1http(n) Tcl Bundled Packages http(n)
2______________________________________________________________________________
6
8 http - Client-side implementation of the HTTP/1.0 protocol.
9
11 package require http ?2.5?
12 ::http::config ?options?
14 ::http::geturl url ?options?
16 ::http::formatQuery key value ?key value ...?
18 ::http::reset token ?why?
20 ::http::wait token
22 ::http::status token
24 ::http::size token
26 ::http::code token
28 ::http::ncode token
30 ::http::data token
32 ::http::error token
34 ::http::cleanup token
36 ::http::register proto port command
38 ::http::unregister proto
40_________________________________________________________________
41
44 The http package provides the client side of the HTTP/1.0 protocol.
45 The package implements the GET, POST, and HEAD operations of HTTP/1.0.
46 It allows configuration of a proxy host to get through firewalls. The
47 package is compatible with the Safesock security policy, so it can be
48 used by untrusted applets to do URL fetching from a restricted set of
49 hosts. This package can be extended to support additional HTTP trans‐
50 port protocols, such as HTTPS, by providing a custom socket command,
51 via ::http::register.
52 The ::http::geturl procedure does a HTTP transaction. Its options
54 determine whether a GET, POST, or HEAD transaction is performed. The
55 return value of ::http::geturl is a token for the transaction. The
56 value is also the name of an array in the ::http namespace that con‐
57 tains state information about the transaction. The elements of this
58 array are described in the STATE ARRAY section.
59 If the -command option is specified, then the HTTP operation is done in
61 the background. ::http::geturl returns immediately after generating
62 the HTTP request and the callback is invoked when the transaction com‐
63 pletes. For this to work, the Tcl event loop must be active. In Tk
64 applications this is always true. For pure-Tcl applications, the call‐
65 er can use ::http::wait after calling ::http::geturl to start the event
66 loop.
67
69 ::http::config ?options?
70 The ::http::config command is used to set and query the name of
71 the proxy server and port, and the User-Agent name used in the
72 HTTP requests. If no options are specified, then the current
73 configuration is returned. If a single argument is specified,
74 then it should be one of the flags described below. In this
75 case the current value of that setting is returned. Otherwise,
76 the options should be a set of flags and values that define the
77 configuration:
78 -accept mimetypes
80 The Accept header of the request. The default is */*,
81 which means that all types of documents are accepted.
82 Otherwise you can supply a comma-separated list of mime
83 type patterns that you are willing to receive. For exam‐
84 ple, "image/gif, image/jpeg, text/*".
85 -proxyhost hostname
87 The name of the proxy host, if any. If this value is the
88 empty string, the URL host is contacted directly.
89 -proxyport number
91 The proxy port number.
92 -proxyfilter command
94 The command is a callback that is made during
95 ::http::geturl to determine if a proxy is required for a
96 given host. One argument, a host name, is added to com‐
97 mand when it is invoked. If a proxy is required, the
98 callback should return a two-element list containing the
99 proxy server and proxy port. Otherwise the filter should
100 return an empty list. The default filter returns the
101 values of the -proxyhost and -proxyport settings if they
102 are non-empty.
103 -urlencoding encoding
105 The encoding used for creating the x-url-encoded URLs
106 with ::http::formatQuery. The default is utf-8, as spec‐
107 ified by RFC 2718. Prior to http 2.5 this was unspeci‐
108 fied, and that behavior can be returned by specifying the
109 empty string ({}), although iso8859-1 is recommended to
110 restore similar behavior but without the ::http::format‐
111 Query throwing an error processing non-latin-1 charac‐
112 ters.
113 -useragent string
115 The value of the User-Agent header in the HTTP request.
116 The default is "Tcl http client package 2.4."
117 ::http::geturl url ?options?
119 The ::http::geturl command is the main procedure in the package.
120 The -query option causes a POST operation and the -validate
121 option causes a HEAD operation; otherwise, a GET operation is
122 performed. The ::http::geturl command returns a token value
123 that can be used to get information about the transaction. See
124 the STATE ARRAY and ERRORS section for details. The
125 ::http::geturl command blocks until the operation completes,
126 unless the -command option specifies a callback that is invoked
127 when the HTTP transaction completes. ::http::geturl takes sev‐
128 eral options:
129 -binary boolean
131 Specifies whether to force interpreting the URL data as
132 binary. Normally this is auto-detected (anything not
133 beginning with a text content type or whose content
134 encoding is gzip or compress is considered binary data).
