… | |
… | |
39 | package AnyEvent::HTTP; |
39 | package AnyEvent::HTTP; |
40 | |
40 | |
41 | use strict; |
41 | use strict; |
42 | no warnings; |
42 | no warnings; |
43 | |
43 | |
44 | use Carp; |
44 | use Errno (); |
45 | |
45 | |
46 | use AnyEvent (); |
46 | use AnyEvent 5.0 (); |
47 | use AnyEvent::Util (); |
47 | use AnyEvent::Util (); |
48 | use AnyEvent::Socket (); |
48 | use AnyEvent::Socket (); |
49 | use AnyEvent::Handle (); |
49 | use AnyEvent::Handle (); |
50 | |
50 | |
51 | use base Exporter::; |
51 | use base Exporter::; |
52 | |
52 | |
53 | our $VERSION = '1.12'; |
53 | our $VERSION = '1.43'; |
54 | |
54 | |
55 | our @EXPORT = qw(http_get http_post http_head http_request); |
55 | our @EXPORT = qw(http_get http_post http_head http_request); |
56 | |
56 | |
57 | our $USERAGENT = "Mozilla/5.0 (compatible; AnyEvent::HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)"; |
57 | our $USERAGENT = "Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)"; |
58 | our $MAX_RECURSE = 10; |
58 | our $MAX_RECURSE = 10; |
59 | our $MAX_PERSISTENT = 8; |
59 | our $MAX_PERSISTENT = 8; |
60 | our $PERSISTENT_TIMEOUT = 2; |
60 | our $PERSISTENT_TIMEOUT = 2; |
61 | our $TIMEOUT = 300; |
61 | our $TIMEOUT = 300; |
62 | |
62 | |
63 | # changing these is evil |
63 | # changing these is evil |
64 | our $MAX_PERSISTENT_PER_HOST = 2; |
64 | our $MAX_PERSISTENT_PER_HOST = 0; |
65 | our $MAX_PER_HOST = 4; |
65 | our $MAX_PER_HOST = 4; |
66 | |
66 | |
67 | our $PROXY; |
67 | our $PROXY; |
68 | our $ACTIVE = 0; |
68 | our $ACTIVE = 0; |
69 | |
69 | |
… | |
… | |
94 | When called in void context, nothing is returned. In other contexts, |
94 | When called in void context, nothing is returned. In other contexts, |
95 | C<http_request> returns a "cancellation guard" - you have to keep the |
95 | 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 |
96 | 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. |
97 | destroyed before the callbakc is called, the request will be cancelled. |
98 | |
98 | |
99 | The callback will be called with the response data as first argument |
99 | The callback will be called with the response body data as first argument |
100 | (or C<undef> if it wasn't available due to errors), and a hash-ref with |
100 | (or C<undef> if an error occured), and a hash-ref with response headers as |
101 | response headers as second argument. |
101 | second argument. |
102 | |
102 | |
103 | All the headers in that hash are lowercased. In addition to the response |
103 | 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> |
104 | headers, the "pseudo-headers" C<HTTPVersion>, C<Status> and C<Reason> |
105 | contain the three parts of the HTTP Status-Line of the same name. The |
105 | contain the three parts of the HTTP Status-Line of the same name. The |
106 | pseudo-header C<URL> contains the original URL (which can differ from the |
106 | pseudo-header C<URL> contains the original URL (which can differ from the |
… | |
… | |
108 | |
108 | |
109 | If the server sends a header multiple times, then their contents will be |
109 | 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. |
110 | joined together with a comma (C<,>), as per the HTTP spec. |
111 | |
111 | |
112 | If an internal error occurs, such as not being able to resolve a hostname, |
112 | If an internal error occurs, such as not being able to resolve a hostname, |
113 | then C<$data> will be C<undef>, C<< $headers->{Status} >> will be C<599> |
113 | then C<$data> will be C<undef>, C<< $headers->{Status} >> will be C<59x> |
114 | and the C<Reason> pseudo-header will contain an error message. |
114 | (usually C<599>) and the C<Reason> pseudo-header will contain an error |
|
|
115 | message. |
115 | |
116 | |
116 | A typical callback might look like this: |
117 | A typical callback might look like this: |
117 | |
118 | |
118 | sub { |
119 | sub { |
119 | my ($body, $hdr) = @_; |
120 | my ($body, $hdr) = @_; |
… | |
… | |
137 | |
138 | |
138 | =item headers => hashref |
139 | =item headers => hashref |
139 | |
140 | |
140 | The request headers to use. Currently, C<http_request> may provide its |
141 | The request headers to use. Currently, C<http_request> may provide its |
141 | own C<Host:>, C<Content-Length:>, C<Connection:> and C<Cookie:> headers |
