… | |
… | |
43 | |
43 | |
44 | use Errno (); |
44 | use Errno (); |
45 | |
45 | |
46 | use AnyEvent 5.0 (); |
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 = '1.44'; |
52 | our $VERSION = '1.5'; |
54 | |
53 | |
55 | our @EXPORT = qw(http_get http_post http_head http_request); |
54 | our @EXPORT = qw(http_get http_post http_head http_request); |
56 | |
55 | |
57 | our $USERAGENT = "Mozilla/5.0 (compatible; U; AnyEvent-HTTP/$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; |
… | |
… | |
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 |
101 | second argument. |
100 | (and trailers) as second argument. |
102 | |
101 | |
103 | All the headers in that hash are lowercased. In addition to the response |
102 | All the headers in that hash are lowercased. In addition to the response |
104 | headers, the "pseudo-headers" C<HTTPVersion>, C<Status> and C<Reason> |
103 | headers, the "pseudo-headers" (uppercase to avoid clashing with possible |
|
|
104 | response headers) C<HTTPVersion>, C<Status> and C<Reason> contain the |
105 | contain the three parts of the HTTP Status-Line of the same name. The |
105 | three parts of the HTTP Status-Line of the same name. If an error occurs |
|
|
106 | during the body phase of a request, then the original C<Status> and |
|
|
107 | C<Reason> values from the header are available as C<OrigStatus> and |
|
|
108 | C<OrigReason>. |
|
|
109 | |
106 | pseudo-header C<URL> contains the original URL (which can differ from the |
110 | The pseudo-header C<URL> contains the actual URL (which can differ from |
107 | requested URL when following redirects). |
111 | the requested URL when following redirects - for example, you might get |
|
|
112 | an error that your URL scheme is not supported even though your URL is a |
|
|
113 | valid http URL because it redirected to an ftp URL, in which case you can |
|
|
114 | look at the URL pseudo header). |
|
|
115 | |
|
|
116 | The pseudo-header C<Redirect> only exists when the request was a result |
|
|
117 | of an internal redirect. In that case it is an array reference with |
|
|
118 | the C<($data, $headers)> from the redirect response. Note that this |
|
|
119 | response could in turn be the result of a redirect itself, and C<< |
|
|
120 | $headers->{Redirect}[1]{Redirect} >> will then contain the original |
|
|
121 | response, and so on. |
108 | |
122 | |
109 | If the server sends a header multiple times, then their contents will be |
123 | If the server sends a header multiple times, then their contents will be |
110 | joined together with a comma (C<,>), as per the HTTP spec. |
124 | joined together with a comma (C<,>), as per the HTTP spec. |
111 | |
125 | |
112 | If an internal error occurs, such as not being able to resolve a hostname, |
126 | If an internal error occurs, such as not being able to resolve a hostname, |
… | |
… | |
136 | Whether to recurse requests or not, e.g. on redirects, authentication |
150 | Whether to recurse requests or not, e.g. on redirects, authentication |
137 | retries and so on, and how often to do so. |
151 | retries and so on, and how often to do so. |
138 | |
152 | |
139 | =item headers => hashref |
153 | =item headers => hashref |
140 | |
154 | |
141 | The request headers to use. Currently, C<http_request> may provide its |
155 | The request headers to use. Currently, C<http_request> may provide its own |
142 | own C<Host:>, C<Content-Length:>, C<Connection:> and C<Cookie:> headers |
156 | C<Host:>, C<Content-Length:>, C<Connection:> and C<Cookie:> headers and |
143 | and will provide defaults for C<User-Agent:> and C<Referer:> (this can be |
157 | will provide defaults at least for C<TE:>, C<Referer:> and C<User-Agent:> |
144 | suppressed by using C<undef> for these headers in which case they won't be |
158 | (this can be suppressed by using C<undef> for these headers in which case |
145 | sent at all). |
159 | they won't be sent at all). |
146 | |
160 | |
147 | =item timeout => $seconds |
161 | =item timeout => $seconds |
148 | |
162 | |
149 | The time-out to use for various stages - each connect attempt will reset |
163 | The time-out to use for various stages - each connect attempt will reset |
150 | the timeout, as will read or write activity, i.e. this is not an overall |
164 | the timeout, as will read or write activity, i.e. this is not an overall |
… | |
… | |
160 | C<$scheme> must be either missing, C<http> for HTTP or C<https> for |
174 | C<$scheme> must be either missing, C<http> for HTTP or C<https> for |
161 | HTTPS. |
175 | HTTPS. |
162 | |
176 | |
163 | =item body => $string |
177 | =item body => $string |
164 | |
178 | |
165 | The request body, usually empty. Will be-sent as-is (future versions of |
179 | The request body, usually empty. Will be sent as-is (future versions of |
166 | this module might offer more options). |
180 | this module might offer more options). |
167 | |
181 | |
168 | =item cookie_jar => $hash_ref |
182 | =item cookie_jar => $hash_ref |
169 | |
183 | |
170 | Passing this parameter enables (simplified) cookie-processing, loosely |
184 | Passing this parameter enables (simplified) cookie-processing, loosely |
… | |
… | |
178 | Note that this cookie implementation is not of very high quality, nor |
192 | Note that this cookie implementation is not of very high quality, nor |
179 | meant to be complete. If you want complete cookie management you have to |
193 | meant to be complete. If you want complete cookie management you have to |
180 | do that on your own. C<cookie_jar> is meant as a quick fix to get some |
194 | do that on your own. C<cookie_jar> is meant as a quick fix to get some |
181 | cookie-using sites working. Cookies are a privacy disaster, do not use |
