- SEE ALSO
- METHODS AND FUNCTIONS
- $prot = new HTTP::Protocol;
- $prot = LWP::Protocol::create($url)
- $class = LWP::Protocol::implementor($scheme, [$class])
- $prot->collect($arg, $response, $collector)
- $prot->collect_once($arg, $response, $content)
LWP::Protocol - Base class for LWP protocols
package LWP::Protocol::foo; require LWP::Protocol; @ISA=qw(LWP::Protocol);
This class is used a the base class for all protocol implementations supported by the LWP library.
When creating an instance of this class using
LWP::Protocol::create($url), and you get an initialised subclass appropriate for that access method. In other words, the LWP::Protocol::create() function calls the constructor for one of its subclasses.
All derived LWP::Protocol classes need to override the request() method which is used to service a request. The overridden method can make use of the collect() function to collect together chunks of data as it is received.
Inspect the LWP/Protocol/file.pm and LWP/Protocol/http.pm files for examples of usage.
The LWP::Protocol constructor is inherited by subclasses. As this is a virtual base class this method should not be called directly.
Create an object of the class implementing the protocol to handle the given scheme. This is a function, not a method. It is more an object factory than a constructor. This is the function user agents should use to access protocols.
Get and/or set implementor class for a scheme. Returns '' if the specified scheme is not supported.
$response = $protocol->request($request, $proxy, undef); $response = $protocol->request($request, $proxy, '/tmp/sss'); $response = $protocol->request($request, $proxy, \&callback, 1024);
Dispactches a request over the protocol, and returns a response object. This method needs to be overridden in subclasses. Referer to LWP::UserAgent for description of the arguments.
Get and set the timeout value in seconds
Indicates if the library is allowed to use the core alarm() function to implement timeouts.
Should we initialize response headers from the <head> section of HTML documents.
Called to collect the content of a request, and process it appropriately into a scalar, file, or by calling a callback. If $arg is undefined, then the content is stored within the $response. If $arg is a simple scalar, then $arg is interpreted as a file name and the content is written to this file. If $arg is a reference to a routine, then content is passed to this routine.
The $collector is a routine that will be called and which is reponsible for returning pieces (as ref to scalar) of the content to process. The $collector signals EOF by returning a reference to an empty sting.
The return value from collect() is the $response object reference.
Note: We will only use the callback or file argument if $response->is_success(). This avoids sendig content data for redirects and authentization responses to the callback which would be confusing.
Can be called when the whole response content is available as $content. This will invoke collect() with a collector callback that returns a reference to $content the first time and an empty string the next.