142 | own C<Host:>, C<Content-Length:>, C<Connection:> and C<Cookie:> headers |
142 | and will provide defaults for C<User-Agent:> and C<Referer:>. |
143 | and will provide defaults for C<User-Agent:> and C<Referer:> (this can be |
|
|
144 | suppressed by using C<undef> for these headers in which case they won't be |
|
|
145 | sent at all). |
143 | |
146 | |
144 | =item timeout => $seconds |
147 | =item timeout => $seconds |
145 | |
148 | |
146 | The time-out to use for various stages - each connect attempt will reset |
149 | The time-out to use for various stages - each connect attempt will reset |
147 | the timeout, as will read or write activity. Default timeout is 5 minutes. |
150 | the timeout, as will read or write activity, i.e. this is not an overall |
|
|
151 | timeout. |
|
|
152 | |
|
|
153 | Default timeout is 5 minutes. |
148 | |
154 | |
149 | =item proxy => [$host, $port[, $scheme]] or undef |
155 | =item proxy => [$host, $port[, $scheme]] or undef |
150 | |
156 | |
151 | Use the given http proxy for all requests. If not specified, then the |
157 | Use the given http proxy for all requests. If not specified, then the |
152 | default proxy (as specified by C<$ENV{http_proxy}>) is used. |
158 | default proxy (as specified by C<$ENV{http_proxy}>) is used. |
153 | |
159 | |
154 | C<$scheme> must be either missing or C<http> for HTTP, or C<https> for |
160 | C<$scheme> must be either missing, C<http> for HTTP or C<https> for |
155 | HTTPS. |
161 | HTTPS. |
156 | |
162 | |
157 | =item body => $string |
163 | =item body => $string |
158 | |
164 | |
159 | The request body, usually empty. Will be-sent as-is (future versions of |
165 | The request body, usually empty. Will be-sent as-is (future versions of |
… | |
… | |
165 | based on the original netscape specification. |
171 | based on the original netscape specification. |
166 | |
172 | |
167 | The C<$hash_ref> must be an (initially empty) hash reference which will |
173 | The C<$hash_ref> must be an (initially empty) hash reference which will |
168 | get updated automatically. It is possible to save the cookie_jar to |
174 | get updated automatically. It is possible to save the cookie_jar to |
169 | persistent storage with something like JSON or Storable, but this is not |
175 | persistent storage with something like JSON or Storable, but this is not |
170 | recommended, as expire times are currently being ignored. |
176 | recommended, as expiry times are currently being ignored. |
171 | |
177 | |
172 | Note that this cookie implementation is not of very high quality, nor |
178 | Note that this cookie implementation is not of very high quality, nor |
173 | meant to be complete. If you want complete cookie management you have to |
179 | meant to be complete. If you want complete cookie management you have to |
174 | do that on your own. C<cookie_jar> is meant as a quick fix to get some |
180 | do that on your own. C<cookie_jar> is meant as a quick fix to get some |
175 | cookie-using sites working. Cookies are a privacy disaster, do not use |
181 | cookie-using sites working. Cookies are a privacy disaster, do not use |
176 | them unless required to. |
182 | them unless required to. |
|
|
183 | |
|
|
184 | =item tls_ctx => $scheme | $tls_ctx |
|
|
185 | |
|
|
186 | Specifies the AnyEvent::TLS context to be used for https connections. This |
|
|
187 | parameter follows the same rules as the C<tls_ctx> parameter to |
|
|
188 | L<AnyEvent::Handle>, but additionally, the two strings C<low> or |
|
|
189 | C<high> can be specified, which give you a predefined low-security (no |
|
|
190 | verification, highest compatibility) and high-security (CA and common-name |
|
|
191 | verification) TLS context. |
|
|
192 | |
|
|
193 | The default for this option is C<low>, which could be interpreted as "give |
|
|
194 | me the page, no matter what". |
|
|
195 | |
|
|
196 | =item on_prepare => $callback->($fh) |
|
|
197 | |
|
|
198 | In rare cases you need to "tune" the socket before it is used to |
|
|
199 | connect (for exmaple, to bind it on a given IP address). This parameter |
|
|
200 | overrides the prepare callback passed to C<AnyEvent::Socket::tcp_connect> |
|
|
201 | and behaves exactly the same way (e.g. it has to provide a |