195 | cookie-using sites working. Cookies are a privacy disaster, do not use |
182 | them unless required to. |
196 | them unless required to. |
|
|
197 | |
|
|
198 | When cookie processing is enabled, the C<Cookie:> and C<Set-Cookie:> |
|
|
199 | headers will be ste and handled by this module, otherwise they will be |
|
|
200 | left untouched. |
183 | |
201 | |
184 | =item tls_ctx => $scheme | $tls_ctx |
202 | =item tls_ctx => $scheme | $tls_ctx |
185 | |
203 | |
186 | Specifies the AnyEvent::TLS context to be used for https connections. This |
204 | Specifies the AnyEvent::TLS context to be used for https connections. This |
187 | parameter follows the same rules as the C<tls_ctx> parameter to |
205 | parameter follows the same rules as the C<tls_ctx> parameter to |
… | |
… | |
200 | overrides the prepare callback passed to C<AnyEvent::Socket::tcp_connect> |
218 | overrides the prepare callback passed to C<AnyEvent::Socket::tcp_connect> |
201 | and behaves exactly the same way (e.g. it has to provide a |
219 | and behaves exactly the same way (e.g. it has to provide a |
202 | timeout). See the description for the C<$prepare_cb> argument of |
220 | timeout). See the description for the C<$prepare_cb> argument of |
203 | C<AnyEvent::Socket::tcp_connect> for details. |
221 | C<AnyEvent::Socket::tcp_connect> for details. |
204 | |
222 | |
|
|
223 | =item tcp_connect => $callback->($host, $service, $connect_cb, $prepare_cb) |
|
|
224 | |
|
|
225 | In even rarer cases you want total control over how AnyEvent::HTTP |
|
|
226 | establishes connections. Normally it uses L<AnyEvent::Socket::tcp_connect> |
|
|
227 | to do this, but you can provide your own C<tcp_connect> function - |
|
|
228 | obviously, it has to follow the same calling conventions, except that it |
|
|
229 | may always return a connection guard object. |
|
|
230 | |
|
|
231 | There are probably lots of weird uses for this function, starting from |
|
|
232 | tracing the hosts C<http_request> actually tries to connect, to (inexact |
|
|
233 | but fast) host => IP address caching or even socks protocol support. |
|
|
234 | |
205 | =item on_header => $callback->($headers) |
235 | =item on_header => $callback->($headers) |
206 | |
236 | |
207 | When specified, this callback will be called with the header hash as soon |
237 | 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 |
238 | as headers have been successfully received from the remote server (not on |
209 | locally-generated errors). |
239 | locally-generated errors). |
… | |
… | |
214 | |
244 | |
215 | This callback is useful, among other things, to quickly reject unwanted |
245 | 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 |
246 | content, which, if it is supposed to be rare, can be faster than first |
217 | doing a C<HEAD> request. |
247 | doing a C<HEAD> request. |
218 | |
248 | |
|
|
249 | The downside is that cancelling the request makes it impossible to re-use |
|
|
250 | the connection. Also, the C<on_header> callback will not receive any |
|
|
251 | trailer (headers sent after the response body). |
|
|
252 | |
219 | Example: cancel the request unless the content-type is "text/html". |
253 | Example: cancel the request unless the content-type is "text/html". |
220 | |
254 | |
221 | on_header => sub { |
255 | on_header => sub { |
222 | $_[0]{"content-type"} =~ /^text\/html\s*(?:;|$)/ |
256 | $_[0]{"content-type"} =~ /^text\/html\s*(?:;|$)/ |
223 | }, |
257 | }, |
… | |
… | |
229 | string instead of the body data. |
263 | string instead of the body data. |
230 | |
264 | |
231 | It has to return either true (in which case AnyEvent::HTTP will continue), |
265 | 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 |
266 | or false, in which case AnyEvent::HTTP will cancel the download (and call |
233 | the completion callback with an error code of C<598>). |
267 | the completion callback with an error code of C<598>). |
|
|
268 | |
|
|
269 | The downside to cancelling the request is that it makes it impossible to |
|
|
270 | re-use the connection. |
234 | |
271 | |
235 | This callback is useful when the data is too large to be held in memory |
272 | 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 |
273 | (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. |
274 | be extracted, or when the body should be processed incrementally. |
238 | |
275 | |
… | |
… | |
264 | If you think you need this, first have a look at C<on_body>, to see if |
301 | 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. |
302 | that doesn't solve your problem in a better way. |
266 | |
303 | |
267 | =back |
304 | =back |
268 | |
305 | |
269 | Example: make a simple HTTP GET request for http://www.nethype.de/ |
306 | Example: do a simple HTTP GET request for http://www.nethype.de/ and print |
|
|
307 | the response body. |
270 | |
308 | |
271 | http_request GET => "http://www.nethype.de/", sub { |
309 | http_request GET => "http://www.nethype.de/", sub { |
272 | my ($body, $hdr) = @_; |
310 | my ($body, $hdr) = @_; |
273 | print "$body\n"; |
311 | print "$body\n"; |
274 | }; |
312 | }; |
275 | |
313 | |
276 | Example: make a HTTP HEAD request on https://www.google.com/, use a |
314 | Example: do a HTTP HEAD request on https://www.google.com/, use a |
277 | timeout of 30 seconds. |
315 | timeout of 30 seconds. |
278 | |
316 | |
279 | http_request |
317 | http_request |
280 | GET => "https://www.google.com", |
318 | GET => "https://www.google.com", |
281 | timeout => 30, |
319 | timeout => 30, |
… | |
… | |
284 | use Data::Dumper; |
322 | use Data::Dumper; |
285 | print Dumper $hdr; |
323 | print Dumper $hdr; |
286 | } |
324 | } |
287 | ; |
325 | ; |
288 | |
326 | |
289 | Example: make another simple HTTP GET request, but immediately try to |