135 -blocksize size
137 The block size used when reading the URL. At most size
138 bytes are read at once. After each block, a call to the
139 -progress callback is made (if that option is specified).
140 -channel name
142 Copy the URL contents to channel name instead of saving
143 it in state(body).
144 -command callback
146 Invoke callback after the HTTP transaction completes.
147 This option causes ::http::geturl to return immediately.
148 The callback gets an additional argument that is the
149 token returned from ::http::geturl. This token is the
150 name of an array that is described in the STATE ARRAY
151 section. Here is a template for the callback:
152 proc httpCallback {token} {
153 upvar #0 $token state
154 # Access state as a Tcl array
155 }
156 -handler callback
158 Invoke callback whenever HTTP data is available; if
159 present, nothing else will be done with the HTTP data.
160 This procedure gets two additional arguments: the socket
161 for the HTTP data and the token returned from
162 ::http::geturl. The token is the name of a global array
163 that is described in the STATE ARRAY section. The proce‐
164 dure is expected to return the number of bytes read from
165 the socket. Here is a template for the callback:
166 proc httpHandlerCallback {socket token} {
167 upvar #0 $token state
168 # Access socket, and state as a Tcl array
169 ...
170 (example: set data [read $socket 1000];set nbytes [string length $data])
171 ...
172 return nbytes
173 }
174 -headers keyvaluelist
176 This option is used to add extra headers to the HTTP
177 request. The keyvaluelist argument must be a list with
178 an even number of elements that alternate between keys
179 and values. The keys become header field names. New‐
180 lines are stripped from the values so the header cannot
181 be corrupted. For example, if keyvaluelist is Pragma no-
182 cache then the following header is included in the HTTP
183 request:
184 Pragma: no-cache
185 -progress callback
187 The callback is made after each transfer of data from the
188 URL. The callback gets three additional arguments: the
189 token from ::http::geturl, the expected total size of the
190 contents from the Content-Length meta-data, and the cur‐
191 rent number of bytes transferred so far. The expected
192 total size may be unknown, in which case zero is passed
193 to the callback. Here is a template for the progress
194 callback:
195 proc httpProgress {token total current} {
196 upvar #0 $token state
197 }
198 -query query
200 This flag causes ::http::geturl to do a POST request that
201 passes the query to the server. The query must be an x-
202 url-encoding formatted query. The ::http::formatQuery
203 procedure can be used to do the formatting.
204 -queryblocksize size
206 The block size used when posting query data to the URL.
207 At most size bytes are written at once. After each
208 block, a call to the -queryprogress callback is made (if
209 that option is specified).
210 -querychannel channelID
212 This flag causes ::http::geturl to do a POST request that
213 passes the data contained in channelID to the server. The
214 data contained in channelID must be an x-url-encoding
215 formatted query unless the -type option below is used.
216 If a Content-Length header is not specified via the
217 -headers options, ::http::geturl attempts to determine
218 the size of the post data in order to create that header.
219 If it is unable to determine the size, it returns an
220 error.
221 -queryprogress callback
223 The callback is made after each transfer of data to the
224 URL (i.e. POST) and acts exactly like the -progress
225 option (the callback format is the same).
226 -timeout milliseconds
228 If milliseconds is non-zero, then ::http::geturl sets up
229 a timeout to occur after the specified number of mil‐
230 liseconds. A timeout results in a call to ::http::reset
231 and to the -command callback, if specified. The return
232 value of ::http::status is timeout after a timeout has
233 occurred.
234 -type mime-type
236 Use mime-type as the Content-Type value, instead of the
237 default value (application/x-www-form-urlencoded) during
238 a POST operation.
239 -validate boolean
241 If boolean is non-zero, then ::http::geturl does an HTTP
242 HEAD request. This request returns meta information
243 about the URL, but the contents are not returned. The
244 meta information is available in the state(meta) vari‐
245 able after the transaction. See the STATE ARRAY section
246 for details.