|
|
202 | timeout). See the description for the C<$prepare_cb> argument of |
|
|
203 | C<AnyEvent::Socket::tcp_connect> for details. |
|
|
204 | |
|
|
205 | =item on_header => $callback->($headers) |
|
|
206 | |
|
|
207 | When specified, this callback will be called with the header hash as soon |
|
|
208 | as headers have been successfully received from the remote server (not on |
|
|
209 | locally-generated errors). |
|
|
210 | |
|
|
211 | It has to return either true (in which case AnyEvent::HTTP will continue), |
|
|
212 | or false, in which case AnyEvent::HTTP will cancel the download (and call |
|
|
213 | the finish callback with an error code of C<598>). |
|
|
214 | |
|
|
215 | This callback is useful, among other things, to quickly reject unwanted |
|
|
216 | content, which, if it is supposed to be rare, can be faster than first |
|
|
217 | doing a C<HEAD> request. |
|
|
218 | |
|
|
219 | Example: cancel the request unless the content-type is "text/html". |
|
|
220 | |
|
|
221 | on_header => sub { |
|
|
222 | $_[0]{"content-type"} =~ /^text\/html\s*(?:;|$)/ |
|
|
223 | }, |
|
|
224 | |
|
|
225 | =item on_body => $callback->($partial_body, $headers) |
|
|
226 | |
|
|
227 | When specified, all body data will be passed to this callback instead of |
|
|
228 | to the completion callback. The completion callback will get the empty |
|
|
229 | string instead of the body data. |
|
|
230 | |
|
|
231 | It has to return either true (in which case AnyEvent::HTTP will continue), |
|
|
232 | or false, in which case AnyEvent::HTTP will cancel the download (and call |
|
|
233 | the completion callback with an error code of C<598>). |
|
|
234 | |
|
|
235 | This callback is useful when the data is too large to be held in memory |
|
|
236 | (so the callback writes it to a file) or when only some information should |
|
|
237 | be extracted, or when the body should be processed incrementally. |
|
|
238 | |
|
|
239 | It is usually preferred over doing your own body handling via |
|
|
240 | C<want_body_handle>, but in case of streaming APIs, where HTTP is |
|
|
241 | only used to create a connection, C<want_body_handle> is the better |
|
|
242 | alternative, as it allows you to install your own event handler, reducing |
|
|
243 | resource usage. |
|
|
244 | |
|
|
245 | =item want_body_handle => $enable |
|
|
246 | |
|
|
247 | When enabled (default is disabled), the behaviour of AnyEvent::HTTP |
|
|
248 | changes considerably: after parsing the headers, and instead of |
|
|
249 | downloading the body (if any), the completion callback will be |
|
|
250 | called. Instead of the C<$body> argument containing the body data, the |
|
|
251 | callback will receive the L<AnyEvent::Handle> object associated with the |
|
|
252 | connection. In error cases, C<undef> will be passed. When there is no body |
|
|
253 | (e.g. status C<304>), the empty string will be passed. |
|
|
254 | |
|
|
255 | The handle object might or might not be in TLS mode, might be connected to |
|
|
256 | a proxy, be a persistent connection etc., and configured in unspecified |
|
|
257 | ways. The user is responsible for this handle (it will not be used by this |
|
|
258 | module anymore). |
|
|
259 | |
|
|
260 | This is useful with some push-type services, where, after the initial |
|
|
261 | headers, an interactive protocol is used (typical example would be the |
|
|
262 | push-style twitter API which starts a JSON/XML stream). |
|
|
263 | |
|
|
264 | If you think you need this, first have a look at C<on_body>, to see if |
|
|
265 | that doesn't solve your problem in a better way. |
177 | |
266 | |
178 | =back |
267 | =back |
179 | |
268 | |
180 | Example: make a simple HTTP GET request for http://www.nethype.de/ |
269 | Example: make a simple HTTP GET request for http://www.nethype.de/ |
181 | |
270 | |
… | |
… | |
237 | push @{ $CO_SLOT{$_[0]}[1] }, $_[1]; |
326 | push @{ $CO_SLOT{$_[0]}[1] }, $_[1]; |
238 | |
327 | |
239 | _slot_schedule $_[0]; |
328 | _slot_schedule $_[0]; |
240 | } |
329 | } |
241 | |
330 | |
242 | our $qr_nl = qr<\015?\012>; |
331 | our $qr_nl = qr{\015?\012}; |
243 | our $qr_nlnl = qr<\015?\012\015?\012>; |
332 | our $qr_nlnl = qr{(?<![^\012])\015?\012}; |
|
|
333 | |
|
|