327 | Example: do another simple HTTP GET request, but immediately try to |
290 | cancel it. |
328 | cancel it. |
291 | |
329 | |
292 | my $request = http_request GET => "http://www.nethype.de/", sub { |
330 | my $request = http_request GET => "http://www.nethype.de/", sub { |
293 | my ($body, $hdr) = @_; |
331 | my ($body, $hdr) = @_; |
294 | print "$body\n"; |
332 | print "$body\n"; |
… | |
… | |
326 | push @{ $CO_SLOT{$_[0]}[1] }, $_[1]; |
364 | push @{ $CO_SLOT{$_[0]}[1] }, $_[1]; |
327 | |
365 | |
328 | _slot_schedule $_[0]; |
366 | _slot_schedule $_[0]; |
329 | } |
367 | } |
330 | |
368 | |
331 | our $qr_nl = qr{\015?\012}; |
369 | # continue to parse $_ for headers and place them into the arg |
|
|
370 | sub parse_hdr() { |
|
|
371 | my %hdr; |
|
|
372 | |
|
|
373 | # things seen, not parsed: |
|
|
374 | # p3pP="NON CUR OTPi OUR NOR UNI" |
|
|
375 | |
|
|
376 | $hdr{lc $1} .= ",$2" |
|
|
377 | while /\G |
|
|
378 | ([^:\000-\037]*): |
|
|
379 | [\011\040]* |
|
|
380 | ((?: [^\012]+ | \012[\011\040] )*) |
|
|
381 | \012 |
|
|
382 | /gxc; |
|
|
383 | |
|
|
384 | /\G$/ |
|
|
385 | or return; |
|
|
386 | |
|
|
387 | # remove the "," prefix we added to all headers above |
|
|
388 | substr $_, 0, 1, "" |
|
|
389 | for values %hdr; |
|
|
390 | |
|
|
391 | \%hdr |
|
|
392 | } |
|
|
393 | |
332 | our $qr_nlnl = qr{(?<![^\012])\015?\012}; |
394 | our $qr_nlnl = qr{(?<![^\012])\015?\012}; |
333 | |
395 | |
334 | our $TLS_CTX_LOW = { cache => 1, sslv2 => 1 }; |
396 | our $TLS_CTX_LOW = { cache => 1, sslv2 => 1 }; |
335 | our $TLS_CTX_HIGH = { cache => 1, verify => 1, verify_peername => "https" }; |
397 | our $TLS_CTX_HIGH = { cache => 1, verify => 1, verify_peername => "https" }; |
336 | |
398 | |
… | |
… | |
349 | while (my ($k, $v) = each %$hdr) { |
411 | while (my ($k, $v) = each %$hdr) { |
350 | $hdr{lc $k} = $v; |
412 | $hdr{lc $k} = $v; |
351 | } |
413 | } |
352 | } |
414 | } |
353 | |
415 | |
|
|
416 | # pseudo headers for all subsequent responses |
|
|
417 | my @pseudo = (URL => $url); |
|
|
418 | push @pseudo, Redirect => delete $arg{Redirect} if exists $arg{Redirect}; |
|
|
419 | |
354 | my $recurse = exists $arg{recurse} ? delete $arg{recurse} : $MAX_RECURSE; |
420 | my $recurse = exists $arg{recurse} ? delete $arg{recurse} : $MAX_RECURSE; |
355 | |
421 | |
356 | return $cb->(undef, { Status => 599, Reason => "Too many redirections", URL => $url }) |
422 | return $cb->(undef, { @pseudo, Status => 599, Reason => "Too many redirections" }) |
357 | if $recurse < 0; |
423 | if $recurse < 0; |
358 | |
424 | |
359 | my $proxy = $arg{proxy} || $PROXY; |
425 | my $proxy = $arg{proxy} || $PROXY; |
360 | my $timeout = $arg{timeout} || $TIMEOUT; |
426 | my $timeout = $arg{timeout} || $TIMEOUT; |
361 | |
427 | |
362 | my ($uscheme, $uauthority, $upath, $query, $fragment) = |
428 | my ($uscheme, $uauthority, $upath, $query, $fragment) = |
363 | $url =~ m|(?:([^:/?#]+):)?(?://([^/?#]*))?([^?#]*)(?:\?([^#]*))?(?:#(.*))?|; |
429 | $url =~ m|(?:([^:/?#]+):)?(?://([^/?#]*))?([^?#]*)(?:(\?[^#]*))?(?:#(.*))?|; |
364 | |
430 | |
365 | $uscheme = lc $uscheme; |
431 | $uscheme = lc $uscheme; |
366 | |
432 | |
367 | my $uport = $uscheme eq "http" ? 80 |
433 | my $uport = $uscheme eq "http" ? 80 |
368 | : $uscheme eq "https" ? 443 |
434 | : $uscheme eq "https" ? 443 |
369 | : return $cb->(undef, { Status => 599, Reason => "Only http and https URL schemes supported", URL => $url }); |
435 | : return $cb->(undef, { @pseudo, Status => 599, Reason => "Only http and https URL schemes supported" }); |
370 | |
436 | |
371 | $uauthority =~ /^(?: .*\@ )? ([^\@:]+) (?: : (\d+) )?$/x |
437 | $uauthority =~ /^(?: .*\@ )? ([^\@:]+) (?: : (\d+) )?$/x |
372 | or return $cb->(undef, { Status => 599, Reason => "Unparsable URL", URL => $url }); |
438 | or return $cb->(undef, { @pseudo, Status => 599, Reason => "Unparsable URL" }); |
373 | |
439 | |
374 | my $uhost = $1; |
440 | my $uhost = $1; |
375 | $uport = $2 if defined $2; |
441 | $uport = $2 if defined $2; |
376 | |
442 | |
377 | $hdr{host} = defined $2 ? "$uhost:$2" : "$uhost" |
443 | $hdr{host} = defined $2 ? "$uhost:$2" : "$uhost" |
378 | unless exists $hdr{host}; |
444 | unless exists $hdr{host}; |
379 | |
445 | |
380 | $uhost =~ s/^\[(.*)\]$/$1/; |
446 | $uhost =~ s/^\[(.*)\]$/$1/; |
381 | $upath .= "?$query" if length $query; |
447 | $upath .= $query if length $query; |
382 | |
448 | |
383 | $upath =~ s%^/?%/%; |
449 | $upath =~ s%^/?%/%; |
384 | |
450 | |
385 | # cookie processing |
451 | # cookie processing |
386 | if (my $jar = $arg{cookie_jar}) { |
452 | if (my $jar = $arg{cookie_jar}) { |
… | |
… | |
426 | } else { |
492 | } else { |
427 | ($rhost, $rport, $rscheme, $rpath) = ($uhost, $uport, $uscheme, $upath); |
493 | ($rhost, $rport, $rscheme, $rpath) = ($uhost, $uport, $uscheme, $upath); |
428 | } |
494 | } |
429 | |
495 | |
430 | # leave out fragment and query string, just a heuristic |
496 | # leave out fragment and query string, just a heuristic |
431 | $hdr{referer} ||= "$uscheme://$uauthority$upath" unless exists $hdr{referer}; |
497 | $hdr{referer} = "$uscheme://$uauthority$upath" unless exists $hdr{referer}; |
432 | $hdr{"user-agent"} ||= $USERAGENT unless exists $hdr{"user-agent"}; |
498 | $hdr{"user-agent"} = $USERAGENT unless exists $hdr{"user-agent"}; |
433 | |
499 | |
434 | $hdr{"content-length"} = length $arg{body} |
500 | $hdr{"content-length"} = length $arg{body} |
435 | if length $arg{body} || $method ne "GET"; |
501 | if length $arg{body} || $method ne "GET"; |
436 | |
502 | |
|
|
503 | $hdr{connection} = "close TE"; #1.1 |
|
|
504 | $hdr{te} = "trailers" unless exists $hdr{te}; #1.1 |
|
|
505 | |
437 | my %state = (connect_guard => 1); |
506 | my %state = (connect_guard => 1); |
438 | |
507 | |
439 | _get_slot $uhost, sub { |
