
|
855 | |
|
|
856 | Socks proxies are not directly supported by AnyEvent::HTTP. You can |
|
|
857 | compile your perl to support socks, or use an external program such as |
|
|
858 | F<socksify> (dante) or F<tsocks> to make your program use a socks proxy |
|
|
859 | transparently. |
|
|
860 | |
|
|
861 | Alternatively, for AnyEvent::HTTP only, you can use your own |
|
|
862 | C<tcp_connect> function that does the proxy handshake - here is an example |
|
|
863 | that works with socks4a proxies: |
|
|
864 | |
|
|
865 | use Errno; |
|
|
866 | use AnyEvent::Util; |
|
|
867 | use AnyEvent::Socket; |
|
|
868 | use AnyEvent::Handle; |
|
|
869 | |
|
|
870 | # host, port and username of/for your socks4a proxy |
|
|
871 | my $socks_host = "10.0.0.23"; |
|
|
872 | my $socks_port = 9050; |
|
|
873 | my $socks_user = ""; |
|
|
874 | |
|
|
875 | sub socks4a_connect { |
|
|
876 | my ($host, $port, $connect_cb, $prepare_cb) = @_; |
|
|
877 | |
|
|
878 | my $hdl = new AnyEvent::Handle |
|
|
879 | connect => [$socks_host, $socks_port], |
|
|
880 | on_prepare => sub { $prepare_cb->($_[0]{fh}) }, |
|
|
881 | on_error => sub { $connect_cb->() }, |
|
|
882 | ; |
|
|
883 | |
|
|
884 | $hdl->push_write (pack "CCnNZ*Z*", 4, 1, $port, 1, $socks_user, $host); |
|
|
885 | |
|
|
886 | $hdl->push_read (chunk => 8, sub { |
|
|
887 | my ($hdl, $chunk) = @_; |
|
|
888 | my ($status, $port, $ipn) = unpack "xCna4", $chunk; |
|
|
889 | |
|
|
890 | if ($status == 0x5a) { |
|
|
891 | $connect_cb->($hdl->{fh}, (format_address $ipn) . ":$port"); |
|
|
892 | } else { |
|
|
893 | $! = Errno::ENXIO; $connect_cb->(); |
|
|
894 | } |
|
|
895 | }); |
|
|
896 | |
|
|
897 | $hdl |
|
|
898 | } |
|
|
899 | |
|
|
900 | Use C<socks4a_connect> instead of C<tcp_connect> when doing C<http_request>s, |
|
|
901 | possibly after switching off other proxy types: |
|
|
902 | |
|
|
903 | AnyEvent::HTTP::set_proxy undef; # usually you do not want other proxies |
|
|
904 | |
|
|
905 | http_get 'http://www.google.com', tcp_connect => \&socks4a_connect, sub { |
|
|
906 | my ($data, $headers) = @_; |
|
|
907 | ... |
|
|
908 | }; |
765 | |
909 | |
766 | =head1 SEE ALSO |
910 | =head1 SEE ALSO |
767 | |
911 | |
768 | L<AnyEvent>. |
912 | L<AnyEvent>. |
769 | |
913 | |