334 | our $TLS_CTX_LOW = { cache => 1, sslv2 => 1 }; |
|
|
335 | our $TLS_CTX_HIGH = { cache => 1, verify => 1, verify_peername => "https" }; |
244 | |
336 | |
245 | sub http_request($$@) { |
337 | sub http_request($$@) { |
246 | my $cb = pop; |
338 | my $cb = pop; |
247 | my ($method, $url, %arg) = @_; |
339 | my ($method, $url, %arg) = @_; |
248 | |
340 | |
249 | my %hdr; |
341 | my %hdr; |
|
|
342 | |
|
|
343 | $arg{tls_ctx} = $TLS_CTX_LOW if $arg{tls_ctx} eq "low" || !exists $arg{tls_ctx}; |
|
|
344 | $arg{tls_ctx} = $TLS_CTX_HIGH if $arg{tls_ctx} eq "high"; |
250 | |
345 | |
251 | $method = uc $method; |
346 | $method = uc $method; |
252 | |
347 | |
253 | if (my $hdr = $arg{headers}) { |
348 | if (my $hdr = $arg{headers}) { |
254 | while (my ($k, $v) = each %$hdr) { |
349 | while (my ($k, $v) = each %$hdr) { |
… | |
… | |
256 | } |
351 | } |
257 | } |
352 | } |
258 | |
353 | |
259 | my $recurse = exists $arg{recurse} ? delete $arg{recurse} : $MAX_RECURSE; |
354 | my $recurse = exists $arg{recurse} ? delete $arg{recurse} : $MAX_RECURSE; |
260 | |
355 | |
261 | return $cb->(undef, { Status => 599, Reason => "recursion limit reached", URL => $url }) |
356 | return $cb->(undef, { Status => 599, Reason => "Too many redirections", URL => $url }) |
262 | if $recurse < 0; |
357 | if $recurse < 0; |
263 | |
358 | |
264 | my $proxy = $arg{proxy} || $PROXY; |
359 | my $proxy = $arg{proxy} || $PROXY; |
265 | my $timeout = $arg{timeout} || $TIMEOUT; |
360 | my $timeout = $arg{timeout} || $TIMEOUT; |
266 | |
361 | |
267 | $hdr{"user-agent"} ||= $USERAGENT; |
|
|
268 | |
|
|
269 | my ($uscheme, $uauthority, $upath, $query, $fragment) = |
362 | my ($uscheme, $uauthority, $upath, $query, $fragment) = |
270 | $url =~ m|(?:([^:/?#]+):)?(?://([^/?#]*))?([^?#]*)(?:\?([^#]*))?(?:#(.*))?|; |
363 | $url =~ m|(?:([^:/?#]+):)?(?://([^/?#]*))?([^?#]*)(?:\?([^#]*))?(?:#(.*))?|; |
271 | |
364 | |
272 | $uscheme = lc $uscheme; |
365 | $uscheme = lc $uscheme; |
273 | |
366 | |
274 | my $uport = $uscheme eq "http" ? 80 |
367 | my $uport = $uscheme eq "http" ? 80 |
275 | : $uscheme eq "https" ? 443 |
368 | : $uscheme eq "https" ? 443 |
276 | : return $cb->(undef, { Status => 599, Reason => "only http and https URL schemes supported", URL => $url }); |
369 | : return $cb->(undef, { Status => 599, Reason => "Only http and https URL schemes supported", URL => $url }); |
277 | |
|
|
278 | $hdr{referer} ||= "$uscheme://$uauthority$upath"; # leave out fragment and query string, just a heuristic |
|
|
279 | |
370 | |
280 | $uauthority =~ /^(?: .*\@ )? ([^\@:]+) (?: : (\d+) )?$/x |
371 | $uauthority =~ /^(?: .*\@ )? ([^\@:]+) (?: : (\d+) )?$/x |
281 | or return $cb->(undef, { Status => 599, Reason => "unparsable URL", URL => $url }); |
372 | or return $cb->(undef, { Status => 599, Reason => "Unparsable URL", URL => $url }); |
282 | |
373 | |
283 | my $uhost = $1; |
374 | my $uhost = $1; |
284 | $uport = $2 if defined $2; |
375 | $uport = $2 if defined $2; |
|
|
376 | |
|
|
377 | $hdr{host} = defined $2 ? "$uhost:$2" : "$uhost"; |
285 | |
378 | |
286 | $uhost =~ s/^\[(.*)\]$/$1/; |
379 | $uhost =~ s/^\[(.*)\]$/$1/; |
287 | $upath .= "?$query" if length $query; |
380 | $upath .= "?$query" if length $query; |
288 | |
381 | |
289 | $upath =~ s%^/?%/%; |
382 | $upath =~ s%^/?%/%; |
… | |
… | |
322 | my ($rhost, $rport, $rscheme, $rpath); # request host, port, path |
415 | my ($rhost, $rport, $rscheme, $rpath); # request host, port, path |
323 | |
416 | |
324 | if ($proxy) { |
417 | if ($proxy) { |
325 | ($rpath, $rhost, $rport, $rscheme) = ($url, @$proxy); |
418 | ($rpath, $rhost, $rport, $rscheme) = ($url, @$proxy); |
326 | |
419 | |
|
|
420 | $rscheme = "http" unless defined $rscheme; |
|
|
421 | |
327 | # don't support https requests over https-proxy transport, |
422 | # don't support https requests over https-proxy transport, |
328 | # can't be done with tls as spec'ed, unless you double-encrypt. |
423 | # can't be done with tls as spec'ed, unless you double-encrypt. |
329 | $rscheme = "http" if $uscheme eq "https" && $rscheme eq "https"; |
424 | $rscheme = "http" if $uscheme eq "https" && $rscheme eq "https"; |
330 | } else { |
425 | } else { |
331 | ($rhost, $rport, $rscheme, $rpath) = ($uhost, $uport, $uscheme, $upath); |