247 ::http::formatQuery key value ?key value ...?
249 This procedure does x-url-encoding of query data. It takes an
250 even number of arguments that are the keys and values of the
251 query. It encodes the keys and values, and generates one string
252 that has the proper & and = separators. The result is suitable
253 for the -query value passed to ::http::geturl.
254 ::http::reset token ?why?
256 This command resets the HTTP transaction identified by token, if
257 any. This sets the state(status) value to why, which defaults
258 to reset, and then calls the registered -command callback.
259 ::http::wait token
261 This is a convenience procedure that blocks and waits for the
262 transaction to complete. This only works in trusted code
263 because it uses vwait. Also, it's not useful for the case where
264 ::http::geturl is called without the -command option because in
265 this case the ::http::geturl call doesn't return until the HTTP
266 transaction is complete, and thus there's nothing to wait for.
267 ::http::data token
269 This is a convenience procedure that returns the body element
270 (i.e., the URL data) of the state array.
271 ::http::error token
273 This is a convenience procedure that returns the error element
274 of the state array.
275 ::http::status token
277 This is a convenience procedure that returns the status element
278 of the state array.
279 ::http::code token
281 This is a convenience procedure that returns the http element of
282 the state array.
283 ::http::ncode token
285 This is a convenience procedure that returns just the numeric
286 return code (200, 404, etc.) from the http element of the state
287 array.
288 ::http::size token
290 This is a convenience procedure that returns the currentsize
291 element of the state array, which represents the number of bytes
292 received from the URL in the ::http::geturl call.
293 ::http::cleanup token
295 This procedure cleans up the state associated with the connec‐
296 tion identified by token. After this call, the procedures like
297 ::http::data cannot be used to get information about the opera‐
298 tion. It is strongly recommended that you call this function
299 after you're done with a given HTTP request. Not doing so will
300 result in memory not being freed, and if your app calls
301 ::http::geturl enough times, the memory leak could cause a per‐
302 formance hit...or worse.
303 ::http::register proto port command
305 This procedure allows one to provide custom HTTP transport types
306 such as HTTPS, by registering a prefix, the default port, and
307 the command to execute to create the Tcl channel. E.g.:
308 package require http
309 package require tls
310 ::http::register https 443 ::tls::socket
312 set token [::http::geturl https://my.secure.site/]
314 ::http::unregister proto
316 This procedure unregisters a protocol handler that was previ‐
317 ously registered via ::http::register.
318
321 The ::http::geturl procedure will raise errors in the following cases:
322 invalid command line options, an invalid URL, a URL on a non-existent
323 host, or a URL at a bad port on an existing host. These errors mean
324 that it cannot even start the network transaction. It will also raise
325 an error if it gets an I/O error while writing out the HTTP request
326 header. For synchronous ::http::geturl calls (where -command is not
327 specified), it will raise an error if it gets an I/O error while read‐
328 ing the HTTP reply headers or data. Because ::http::geturl doesn't
329 return a token in these cases, it does all the required cleanup and
330 there's no issue of your app having to call ::http::cleanup.
331 For asynchronous ::http::geturl calls, all of the above error situa‐
333 tions apply, except that if there's any error while reading the HTTP
334 reply headers or data, no exception is thrown. This is because after
335 writing the HTTP headers, ::http::geturl returns, and the rest of the
336 HTTP transaction occurs in the background. The command callback can
337 check if any error occurred during the read by calling ::http::status
338 to check the status and if its error, calling ::http::error to get the
339 error message.
340 Alternatively, if the main program flow reaches a point where it needs
342 to know the result of the asynchronous HTTP request, it can call
343 ::http::wait and then check status and error, just as the callback
344 does.
345 In any case, you must still call ::http::cleanup to delete the state
347 array when you're done.
348 There are other possible results of the HTTP transaction determined by
350 examining the status from ::http::status. These are described below.
351 ok If the HTTP transaction completes entirely, then status will be
353 ok. However, you should still check the ::http::code value to
354 get the HTTP status. The ::http::ncode procedure provides just
355 the numeric error (e.g., 200, 404 or 500) while the ::http::code
356 procedure returns a value like "HTTP 404 File not found".