508 | _get_slot $uhost, sub { |
440 | $state{slot_guard} = shift; |
509 | $state{slot_guard} = shift; |
441 | |
510 | |
442 | return unless $state{connect_guard}; |
511 | return unless $state{connect_guard}; |
443 | |
512 | |
444 | $state{connect_guard} = AnyEvent::Socket::tcp_connect $rhost, $rport, sub { |
513 | my $connect_cb = sub { |
445 | $state{fh} = shift |
514 | $state{fh} = shift |
446 | or do { |
515 | or do { |
447 | my $err = "$!"; |
516 | my $err = "$!"; |
448 | %state = (); |
517 | %state = (); |
449 | return $cb->(undef, { Status => 599, Reason => $err, URL => $url }); |
518 | return $cb->(undef, { @pseudo, Status => 599, Reason => $err }); |
450 | }; |
519 | }; |
451 | |
520 | |
452 | pop; # free memory, save a tree |
521 | pop; # free memory, save a tree |
453 | |
522 | |
454 | return unless delete $state{connect_guard}; |
523 | return unless delete $state{connect_guard}; |
455 | |
524 | |
456 | # get handle |
525 | # get handle |
457 | $state{handle} = new AnyEvent::Handle |
526 | $state{handle} = new AnyEvent::Handle |
458 | fh => $state{fh}, |
527 | fh => $state{fh}, |
|
|
528 | peername => $rhost, |
|
|
529 | tls_ctx => $arg{tls_ctx}, |
|
|
530 | # these need to be reconfigured on keepalive handles |
459 | timeout => $timeout, |
531 | timeout => $timeout, |
460 | peername => $rhost, |
532 | on_error => sub { |
461 | tls_ctx => $arg{tls_ctx}; |
533 | %state = (); |
|
|
534 | $cb->(undef, { @pseudo, Status => 599, Reason => $_[2] }); |
|
|
535 | }, |
|
|
536 | on_eof => sub { |
|
|
537 | %state = (); |
|
|
538 | $cb->(undef, { @pseudo, Status => 599, Reason => "Unexpected end-of-file" }); |
|
|
539 | }, |
|
|
540 | ; |
462 | |
541 | |
463 | # limit the number of persistent connections |
542 | # limit the number of persistent connections |
464 | # keepalive not yet supported |
543 | # keepalive not yet supported |
465 | if ($KA_COUNT{$_[1]} < $MAX_PERSISTENT_PER_HOST) { |
544 | # if ($KA_COUNT{$_[1]} < $MAX_PERSISTENT_PER_HOST) { |
466 | ++$KA_COUNT{$_[1]}; |
545 | # ++$KA_COUNT{$_[1]}; |
467 | $state{handle}{ka_count_guard} = AnyEvent::Util::guard { |
546 | # $state{handle}{ka_count_guard} = AnyEvent::Util::guard { |
468 | --$KA_COUNT{$_[1]} |
547 | # --$KA_COUNT{$_[1]} |
469 | }; |
548 | # }; |
470 | $hdr{connection} = "keep-alive"; |
549 | # $hdr{connection} = "keep-alive"; |
471 | } else { |
|
|
472 | delete $hdr{connection}; |
|
|
473 | } |
550 | # } |
474 | |
|
|
475 | # (re-)configure handle |
|
|
476 | $state{handle}->on_error (sub { |
|
|
477 | %state = (); |
|
|
478 | $cb->(undef, { Status => 599, Reason => $_[2], URL => $url }); |
|
|
479 | }); |
|
|
480 | $state{handle}->on_eof (sub { |
|
|
481 | %state = (); |
|
|
482 | $cb->(undef, { Status => 599, Reason => "Unexpected end-of-file", URL => $url }); |
|
|
483 | }); |
|
|
484 | |
551 | |
485 | $state{handle}->starttls ("connect") if $rscheme eq "https"; |
552 | $state{handle}->starttls ("connect") if $rscheme eq "https"; |
486 | |
553 | |
487 | # handle actual, non-tunneled, request |
554 | # handle actual, non-tunneled, request |
488 | my $handle_actual_request = sub { |
555 | my $handle_actual_request = sub { |
489 | $state{handle}->starttls ("connect") if $uscheme eq "https" && !exists $state{handle}{tls}; |
556 | $state{handle}->starttls ("connect") if $uscheme eq "https" && !exists $state{handle}{tls}; |
490 | |
557 | |
491 | # send request |
558 | # send request |
492 | $state{handle}->push_write ( |
559 | $state{handle}->push_write ( |
493 | "$method $rpath HTTP/1.0\015\012" |
560 | "$method $rpath HTTP/1.1\015\012" |
494 | . (join "", map "\u$_: $hdr{$_}\015\012", grep defined $hdr{$_}, keys %hdr) |
561 | . (join "", map "\u$_: $hdr{$_}\015\012", grep defined $hdr{$_}, keys %hdr) |
495 | . "\015\012" |
562 | . "\015\012" |
496 | . (delete $arg{body}) |
563 | . (delete $arg{body}) |
497 | ); |
564 | ); |
498 | |
565 | |
499 | # return if error occured during push_write() |
566 | # return if error occured during push_write() |
500 | return unless %state; |
567 | return unless %state; |
501 | |
568 | |
502 | %hdr = (); # reduce memory usage, save a kitten |
569 | %hdr = (); # reduce memory usage, save a kitten, also make it possible to re-use |
503 | |
570 | |
504 | # status line |
571 | # status line and headers |
505 | $state{handle}->push_read (line => $qr_nl, sub { |
572 | $state{read_response} = sub { |
506 | $_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix |
|
|
507 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "Invalid server response ($_[1])", URL => $url })); |
|
|
508 | |
|
|
509 | my %hdr = ( # response headers |
|
|
510 | HTTPVersion => ",$1", |
|
|
511 | Status => ",$2", |
|
|
512 | Reason => ",$3", |
|
|
513 | URL => ",$url" |
|
|
514 | ); |
|
|
515 | |
|
|
516 | # headers, could be optimized a bit |
|
|
517 | $state{handle}->unshift_read (line => $qr_nlnl, sub { |
|
|
518 | for ("$_[1]") { |
573 | for ("$_[1]") { |
519 | y/\015//d; # weed out any \015, as they show up in the weirdest of places. |
574 | y/\015//d; # weed out any \015, as they show up in the weirdest of places. |
520 | |
575 | |
521 | # things seen, not parsed: |
576 | /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\012]*) )? \012/igxc |
522 | # p3pP="NON CUR OTPi OUR NOR UNI" |
577 | or return (%state = (), $cb->(undef, { @pseudo, Status => 599, Reason => "Invalid server response" })); |
523 | |
578 | |
524 | $hdr{lc $1} .= ",$2" |
579 | # 100 Continue handling |
525 | while /\G |
580 | # should not happen as we don't send expect: 100-continue, |
526 | ([^:\000-\037]*): |
581 | # but we handle it just in case. |
527 | [\011\040]* |
582 | # since we send the request body regardless, if we get an error |
528 | ((?: [^\012]+ | \012[\011\040] )*) |