426 | ($rhost, $rport, $rscheme, $rpath) = ($uhost, $uport, $uscheme, $upath); |
332 | } |
427 | } |
333 | |
428 | |
334 | $hdr{host} = $uhost; |
429 | # leave out fragment and query string, just a heuristic |
|
|
430 | $hdr{referer} ||= "$uscheme://$uauthority$upath" unless exists $hdr{referer}; |
|
|
431 | $hdr{"user-agent"} ||= $USERAGENT unless exists $hdr{"user-agent"}; |
|
|
432 | |
335 | $hdr{"content-length"} = length $arg{body}; |
433 | $hdr{"content-length"} = length $arg{body}; |
336 | |
434 | |
337 | my %state = (connect_guard => 1); |
435 | my %state = (connect_guard => 1); |
338 | |
436 | |
339 | _get_slot $uhost, sub { |
437 | _get_slot $uhost, sub { |
… | |
… | |
341 | |
439 | |
342 | return unless $state{connect_guard}; |
440 | return unless $state{connect_guard}; |
343 | |
441 | |
344 | $state{connect_guard} = AnyEvent::Socket::tcp_connect $rhost, $rport, sub { |
442 | $state{connect_guard} = AnyEvent::Socket::tcp_connect $rhost, $rport, sub { |
345 | $state{fh} = shift |
443 | $state{fh} = shift |
|
|
444 | or do { |
|
|
445 | my $err = "$!"; |
|
|
446 | %state = (); |
346 | or return $cb->(undef, { Status => 599, Reason => "$!", URL => $url }); |
447 | return $cb->(undef, { Status => 599, Reason => $err, URL => $url }); |
|
|
448 | }; |
|
|
449 | |
347 | pop; # free memory, save a tree |
450 | pop; # free memory, save a tree |
348 | |
451 | |
349 | return unless delete $state{connect_guard}; |
452 | return unless delete $state{connect_guard}; |
350 | |
453 | |
351 | # get handle |
454 | # get handle |
352 | $state{handle} = new AnyEvent::Handle |
455 | $state{handle} = new AnyEvent::Handle |
353 | fh => $state{fh}, |
456 | fh => $state{fh}, |
354 | timeout => $timeout; |
457 | timeout => $timeout, |
|
|
458 | peername => $rhost, |
|
|
459 | tls_ctx => $arg{tls_ctx}; |
355 | |
460 | |
356 | # limit the number of persistent connections |
461 | # limit the number of persistent connections |
357 | # keepalive not yet supported |
462 | # keepalive not yet supported |
358 | if ($KA_COUNT{$_[1]} < $MAX_PERSISTENT_PER_HOST) { |
463 | if ($KA_COUNT{$_[1]} < $MAX_PERSISTENT_PER_HOST) { |
359 | ++$KA_COUNT{$_[1]}; |
464 | ++$KA_COUNT{$_[1]}; |
… | |
… | |
365 | delete $hdr{connection}; |
470 | delete $hdr{connection}; |
366 | } |
471 | } |
367 | |
472 | |
368 | # (re-)configure handle |
473 | # (re-)configure handle |
369 | $state{handle}->on_error (sub { |
474 | $state{handle}->on_error (sub { |
370 | my $errno = "$!"; |
|
|
371 | %state = (); |
475 | %state = (); |
372 | $cb->(undef, { Status => 599, Reason => $errno, URL => $url }); |
476 | $cb->(undef, { Status => 599, Reason => $_[2], URL => $url }); |
373 | }); |
477 | }); |
374 | $state{handle}->on_eof (sub { |
478 | $state{handle}->on_eof (sub { |
375 | %state = (); |
479 | %state = (); |
376 | $cb->(undef, { Status => 599, Reason => "unexpected end-of-file", URL => $url }); |
480 | $cb->(undef, { Status => 599, Reason => "Unexpected end-of-file", URL => $url }); |
377 | }); |
481 | }); |
378 | |
482 | |
379 | $state{handle}->starttls ("connect") if $rscheme eq "https"; |
483 | $state{handle}->starttls ("connect") if $rscheme eq "https"; |
380 | |
484 | |
381 | # handle actual, non-tunneled, request |
485 | # handle actual, non-tunneled, request |
… | |
… | |
383 | $state{handle}->starttls ("connect") if $uscheme eq "https" && !exists $state{handle}{tls}; |
487 | $state{handle}->starttls ("connect") if $uscheme eq "https" && !exists $state{handle}{tls}; |
384 | |
488 | |
385 | # send request |
489 | # send request |
386 | $state{handle}->push_write ( |
490 | $state{handle}->push_write ( |
387 | "$method $rpath HTTP/1.0\015\012" |
491 | "$method $rpath HTTP/1.0\015\012" |
388 | . (join "", map "\u$_: $hdr{$_}\015\012", keys %hdr) |
492 | . (join "", map "\u$_: $hdr{$_}\015\012", grep defined $hdr{$_}, keys %hdr) |
389 | . "\015\012" |
493 | . "\015\012" |
390 | . (delete $arg{body}) |
494 | . (delete $arg{body}) |
391 | ); |
495 | ); |
392 | |
496 | |
393 | %hdr = (); # reduce memory usage, save a kitten |
497 | %hdr = (); # reduce memory usage, save a kitten |
394 | |
498 | |
395 | # status line |
499 | # status line |
396 | $state{handle}->push_read (line => $qr_nl, sub { |
500 | $state{handle}->push_read (line => $qr_nl, sub { |
