NAME IO::Callback::HTTP - read/write from HTTP URIs as if they were filehandles SYNOPSIS use IO::Callback::HTTP; my $fh = IO::Callback::HTTP->new("<", "http://www.example.com/"); while (my $line = <$fh>) { print $line; } DESCRIPTION This module allows you to read from and write to HTTP resources as if they were normal file handles (in fact, any non-HTTP resources supported by LWP::UserAgent ought to be OK too, including FTP, Gopher, etc). Why would you do this? Not for efficiency reasons, that's for sure. However, certain APIs expect to be passed filehandles; this module gives you those filehandles. Files can be opened in either read mode, using: my $fh = IO::Callback::HTTP->new('<', $request, %options); or write mode: my $fh = IO::Callback::HTTP->new('>', $request, %options); The $fh variable will then act like a normal Perl filehandle, but instead of interacting with a local file on disk, you'll be interacting with an HTTP resource on a remote server. $request can be a URI (either a string, or a blessed URI object), or it can be an HTTP::Request object. A URI is obviously simpler, but using an HTTP::Request object offers you more flexibility, such as the ability to change the HTTP method (defaults to GET for filehandles opened in read mode, and PUT for filehandles opened in write mode) or include particular HTTP headers (some of which are very useful: Accept, Content-Type, User-Agent, etc). Note that for a single filehandle, only one HTTP request is actually made. In the case of read mode, this happens on the first read. If no characters are read from the handle, then no request is made. In the case of write mode, the request happens once the file is closed. There are also a few options which can be passed to the constructor: "agent" An LWP::UserAgent object (or a subclass, such as WWW::Mechanize or LWPx::ParanoidAgent) that will actually make the request. This is optional; IO::Callback::HTTP does have its own pet UA that it can use if you don't provide one. "bytes" In read mode, if true, will make sure the data read from the handle is returned encoded as a UTF-8 byte string. If false, then the data read will be returned as a utf8 character string. In write mode, if true, will assume that you're writing bytes to the filehandle. If false, will assume that you're writing utf8 character strings to the filehandle, so will deal with encoding them to UTF-8 octets. Defaults to true. "failure" Set this to a coderef to trigger when the HTTP request fails (i.e. times out or non-2XX HTTP response code). It is passed a single parameter, which is the HTTP::Response object. Either way, IO::Callback::HTTP should do the correct thing, setting $! and so on. "success" Set this to a coderef to trigger when the HTTP request succeeds (i.e. 2XX HTTP response code). It is passed a single parameter, which is the HTTP::Response object. For filehandles in read mode, this is probably not especially useful, the fact that you can read from the file handle at all indicates that the request was successful. In write mode, it's more interesting as you may be interested in the result of a POST or PUT request. BUGS Please report any bugs to . SEE ALSO IO::Callback, LWP::UserAgent. IO::All::LWP does something similar, though it's less flexible. AUTHOR Toby Inkster . COPYRIGHT AND LICENCE This software is copyright (c) 2012 by Toby Inkster. This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language system itself. DISCLAIMER OF WARRANTIES THIS PACKAGE IS PROVIDED "AS IS" AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR A PARTICULAR PURPOSE.