583 | # we are out of-sync, which we currently do NOT handle correctly. |
529 | \012 |
584 | return $state{handle}->push_read (line => $qr_nlnl, $state{read_response}) |
530 | /gxc; |
585 | if $2 eq 100; |
531 | |
586 | |
|
|
587 | push @pseudo, |
|
|
588 | HTTPVersion => $1, |
|
|
589 | Status => $2, |
|
|
590 | Reason => $3, |
532 | /\G$/ |
591 | ; |
|
|
592 | |
|
|
593 | my $hdr = parse_hdr |
533 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "Garbled response headers", URL => $url })); |
594 | or return (%state = (), $cb->(undef, { @pseudo, Status => 599, Reason => "Garbled response headers" })); |
|
|
595 | |
|
|
596 | %hdr = (%$hdr, @pseudo); |
|
|
597 | } |
|
|
598 | |
|
|
599 | # redirect handling |
|
|
600 | # microsoft and other shitheads don't give a shit for following standards, |
|
|
601 | # try to support some common forms of broken Location headers. |
|
|
602 | if ($hdr{location} !~ /^(?: $ | [^:\/?\#]+ : )/x) { |
|
|
603 | $hdr{location} =~ s/^\.\/+//; |
|
|
604 | |
|
|
605 | my $url = "$rscheme://$uhost:$uport"; |
|
|
606 | |
|
|
607 | unless ($hdr{location} =~ s/^\///) { |
|
|
608 | $url .= $upath; |
|
|
609 | $url =~ s/\/[^\/]*$//; |
534 | } |
610 | } |
535 | |
611 | |
536 | substr $_, 0, 1, "" |
|
|
537 | for values %hdr; |
|
|
538 | |
|
|
539 | # redirect handling |
|
|
540 | # microsoft and other shitheads don't give a shit for following standards, |
|
|
541 | # try to support some common forms of broken Location headers. |
|
|
542 | if ($hdr{location} !~ /^(?: $ | [^:\/?\#]+ : )/x) { |
|
|
543 | $hdr{location} =~ s/^\.\/+//; |
|
|
544 | |
|
|
545 | my $url = "$rscheme://$uhost:$uport"; |
|
|
546 | |
|
|
547 | unless ($hdr{location} =~ s/^\///) { |
|
|
548 | $url .= $upath; |
|
|
549 | $url =~ s/\/[^\/]*$//; |
|
|
550 | } |
|
|
551 | |
|
|
552 | $hdr{location} = "$url/$hdr{location}"; |
612 | $hdr{location} = "$url/$hdr{location}"; |
|
|
613 | } |
|
|
614 | |
|
|
615 | my $redirect; |
|
|
616 | |
|
|
617 | if ($recurse) { |
|
|
618 | my $status = $hdr{Status}; |
|
|
619 | |
|
|
620 | # industry standard is to redirect POST as GET for |
|
|
621 | # 301, 302 and 303, in contrast to http/1.0 and 1.1. |
|
|
622 | # also, the UA should ask the user for 301 and 307 and POST, |
|
|
623 | # industry standard seems to be to simply follow. |
|
|
624 | # we go with the industry standard. |
|
|
625 | if ($status == 301 or $status == 302 or $status == 303) { |
|
|
626 | # HTTP/1.1 is unclear on how to mutate the method |
|
|
627 | $method = "GET" unless $method eq "HEAD"; |
|
|
628 | $redirect = 1; |
|
|
629 | } elsif ($status == 307) { |
|
|
630 | $redirect = 1; |
553 | } |
631 | } |
|
|
632 | } |
554 | |
633 | |
|
|
634 | my $finish = sub { # ($data, $err_status, $err_reason[, $keepalive]) |
555 | my $redirect; |
635 | my $keepalive = pop; |
556 | |
636 | |
|
|
637 | $state{handle}->destroy if $state{handle}; |
|
|
638 | %state = (); |
|
|
639 | |
|
|
640 | if (defined $_[1]) { |
|
|
641 | $hdr{OrigStatus} = $hdr{Status}; $hdr{Status} = $_[1]; |
|
|
642 | $hdr{OrigReason} = $hdr{Reason}; $hdr{Reason} = $_[2]; |
|
|
643 | } |
|
|
644 | |
|
|
645 | # set-cookie processing |
557 | if ($recurse) { |
646 | if ($arg{cookie_jar}) { |
558 | if ($hdr{Status} =~ /^30[12]$/ && $method ne "POST") { |
647 | for ($hdr{"set-cookie"}) { |
559 | # apparently, mozilla et al. just change POST to GET here |
648 | # parse NAME=VALUE |
560 | # more research is needed before we do the same |
649 | my @kv; |
|
|
650 | |
|
|
651 | while (/\G\s* ([^=;,[:space:]]+) \s*=\s* (?: "((?:[^\\"]+|\\.)*)" | ([^=;,[:space:]]*) )/gcxs) { |
561 | $redirect = 1; |
652 | my $name = $1; |
562 | } elsif ($hdr{Status} == 303) { |
653 | my $value = $3; |
563 | # even http/1.1 is unclear on how to mutate the method |
654 | |
564 | $method = "GET" unless $method eq "HEAD"; |
655 | unless ($value) { |
|
|
656 | $value = $2; |
|
|
657 | $value =~ s/\\(.)/$1/gs; |
|
|
658 | } |
|
|
659 | |
|
|
660 | push @kv, $name => $value; |
|
|
661 | |
|
|
662 | last unless /\G\s*;/gc; |
|
|
663 | } |
|
|
664 | |
|
|
665 | last unless @kv; |
|
|
666 | |
|
|
667 | my $name = shift @kv; |
|
|
668 | my %kv = (value => shift @kv, @kv); |
|
|
669 | |
|
|
670 | my $cdom; |
|
|
671 | my $cpath = (delete $kv{path}) || "/"; |
|
|
672 | |
|
|
673 | if (exists $kv{domain}) { |
|
|
674 | $cdom = delete $kv{domain}; |
|
|
675 | |
|
|
676 | $cdom =~ s/^\.?/./; # make sure it starts with a "." |
|
|
677 | |
|
|
678 | next if $cdom =~ /\.$/; |
|
|
679 | |
|
|
680 | # this is not rfc-like and not netscape-like. go figure. |
|
|
681 | my $ndots = $cdom =~ y/.//; |
|
|
682 | next if $ndots < ($cdom =~ /\.[^.][^.]\.[^.][^.]$/ ? 3 : 2); |
|
|
683 | } else { |
|
|
684 | $cdom = $uhost; |
|
|
685 | } |
|
|
686 | |
565 | $redirect = 1; |
687 | # store it |
566 | } elsif ($hdr{Status} == 307 && $method =~ /^(?:GET|HEAD)$/) { |
688 | $arg{cookie_jar}{version} = 1; |
567 | $redirect = 1; |
689 | $arg{cookie_jar}{$cdom}{$cpath}{$name} = \%kv; |
|
|
690 | |
|
|
691 | redo if /\G\s*,/gc; |
568 | } |
692 | } |
569 | } |
693 | } |
570 | |
694 | |
571 | my $finish = sub { |
|
|
572 | $state{handle}->destroy if $state{handle}; |
|
|
573 | %state = (); |
|
|
574 | |
|
|
575 | # set-cookie processing |
|
|
576 | if ($arg{cookie_jar}) { |
|
|
577 | for ($_[1]{"set-cookie"}) { |
|
|
578 | # parse NAME=VALUE |
|
|
579 | my @kv; |
|
|
580 | |
|
|
581 | while (/\G\s* ([^=;,[:space:]]+) \s*=\s* (?: "((?:[^\\"]+|\\.)*)" | ([^=;,[:space:]]*) )/gcxs) { |
|
|
582 | my $name = $1; |
|
|
583 | my $value = $3; |
|
|
584 | |
|
|
585 | unless ($value) { |
|
|
586 | $value = $2; |
|
|
587 | $value =~ s/\\(.)/$1/gs; |
|
|
588 | } |
|
|
589 | |
|
|
590 | push @kv, $name => $value; |
|
|
591 | |
|
|
592 | last unless /\G\s*;/gc; |
|
|
593 | } |