397 | $_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix |
501 | $_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix |
398 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "invalid server response ($_[1])", URL => $url })); |
502 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "Invalid server response ($_[1])", URL => $url })); |
399 | |
503 | |
400 | my %hdr = ( # response headers |
504 | my %hdr = ( # response headers |
401 | HTTPVersion => ",$1", |
505 | HTTPVersion => ",$1", |
402 | Status => ",$2", |
506 | Status => ",$2", |
403 | Reason => ",$3", |
507 | Reason => ",$3", |
404 | URL => ",$url" |
508 | URL => ",$url" |
405 | ); |
509 | ); |
406 | |
510 | |
407 | # headers, could be optimized a bit |
511 | # headers, could be optimized a bit |
408 | $state{handle}->unshift_read (line => $qr_nlnl, sub { |
512 | $state{handle}->unshift_read (line => $qr_nlnl, sub { |
409 | for ("$_[1]\012") { |
513 | for ("$_[1]") { |
410 | y/\015//d; # weed out any \015, as they show up in the weirdest of places. |
514 | y/\015//d; # weed out any \015, as they show up in the weirdest of places. |
411 | |
515 | |
412 | # we support spaces in field names, as lotus domino |
516 | # things seen, not parsed: |
413 | # creates them (actually spaces around seperators |
517 | # p3pP="NON CUR OTPi OUR NOR UNI" |
414 | # are strictly allowed in http, they are a security issue). |
518 | |
415 | $hdr{lc $1} .= ",$2" |
519 | $hdr{lc $1} .= ",$2" |
416 | while /\G |
520 | while /\G |
417 | ([^:\000-\037]+): |
521 | ([^:\000-\037]*): |
418 | [\011\040]* |
522 | [\011\040]* |
419 | ((?: [^\012]+ | \012[\011\040] )*) |
523 | ((?: [^\012]+ | \012[\011\040] )*) |
420 | \012 |
524 | \012 |
421 | /gxc; |
525 | /gxc; |
422 | |
526 | |
423 | /\G$/ |
527 | /\G$/ |
424 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "garbled response headers", URL => $url })); |
528 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "Garbled response headers", URL => $url })); |
425 | } |
529 | } |
426 | |
530 | |
427 | substr $_, 0, 1, "" |
531 | substr $_, 0, 1, "" |
428 | for values %hdr; |
532 | for values %hdr; |
429 | |
533 | |
|
|
534 | # redirect handling |
|
|
535 | # microsoft and other shitheads don't give a shit for following standards, |
|
|
536 | # try to support some common forms of broken Location headers. |
|
|
537 | if ($hdr{location} !~ /^(?: $ | [^:\/?\#]+ : )/x) { |
|
|
538 | $hdr{location} =~ s/^\.\/+//; |
|
|
539 | |
|
|
540 | my $url = "$rscheme://$uhost:$uport"; |
|
|
541 | |
|
|
542 | unless ($hdr{location} =~ s/^\///) { |
|
|
543 | $url .= $upath; |
|
|
544 | $url =~ s/\/[^\/]*$//; |
|
|
545 | } |
|
|
546 | |
|
|
547 | $hdr{location} = "$url/$hdr{location}"; |
|
|
548 | } |
|
|
549 | |
|
|
550 | my $redirect; |
|
|
551 | |
|
|
552 | if ($recurse) { |
|
|
553 | if ($hdr{Status} =~ /^30[12]$/ && $method ne "POST") { |
|
|
554 | # apparently, mozilla et al. just change POST to GET here |
|
|
555 | # more research is needed before we do the same |
|
|
556 | $redirect = 1; |
|
|
557 | } elsif ($hdr{Status} == 303) { |
|
|
558 | # even http/1.1 is unclear on how to mutate the method |
|
|
559 | $method = "GET" unless $method eq "HEAD"; |
|
|
560 | $redirect = 1; |
|
|
561 | } elsif ($hdr{Status} == 307 && $method =~ /^(?:GET|HEAD)$/) { |
|
|
562 | $redirect = 1; |
|
|
563 | } |
|
|
564 | } |
|
|
565 | |
430 | my $finish = sub { |
566 | my $finish = sub { |
431 | $state{handle}->destroy; |
567 | $state{handle}->destroy if $state{handle}; |
432 | %state = (); |
568 | %state = (); |
433 | |
569 | |
434 | # set-cookie processing |
570 | # set-cookie processing |
435 | if ($arg{cookie_jar}) { |
571 | if ($arg{cookie_jar}) { |
436 | for ($hdr{"set-cookie"}) { |
572 | for ($_[1]{"set-cookie"}) { |
437 | # parse NAME=VALUE |
573 | # parse NAME=VALUE |
438 | my @kv; |
574 | my @kv; |
439 | |
575 | |
440 | while (/\G\s* ([^=;,[:space:]]+) \s*=\s* (?: "((?:[^\\"]+|\\.)*)" | ([^=;,[:space:]]*) )/gcxs) { |
576 | while (/\G\s* ([^=;,[:space:]]+) \s*=\s* (?: "((?:[^\\"]+|\\.)*)" | ([^=;,[:space:]]*) )/gcxs) { |
441 | my $name = $1; |
577 | my $name = $1; |
… | |
… | |
479 | |
615 | |
480 | redo if /\G\s*,/gc; |
616 | redo if /\G\s*,/gc; |
481 | } |
617 | } |
482 | } |
618 | } |
483 | |
619 | |