357 eof If the server closes the socket without replying, then no error
359 is raised, but the status of the transaction will be eof.
360 error The error message will also be stored in the error status array
362 element, accessible via ::http::error.
363 Another error possibility is that ::http::geturl is unable to write all
365 the post query data to the server before the server responds and closes
366 the socket. The error message is saved in the posterror status array
367 element and then ::http::geturl attempts to complete the transaction.
368 If it can read the server's response it will end up with an ok status,
369 otherwise it will have an eof status.
370
373 The ::http::geturl procedure returns a token that can be used to get to
374 the state of the HTTP transaction in the form of a Tcl array. Use this
375 construct to create an easy-to-use array variable:
376 upvar #0 $token state
377 Once the data associated with the URL is no longer needed, the state
378 array should be unset to free up storage. The ::http::cleanup proce‐
379 dure is provided for that purpose. The following elements of the array
380 are supported:
381 body The contents of the URL. This will be empty if the
383 -channel option has been specified. This value is
384 returned by the ::http::data command.
385 charset
387 The value of the charset attribute from the Content-Type
388 meta-data value. If none was specified, this defaults to
389 the RFC standard iso8859-1, or the value of
390 $::http::defaultCharset. Incoming text data will be
391 automatically converted from this charset to utf-8.
392 coding A copy of the Content-Encoding meta-data value.
394 currentsize
396 The current number of bytes fetched from the URL. This
397 value is returned by the ::http::size command.
398 error If defined, this is the error string seen when the HTTP
400 transaction was aborted.
401 http The HTTP status reply from the server. This value is
403 returned by the ::http::code command. The format of this
404 value is:
405 HTTP/1.0 code string
406 The code is a three-digit number defined in the HTTP
407 standard. A code of 200 is OK. Codes beginning with 4
408 or 5 indicate errors. Codes beginning with 3 are redi‐
409 rection errors. In this case the Location meta-data
410 specifies a new URL that contains the requested informa‐
411 tion.
412 meta The HTTP protocol returns meta-data that describes the
414 URL contents. The meta element of the state array is a
415 list of the keys and values of the meta-data. This is in
416 a format useful for initializing an array that just con‐
417 tains the meta-data:
418 array set meta $state(meta)
419 Some of the meta-data keys are listed below, but the HTTP
420 standard defines more, and servers are free to add their
421 own.
422 Content-Type
424 The type of the URL contents. Examples include
425 text/html, image/gif, application/postscript and
426 application/x-tcl.
427 Content-Length
429 The advertised size of the contents. The actual
430 size obtained by ::http::geturl is available as
431 state(size).
432 Location
434 An alternate URL that contains the requested data.
435 posterror
437 The error, if any, that occurred while writing the post
438 query data to the server.
439 status Either ok, for successful completion, reset for user-
441 reset, timeout if a timeout occurred before the transac‐
442 tion could complete, or error for an error condition.
443 During the transaction this value is the empty string.
444 totalsize
446 A copy of the Content-Length meta-data value.
447 type A copy of the Content-Type meta-data value.
449 url The requested URL.
451
453 # Copy a URL to a file and print meta-data
454 proc httpcopy { url file {chunk 4096} } {
455 set out [open $file w]
456 set token [::http::geturl $url -channel $out \
457 -progress httpCopyProgress -blocksize $chunk]
458 close $out
459 # This ends the line started by httpCopyProgress
461 puts stderr ""
462 upvar #0 $token state
464 set max 0
465 foreach {name value} $state(meta) {
466 if {[string length $name] > $max} {
467 set max [string length $name]
468 }
469 if {[regexp -nocase ^location$ $name]} {
470 # Handle URL redirects
471 puts stderr "Location:$value"
472 return [httpcopy [string trim $value] $file $chunk]
473 }
474 }
475 incr max
476 foreach {name value} $state(meta) {
477 puts [format "%-*s %s" $max $name: $value]
478 }
479 return $token
481 }
482 proc httpCopyProgress {args} {
483 puts -nonewline stderr .
484 flush stderr
485 }
486
489 safe(n), socket(n), safesock(n)
490
493 security policy, socket
494http 2.5 http(n)