|
|
594 | |
|
|
595 | last unless @kv; |
|
|
596 | |
|
|
597 | my $name = shift @kv; |
|
|
598 | my %kv = (value => shift @kv, @kv); |
|
|
599 | |
|
|
600 | my $cdom; |
|
|
601 | my $cpath = (delete $kv{path}) || "/"; |
|
|
602 | |
|
|
603 | if (exists $kv{domain}) { |
|
|
604 | $cdom = delete $kv{domain}; |
|
|
605 | |
|
|
606 | $cdom =~ s/^\.?/./; # make sure it starts with a "." |
|
|
607 | |
|
|
608 | next if $cdom =~ /\.$/; |
|
|
609 | |
|
|
610 | # this is not rfc-like and not netscape-like. go figure. |
|
|
611 | my $ndots = $cdom =~ y/.//; |
|
|
612 | next if $ndots < ($cdom =~ /\.[^.][^.]\.[^.][^.]$/ ? 3 : 2); |
|
|
613 | } else { |
|
|
614 | $cdom = $uhost; |
|
|
615 | } |
|
|
616 | |
|
|
617 | # store it |
|
|
618 | $arg{cookie_jar}{version} = 1; |
|
|
619 | $arg{cookie_jar}{$cdom}{$cpath}{$name} = \%kv; |
|
|
620 | |
|
|
621 | redo if /\G\s*,/gc; |
|
|
622 | } |
|
|
623 | } |
|
|
624 | |
|
|
625 | if ($redirect && exists $hdr{location}) { |
695 | if ($redirect && exists $hdr{location}) { |
626 | # we ignore any errors, as it is very common to receive |
696 | # we ignore any errors, as it is very common to receive |
627 | # Content-Length != 0 but no actual body |
697 | # Content-Length != 0 but no actual body |
628 | # we also access %hdr, as $_[1] might be an erro |
698 | # we also access %hdr, as $_[1] might be an erro |
629 | http_request ($method => $hdr{location}, %arg, recurse => $recurse - 1, $cb); |
699 | http_request ( |
630 | } else { |
700 | $method => $hdr{location}, |
631 | $cb->($_[0], $_[1]); |
|
|
632 | } |
701 | %arg, |
|
|
702 | recurse => $recurse - 1, |
|
|
703 | Redirect => [$_[0], \%hdr], |
633 | }; |
704 | $cb); |
634 | |
|
|
635 | my $len = $hdr{"content-length"}; |
|
|
636 | |
|
|
637 | if (!$redirect && $arg{on_header} && !$arg{on_header}(\%hdr)) { |
|
|
638 | $finish->(undef, { Status => 598, Reason => "Request cancelled by on_header", URL => $url }); |
|
|
639 | } elsif ( |
|
|
640 | $hdr{Status} =~ /^(?:1..|[23]04)$/ |
|
|
641 | or $method eq "HEAD" |
|
|
642 | or (defined $len && !$len) |
|
|
643 | ) { |
|
|
644 | # no body |
|
|
645 | $finish->("", \%hdr); |
|
|
646 | } else { |
705 | } else { |
|
|
706 | $cb->($_[0], \%hdr); |
|
|
707 | } |
|
|
708 | }; |
|
|
709 | |
|
|
710 | my $len = $hdr{"content-length"}; |
|
|
711 | |
|
|
712 | if (!$redirect && $arg{on_header} && !$arg{on_header}(\%hdr)) { |
|
|
713 | $finish->(undef, 598 => "Request cancelled by on_header"); |
|
|
714 | } elsif ( |
|
|
715 | $hdr{Status} =~ /^(?:1..|204|205|304)$/ |
|
|
716 | or $method eq "HEAD" |
|
|
717 | or (defined $len && !$len) |
|
|
718 | ) { |
|
|
719 | # no body |
|
|
720 | $finish->("", undef, undef, 1); |
|
|
721 | } else { |
647 | # body handling, four different code paths |
722 | # body handling, many different code paths |
648 | # for want_body_handle, on_body (2x), normal (2x) |
723 | # - no body expected |
649 | # we might read too much here, but it does not matter yet (no pers. connections) |
724 | # - want_body_handle |
|
|
725 | # - te chunked |
|
|
726 | # - 2x length known (with or without on_body) |
|
|
727 | # - 2x length not known (with or without on_body) |
650 | if (!$redirect && $arg{want_body_handle}) { |
728 | if (!$redirect && $arg{want_body_handle}) { |
651 | $_[0]->on_eof (undef); |
729 | $_[0]->on_eof (undef); |
652 | $_[0]->on_error (undef); |
730 | $_[0]->on_error (undef); |
653 | $_[0]->on_read (undef); |
731 | $_[0]->on_read (undef); |
654 | |
732 | |
655 | $finish->(delete $state{handle}, \%hdr); |
733 | $finish->(delete $state{handle}); |
656 | |
734 | |
657 | } elsif ($arg{on_body}) { |
735 | } elsif ($hdr{"transfer-encoding"} =~ /\bchunked\b/i) { |
|
|
736 | my $cl = 0; |
|
|
737 | my $body = undef; |
|
|
738 | my $on_body = $arg{on_body} || sub { $body .= shift; 1 }; |
|
|
739 | |
658 | $_[0]->on_error (sub { $finish->(undef, { Status => 599, Reason => $_[2], URL => $url }) }); |
740 | $_[0]->on_error (sub { $finish->(undef, 599 => $_[2]) }); |
|
|
741 | |
|
|
742 | my $read_chunk; $read_chunk = sub { |
|
|
743 | $_[1] =~ /^([0-9a-fA-F]+)/ |
|
|
744 | or $finish->(undef, 599 => "Garbled chunked transfer encoding"); |
|
|
745 | |
|
|
746 | my $len = hex $1; |
|
|
747 | |
659 | if ($len) { |
748 | if ($len) { |
660 | $_[0]->on_eof (undef); |
749 | $cl += $len; |
|
|
750 | |
|
|
751 | $_[0]->push_read (chunk => $len, sub { |
|
|
752 | $on_body->($_[1], \%hdr) |
|
|
753 | or return $finish->(undef, 598 => "Request cancelled by on_body"); |
|
|
754 | |
661 | $_[0]->on_read (sub { |
755 | $_[0]->push_read (line => sub { |
662 | $len -= length $_[0]{rbuf}; |
756 | length $_[1] |
663 | |
757 | and return $finish->(undef, 599 => "Garbled chunked transfer encoding"); |
664 | $arg{on_body}(delete $_[0]{rbuf}, \%hdr) |
758 | $_[0]->push_read (line => $read_chunk); |
665 | or $finish->(undef, { Status => 598, Reason => "Request cancelled by on_body", URL => $url }); |
|
|
666 | |
|
|
667 | $len > 0 |
759 | }); |
668 | or $finish->("", \%hdr); |
|
|
669 | }); |
760 | }); |
670 | } else { |
761 | } else { |
671 | $_[0]->on_eof (sub { |
762 | $hdr{"content-length"} ||= $cl; |
672 | $finish->("", \%hdr); |
763 | |
|
|
764 | $_[0]->push_read (line => $qr_nlnl, sub { |
|
|
765 | if (length $_[1]) { |
|
|
766 | for ("$_[1]") { |
|
|
767 | y/\015//d; # weed out any \015, as they show up in the weirdest of places. |
|
|
768 | |
|
|
769 | my $hdr = parse_hdr |
|
|
770 | or return $finish->(undef, 599 => "Garbled response trailers"); |
|
|
771 | |
|
|
772 | %hdr = (%hdr, %$hdr); |
|
|
773 | } |
673 | }); |
774 | } |
674 | $_[0]->on_read (sub { |
775 | |
675 | $arg{on_body}(delete $_[0]{rbuf}, \%hdr) |
776 | $finish->($body, undef, undef, 1); |
676 | or $finish->(undef, { Status => 598, Reason => "Request cancelled by on_body", URL => $url }); |