484 | # microsoft and other shitheads don't give a shit for following standards, |
620 | if ($redirect && exists $hdr{location}) { |
485 | # try to support some common forms of broken Location headers. |
621 | # we ignore any errors, as it is very common to receive |
486 | if ($_[1]{location} !~ /^(?: $ | [^:\/?\#]+ : )/x) { |
622 | # Content-Length != 0 but no actual body |
487 | $_[1]{location} =~ s/^\.\/+//; |
623 | # we also access %hdr, as $_[1] might be an erro |
488 | |
|
|
489 | my $url = "$rscheme://$uhost:$uport"; |
|
|
490 | |
|
|
491 | unless ($_[1]{location} =~ s/^\///) { |
|
|
492 | $url .= $upath; |
|
|
493 | $url =~ s/\/[^\/]*$//; |
|
|
494 | } |
|
|
495 | |
|
|
496 | $_[1]{location} = "$url/$_[1]{location}"; |
|
|
497 | } |
|
|
498 | |
|
|
499 | if ($_[1]{Status} =~ /^30[12]$/ && $recurse && $method ne "POST") { |
|
|
500 | # apparently, mozilla et al. just change POST to GET here |
|
|
501 | # more research is needed before we do the same |
|
|
502 | http_request ($method, $_[1]{location}, %arg, recurse => $recurse - 1, $cb); |
|
|
503 | } elsif ($_[1]{Status} == 303 && $recurse) { |
|
|
504 | # even http/1.1 is unclear on how to mutate the method |
|
|
505 | $method = "GET" unless $method eq "HEAD"; |
|
|
506 | http_request ($method => $_[1]{location}, %arg, recurse => $recurse - 1, $cb); |
624 | http_request ($method => $hdr{location}, %arg, recurse => $recurse - 1, $cb); |
507 | } elsif ($_[1]{Status} == 307 && $recurse && $method =~ /^(?:GET|HEAD)$/) { |
|
|
508 | http_request ($method => $_[1]{location}, %arg, recurse => $recurse - 1, $cb); |
|
|
509 | } else { |
625 | } else { |
510 | $cb->($_[0], $_[1]); |
626 | $cb->($_[0], $_[1]); |
511 | } |
627 | } |
512 | }; |
628 | }; |
513 | |
629 | |
514 | if ($hdr{Status} =~ /^(?:1..|204|304)$/ or $method eq "HEAD") { |
630 | my $len = $hdr{"content-length"}; |
|
|
631 | |
|
|
632 | if (!$redirect && $arg{on_header} && !$arg{on_header}(\%hdr)) { |
|
|
633 | $finish->(undef, { Status => 598, Reason => "Request cancelled by on_header", URL => $url }); |
|
|
634 | } elsif ( |
|
|
635 | $hdr{Status} =~ /^(?:1..|[23]04)$/ |
|
|
636 | or $method eq "HEAD" |
|
|
637 | or (defined $len && !$len) |
|
|
638 | ) { |
|
|
639 | # no body |
515 | $finish->(undef, \%hdr); |
640 | $finish->("", \%hdr); |
516 | } else { |
641 | } else { |
517 | if (exists $hdr{"content-length"}) { |
642 | # body handling, four different code paths |
518 | $_[0]->unshift_read (chunk => $hdr{"content-length"}, sub { |
643 | # for want_body_handle, on_body (2x), normal (2x) |
519 | # could cache persistent connection now |
644 | # we might read too much here, but it does not matter yet (no pers. connections) |
520 | if ($hdr{connection} =~ /\bkeep-alive\b/i) { |
645 | if (!$redirect && $arg{want_body_handle}) { |
521 | # but we don't, due to misdesigns, this is annoyingly complex |
646 | $_[0]->on_eof (undef); |
|
|
647 | $_[0]->on_error (undef); |
|
|
648 | $_[0]->on_read (undef); |
|
|
649 | |
|
|
650 | $finish->(delete $state{handle}, \%hdr); |
|
|
651 | |
|
|
652 | } elsif ($arg{on_body}) { |
|
|
653 | $_[0]->on_error (sub { $finish->(undef, { Status => 599, Reason => $_[2], URL => $url }) }); |
|
|
654 | if ($len) { |
|
|
655 | $_[0]->on_eof (undef); |
|
|
656 | $_[0]->on_read (sub { |
|
|
657 | $len -= length $_[0]{rbuf}; |
|
|
658 | |
|
|
659 | $arg{on_body}(delete $_[0]{rbuf}, \%hdr) |
|
|
660 | or $finish->(undef, { Status => 598, Reason => "Request cancelled by on_body", URL => $url }); |
|
|
661 | |
|
|
662 | $len > 0 |
|
|
663 | or $finish->("", \%hdr); |
522 | }; |
664 | }); |
523 | |
665 | } else { |
|
|
666 | $_[0]->on_eof (sub { |
524 | $finish->($_[1], \%hdr); |
667 | $finish->("", \%hdr); |
|
|
668 | }); |
|
|
669 | $_[0]->on_read (sub { |
|
|
670 | $arg{on_body}(delete $_[0]{rbuf}, \%hdr) |
|
|
671 | or $finish->(undef, { Status => 598, Reason => "Request cancelled by on_body", URL => $url }); |
|
|
672 | }); |
525 | }); |
673 | } |
526 | } else { |
674 | } else { |
527 | # too bad, need to read until we get an error or EOF, |
|
|
528 | # no way to detect winged data. |
|
|
529 | $_[0]->on_error (sub { |
|
|
530 | # delete ought to be more efficient, as we would have to make |
|
|
531 | # a copy otherwise as $_[0] gets destroyed. |
|
|