|
|
677 | }); |
777 | }); |
678 | } |
778 | } |
|
|
779 | }; |
|
|
780 | |
|
|
781 | $_[0]->push_read (line => $read_chunk); |
|
|
782 | |
|
|
783 | } elsif ($arg{on_body}) { |
|
|
784 | $_[0]->on_error (sub { $finish->(undef, 599 => $_[2]) }); |
|
|
785 | |
|
|
786 | if ($len) { |
|
|
787 | $_[0]->on_read (sub { |
|
|
788 | $len -= length $_[0]{rbuf}; |
|
|
789 | |
|
|
790 | $arg{on_body}(delete $_[0]{rbuf}, \%hdr) |
|
|
791 | or return $finish->(undef, 598 => "Request cancelled by on_body"); |
|
|
792 | |
|
|
793 | $len > 0 |
|
|
794 | or $finish->("", undef, undef, 1); |
|
|
795 | }); |
679 | } else { |
796 | } else { |
680 | $_[0]->on_eof (undef); |
797 | $_[0]->on_eof (sub { |
681 | |
798 | $finish->(""); |
682 | if ($len) { |
799 | }); |
683 | $_[0]->on_error (sub { $finish->(undef, { Status => 599, Reason => $_[2], URL => $url }) }); |
|
|
684 | $_[0]->on_read (sub { |
800 | $_[0]->on_read (sub { |
|
|
801 | $arg{on_body}(delete $_[0]{rbuf}, \%hdr) |
|
|
802 | or $finish->(undef, 598 => "Request cancelled by on_body"); |
|
|
803 | }); |
|
|
804 | } |
|
|
805 | } else { |
|
|
806 | $_[0]->on_eof (undef); |
|
|
807 | |
|
|
808 | if ($len) { |
|
|
809 | $_[0]->on_error (sub { $finish->(undef, 599 => $_[2]) }); |
|
|
810 | $_[0]->on_read (sub { |
685 | $finish->((substr delete $_[0]{rbuf}, 0, $len, ""), \%hdr) |
811 | $finish->((substr delete $_[0]{rbuf}, 0, $len, ""), undef, undef, 1) |
686 | if $len <= length $_[0]{rbuf}; |
812 | if $len <= length $_[0]{rbuf}; |
687 | }); |
813 | }); |
688 | } else { |
814 | } else { |
689 | $_[0]->on_error (sub { |
815 | $_[0]->on_error (sub { |
690 | $! == Errno::EPIPE || !$! |
816 | ($! == Errno::EPIPE || !$!) |
691 | ? $finish->(delete $_[0]{rbuf}, \%hdr) |
817 | ? $finish->(delete $_[0]{rbuf}) |
692 | : $finish->(undef, { Status => 599, Reason => $_[2], URL => $url }); |
818 | : $finish->(undef, 599 => $_[2]); |
693 | }); |
819 | }); |
694 | $_[0]->on_read (sub { }); |
820 | $_[0]->on_read (sub { }); |
695 | } |
|
|
696 | } |
821 | } |
697 | } |
822 | } |
698 | }); |
823 | } |
699 | }); |
824 | }; |
|
|
825 | |
|
|
826 | $state{handle}->push_read (line => $qr_nlnl, $state{read_response}); |
700 | }; |
827 | }; |
701 | |
828 | |
702 | # now handle proxy-CONNECT method |
829 | # now handle proxy-CONNECT method |
703 | if ($proxy && $uscheme eq "https") { |
830 | if ($proxy && $uscheme eq "https") { |
704 | # oh dear, we have to wrap it into a connect request |
831 | # oh dear, we have to wrap it into a connect request |
705 | |
832 | |
706 | # maybe re-use $uauthority with patched port? |
833 | # maybe re-use $uauthority with patched port? |
707 | $state{handle}->push_write ("CONNECT $uhost:$uport HTTP/1.0\015\012Host: $uhost\015\012\015\012"); |
834 | $state{handle}->push_write ("CONNECT $uhost:$uport HTTP/1.0\015\012Host: $uhost\015\012\015\012"); |
708 | $state{handle}->push_read (line => $qr_nlnl, sub { |
835 | $state{handle}->push_read (line => $qr_nlnl, sub { |
709 | $_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix |
836 | $_[1] =~ /^HTTP\/([0-9\.]+) \s+ ([0-9]{3}) (?: \s+ ([^\015\012]*) )?/ix |
710 | or return (%state = (), $cb->(undef, { Status => 599, Reason => "Invalid proxy connect response ($_[1])", URL => $url })); |
837 | or return (%state = (), $cb->(undef, { @pseudo, Status => 599, Reason => "Invalid proxy connect response ($_[1])" })); |
711 | |
838 | |
712 | if ($2 == 200) { |
839 | if ($2 == 200) { |
713 | $rpath = $upath; |
840 | $rpath = $upath; |
714 | &$handle_actual_request; |
841 | &$handle_actual_request; |
715 | } else { |
842 | } else { |
716 | %state = (); |
843 | %state = (); |
717 | $cb->(undef, { Status => $2, Reason => $3, URL => $url }); |
844 | $cb->(undef, { @pseudo, Status => $2, Reason => $3 }); |
718 | } |
845 | } |
719 | }); |
846 | }); |
720 | } else { |
847 | } else { |
721 | &$handle_actual_request; |
848 | &$handle_actual_request; |
722 | } |
849 | } |
|
|
850 | }; |
723 | |
851 | |
724 | }, $arg{on_prepare} || sub { $timeout }; |
852 | my $tcp_connect = $arg{tcp_connect} |
|
|
853 | || do { require AnyEvent::Socket; \&AnyEvent::Socket::tcp_connect }; |
|
|
854 | |
|
|
855 | $state{connect_guard} = $tcp_connect->($rhost, $rport, $connect_cb, $arg{on_prepare} || sub { $timeout }); |
|
|
856 | |
725 | }; |
857 | }; |
726 | |
858 | |
727 | defined wantarray && AnyEvent::Util::guard { %state = () } |
859 | defined wantarray && AnyEvent::Util::guard { %state = () } |
728 | } |
860 | } |
729 | |
861 | |
… | |
… | |
743 | &http_request |
875 | &http_request |
744 | } |
876 | } |
745 | |
877 | |
746 | =back |
878 | =back |
747 | |
879 | |
|
|
880 | =head2 DNS CACHING |
|
|
881 | |
|
|
882 | AnyEvent::HTTP uses the AnyEvent::Socket::tcp_connect function for |
|
|
883 | the actual connection, which in turn uses AnyEvent::DNS to resolve |
|
|
884 | hostnames. The latter is a simple stub resolver and does no caching |
|
|
885 | on its own. If you want DNS caching, you currently have to provide |
|
|
886 | your own default resolver (by storing a suitable resolver object in |
|
|
887 | C<$AnyEvent::DNS::RESOLVER>). |
|
|
888 | |
748 | =head2 GLOBAL FUNCTIONS AND VARIABLES |
889 | =head2 GLOBAL FUNCTIONS AND VARIABLES |
749 | |
890 | |
750 | =over 4 |
891 | =over 4 |
751 | |
892 | |
752 | =item AnyEvent::HTTP::set_proxy "proxy-url" |
893 | =item AnyEvent::HTTP::set_proxy "proxy-url" |
… | |
… | |
755 | string of the form C<http://host:port> (optionally C<https:...>), croaks |
896 | string of the form C<http://host:port> (optionally C<https:...>), croaks |
756 | otherwise. |
897 | otherwise. |
757 | |
898 | |
758 | To clear an already-set proxy, use C<undef>. |
899 | To clear an already-set proxy, use C<undef>. |
759 | |
900 | |