532 | $finish->(delete $_[0]{rbuf}, \%hdr); |
|
|
533 | }); |
|
|
534 | $_[0]->on_eof (undef); |
675 | $_[0]->on_eof (undef); |
|
|
676 | |
|
|
677 | if ($len) { |
|
|
678 | $_[0]->on_error (sub { $finish->(undef, { Status => 599, Reason => $_[2], URL => $url }) }); |
|
|
679 | $_[0]->on_read (sub { |
|
|
680 | $finish->((substr delete $_[0]{rbuf}, 0, $len, ""), \%hdr) |
|
|
681 | if $len <= length $_[0]{rbuf}; |
|
|
682 | }); |
|
|
683 | } else { |
|
|
684 | $_[0]->on_error (sub { |
|
|
685 | $! == Errno::EPIPE || !$! |
|
|
686 | ? $finish->(delete $_[0]{rbuf}, \%hdr) |
|
|
687 | : $finish->(undef, { Status => 599, Reason => $_[2], URL => $url }); |
|
|
688 | }); |
535 | $_[0]->on_read (sub { }); |
689 | $_[0]->on_read (sub { }); |
|
|
690 | } |
536 | } |
691 | } |
537 | } |
692 | } |
538 | }); |
693 | }); |
539 | }); |
694 | }); |
540 | }; |
695 | }; |
… | |
… | |
545 | |
700 | |
546 | # maybe re-use $uauthority with patched port? |
701 | # maybe re-use $uauthority with patched port? |
547 | $state{handle}->push_write ("CONNECT $uhost:$uport HTTP/1.0\015\012Host: $uhost\015\012\015\012"); |
702 | $state{handle}->push_write ("CONNECT $uhost:$uport HTTP/1.0\015\012Host: $uhost\015\012\015\012"); |
548 | $state{handle}->push_read (line => $qr_nlnl, sub { |
703 | $state{handle}->push_read (line => $qr_nlnl, sub { |
549 | $_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix |
704 | $_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix |
550 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "invalid proxy connect response ($_[1])", URL => $url })); |
705 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "Invalid proxy connect response ($_[1])", URL => $url })); |
551 | |
706 | |
552 | if ($2 == 200) { |
707 | if ($2 == 200) { |
553 | $rpath = $upath; |
708 | $rpath = $upath; |
554 | &$handle_actual_request; |
709 | &$handle_actual_request; |
555 | } else { |
710 | } else { |
… | |
… | |
559 | }); |
714 | }); |
560 | } else { |
715 | } else { |
561 | &$handle_actual_request; |
716 | &$handle_actual_request; |
562 | } |
717 | } |
563 | |
718 | |
564 | }, sub { |
719 | }, $arg{on_prepare} || sub { $timeout }; |
565 | $timeout |
|
|
566 | }; |
|
|
567 | }; |
720 | }; |
568 | |
721 | |
569 | defined wantarray && AnyEvent::Util::guard { %state = () } |
722 | defined wantarray && AnyEvent::Util::guard { %state = () } |
570 | } |
723 | } |
571 | |
724 | |
… | |
… | |
601 | The default value for the C<recurse> request parameter (default: C<10>). |
754 | The default value for the C<recurse> request parameter (default: C<10>). |
602 | |
755 | |
603 | =item $AnyEvent::HTTP::USERAGENT |
756 | =item $AnyEvent::HTTP::USERAGENT |
604 | |
757 | |
605 | The default value for the C<User-Agent> header (the default is |
758 | The default value for the C<User-Agent> header (the default is |
606 | C<Mozilla/5.0 (compatible; AnyEvent::HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)>). |
759 | C<Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$VERSION; +http://software.schmorp.de/pkg/AnyEvent)>). |
607 | |
760 | |
608 | =item $AnyEvent::HTTP::MAX_PERSISTENT |
761 | =item $AnyEvent::HTTP::MAX_PER_HOST |
609 | |
762 | |
610 | The maximum number of persistent connections to keep open (default: 8). |
763 | The maximum number of concurrent connections to the same host (identified |
|
|
764 | by the hostname). If the limit is exceeded, then the additional requests |
|
|
765 | are queued until previous connections are closed. |
611 | |
766 | |
612 | Not implemented currently. |
767 | The default value for this is C<4>, and it is highly advisable to not |
613 | |
768 | increase it. |
614 | =item $AnyEvent::HTTP::PERSISTENT_TIMEOUT |
|
|
615 | |
|
|
616 | The maximum time to cache a persistent connection, in seconds (default: 2). |
|
|
617 | |
|
|
618 | Not implemented currently. |
|
|
619 | |
769 | |
620 | =item $AnyEvent::HTTP::ACTIVE |
770 | =item $AnyEvent::HTTP::ACTIVE |
621 | |
771 | |
622 | The number of active connections. This is not the number of currently |
772 | The number of active connections. This is not the number of currently |
623 | running requests, but the number of currently open and non-idle TCP |
773 | running requests, but the number of currently open and non-idle TCP |