|
|
901 | =item $date = AnyEvent::HTTP::format_date $timestamp |
|
|
902 | |
|
|
903 | Takes a POSIX timestamp (seconds since the epoch) and formats it as a HTTP |
|
|
904 | Date (RFC 2616). |
|
|
905 | |
|
|
906 | =item $timestamp = AnyEvent::HTTP::parse_date $date |
|
|
907 | |
|
|
908 | Takes a HTTP Date (RFC 2616) and returns the corresponding POSIX |
|
|
909 | timestamp, or C<undef> if the date cannot be parsed. |
|
|
910 | |
760 | =item $AnyEvent::HTTP::MAX_RECURSE |
911 | =item $AnyEvent::HTTP::MAX_RECURSE |
761 | |
912 | |
762 | The default value for the C<recurse> request parameter (default: C<10>). |
913 | The default value for the C<recurse> request parameter (default: C<10>). |
763 | |
914 | |
764 | =item $AnyEvent::HTTP::USERAGENT |
915 | =item $AnyEvent::HTTP::USERAGENT |
… | |
… | |
782 | connections. This number of can be useful for load-leveling. |
933 | connections. This number of can be useful for load-leveling. |
783 | |
934 | |
784 | =back |
935 | =back |
785 | |
936 | |
786 | =cut |
937 | =cut |
|
|
938 | |
|
|
939 | our @month = qw(Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec); |
|
|
940 | our @weekday = qw(Sun Mon Tue Wed Thu Fri Sat); |
|
|
941 | |
|
|
942 | sub format_date($) { |
|
|
943 | my ($time) = @_; |
|
|
944 | |
|
|
945 | # RFC 822/1123 format |
|
|
946 | my ($S, $M, $H, $mday, $mon, $year, $wday, $yday, undef) = gmtime $time; |
|
|
947 | |
|
|
948 | sprintf "%s, %02d %s %04d %02d:%02d:%02d GMT", |
|
|
949 | $weekday[$wday], $mday, $month[$mon], $year + 1900, |
|
|
950 | $H, $M, $S; |
|
|
951 | } |
|
|
952 | |
|
|
953 | sub parse_date($) { |
|
|
954 | my ($date) = @_; |
|
|
955 | |
|
|
956 | my ($d, $m, $y, $H, $M, $S); |
|
|
957 | |
|
|
958 | if ($date =~ /^[A-Z][a-z][a-z], ([0-9][0-9]) ([A-Z][a-z][a-z]) ([0-9][0-9][0-9][0-9]) ([0-9][0-9]):([0-9][0-9]):([0-9][0-9]) GMT$/) { |
|
|
959 | # RFC 822/1123, required by RFC 2616 |
|
|
960 | ($d, $m, $y, $H, $M, $S) = ($1, $2, $3, $4, $5, $6); |
|
|
961 | |
|
|
962 | } elsif ($date =~ /^[A-Z][a-z]+, ([0-9][0-9])-([A-Z][a-z][a-z])-([0-9][0-9]) ([0-9][0-9]):([0-9][0-9]):([0-9][0-9]) GMT$/) { |
|
|
963 | # RFC 850 |
|
|
964 | ($d, $m, $y, $H, $M, $S) = ($1, $2, $3 < 69 ? $3 + 2000 : $3 + 1900, $4, $5, $6); |
|
|
965 | |
|
|
966 | } elsif ($date =~ /^[A-Z][a-z][a-z] ([A-Z][a-z][a-z]) ([0-9 ][0-9]) ([0-9][0-9]):([0-9][0-9]):([0-9][0-9]) ([0-9][0-9][0-9][0-9])$/) { |
|
|
967 | # ISO C's asctime |
|
|
968 | ($d, $m, $y, $H, $M, $S) = ($2, $1, $6, $3, $4, $5); |
|
|
969 | } |
|
|
970 | # other formats fail in the loop below |
|
|
971 | |
|
|
972 | for (0..11) { |
|
|
973 | if ($m eq $month[$_]) { |
|
|
974 | require Time::Local; |
|
|
975 | return Time::Local::timegm ($S, $M, $H, $d, $_, $y); |
|
|
976 | } |
|
|
977 | } |
|
|
978 | |
|
|
979 | undef |
|
|
980 | } |
787 | |
981 | |
788 | sub set_proxy($) { |
982 | sub set_proxy($) { |
789 | if (length $_[0]) { |
983 | if (length $_[0]) { |
790 | $_[0] =~ m%^(https?):// ([^:/]+) (?: : (\d*) )?%ix |
984 | $_[0] =~ m%^(https?):// ([^:/]+) (?: : (\d*) )?%ix |
791 | or Carp::croak "$_[0]: invalid proxy URL"; |
985 | or Carp::croak "$_[0]: invalid proxy URL"; |
… | |
… | |
798 | # initialise proxy from environment |
992 | # initialise proxy from environment |
799 | eval { |
993 | eval { |
800 | set_proxy $ENV{http_proxy}; |
994 | set_proxy $ENV{http_proxy}; |
801 | }; |
995 | }; |
802 | |
996 | |
|
|
997 | =head2 SOCKS PROXIES |
|
|
998 | |
|
|
999 | Socks proxies are not directly supported by AnyEvent::HTTP. You can |
|
|
1000 | compile your perl to support socks, or use an external program such as |
|
|
1001 | F<socksify> (dante) or F<tsocks> to make your program use a socks proxy |
|
|
1002 | transparently. |
|
|
1003 | |
|
|
1004 | Alternatively, for AnyEvent::HTTP only, you can use your own |
|
|
1005 | C<tcp_connect> function that does the proxy handshake - here is an example |
|
|
1006 | that works with socks4a proxies: |
|
|
1007 | |
|
|
1008 | use Errno; |
|
|
1009 | use AnyEvent::Util; |
|
|
1010 | use AnyEvent::Socket; |
|
|
1011 | use AnyEvent::Handle; |
|
|
1012 | |
|
|
1013 | # host, port and username of/for your socks4a proxy |
|
|
1014 | my $socks_host = "10.0.0.23"; |
|
|
1015 | my $socks_port = 9050; |
|
|
1016 | my $socks_user = ""; |
|
|
1017 | |
|
|
1018 | sub socks4a_connect { |
|
|
1019 | my ($host, $port, $connect_cb, $prepare_cb) = @_; |
|
|
1020 | |
|
|
1021 | my $hdl = new AnyEvent::Handle |
|
|
1022 | connect => [$socks_host, $socks_port], |
|
|
1023 | on_prepare => sub { $prepare_cb->($_[0]{fh}) }, |
|
|
1024 | on_error => sub { $connect_cb->() }, |
|
|
1025 | ; |
|
|
1026 | |
|
|
1027 | $hdl->push_write (pack "CCnNZ*Z*", 4, 1, $port, 1, $socks_user, $host); |
|
|
1028 | |
|
|
1029 | $hdl->push_read (chunk => 8, sub { |
|
|
1030 | my ($hdl, $chunk) = @_; |
|
|
1031 | my ($status, $port, $ipn) = unpack "xCna4", $chunk; |
|
|
1032 | |
|
|
1033 | if ($status == 0x5a) { |
|
|
1034 | $connect_cb->($hdl->{fh}, (format_address $ipn) . ":$port"); |
|
|
1035 | } else { |
|
|
1036 | $! = Errno::ENXIO; $connect_cb->(); |
|
|
1037 | } |
|
|
1038 | }); |
|
|
1039 | |
|
|
1040 | $hdl |
|
|
1041 | } |
|
|
1042 | |
|
|
1043 | Use C<socks4a_connect> instead of C<tcp_connect> when doing C<http_request>s, |
|
|
1044 | possibly after switching off other proxy types: |
|
|
1045 | |
|
|
1046 | AnyEvent::HTTP::set_proxy undef; # usually you do not want other proxies |
|
|
1047 | |
|
|
1048 | http_get 'http://www.google.com', tcp_connect => \&socks4a_connect, sub { |
|
|
1049 | my ($data, $headers) = @_; |
|
|
1050 | ... |
|
|
1051 | }; |
|
|
1052 | |
803 | =head1 SEE ALSO |
1053 | =head1 SEE ALSO |
804 | |
1054 | |
805 | L<AnyEvent>. |
1055 | L<AnyEvent>. |
806 | |
1056 | |
807 | =head1 AUTHOR |
1057 | =head1 AUTHOR |