use strict;
use Carp qw[ ];
-use List::Util qw[ first ];
-our $VERSION = '0.01';
+our $VERSION = '1.01';
our $TYPES = {
'application/octet-stream' => 'HTTP::Body::OctetStream',
'application/x-www-form-urlencoded' => 'HTTP::Body::UrlEncoded',
- 'multipart/form-data' => 'HTTP::Body::MultiPart'
+ 'multipart/form-data' => 'HTTP::Body::MultiPart',
+ 'multipart/related' => 'HTTP::Body::XFormsMultipart',
+ 'application/xml' => 'HTTP::Body::XForms'
};
+require HTTP::Body::OctetStream;
+require HTTP::Body::UrlEncoded;
+require HTTP::Body::MultiPart;
+require HTTP::Body::XFormsMultipart;
+require HTTP::Body::XForms;
+
+use HTTP::Headers;
+use HTTP::Message;
+
=head1 NAME
HTTP::Body - HTTP Body Parser
=head1 DESCRIPTION
-HTTP Body Parser.
+HTTP::Body parses chunks of HTTP POST data and supports
+application/octet-stream, application/x-www-form-urlencoded, and
+multipart/form-data.
+
+Chunked bodies are supported by not passing a length value to new().
+
+It is currently used by L<Catalyst> to parse POST bodies.
+
+=head1 NOTES
+
+When parsing multipart bodies, temporary files are created to store any
+uploaded files. You must delete these temporary files yourself after
+processing them.
=head1 METHODS
-=over 4
+=over 4
+
+=item new
+
+Constructor. Takes content type and content length as parameters,
+returns a L<HTTP::Body> object.
=cut
sub new {
my ( $class, $content_type, $content_length ) = @_;
- unless ( @_ == 3 ) {
- Carp::croak( $class, '->new( $content_type, $content_length )' );
+ unless ( @_ >= 2 ) {
+ Carp::croak( $class, '->new( $content_type, [ $content_length ] )' );
+ }
+
+ my $type;
+ foreach my $supported ( keys %{$TYPES} ) {
+ if ( index( lc($content_type), $supported ) >= 0 ) {
+ $type = $supported;
+ }
}
- my $type = first { index( lc($content_type), $_ ) >= 0 } keys %{$TYPES};
my $body = $TYPES->{ $type || 'application/octet-stream' };
eval "require $body";
my $self = {
buffer => '',
+ chunk_buffer => '',
body => undef,
- content_length => $content_length,
+ chunked => !defined $content_length,
+ content_length => defined $content_length ? $content_length : -1,
content_type => $content_type,
length => 0,
param => {},
=item add
+Add string to internal buffer. Will call spin unless done. returns
+length before adding self.
+
=cut
sub add {
my $self = shift;
+
+ if ( $self->{chunked} ) {
+ $self->{chunk_buffer} .= $_[0];
+
+ while ( $self->{chunk_buffer} =~ m/^([\da-fA-F]+).*\x0D\x0A/ ) {
+ my $chunk_len = hex($1);
+
+ if ( $chunk_len == 0 ) {
+ # Strip chunk len
+ $self->{chunk_buffer} =~ s/^([\da-fA-F]+).*\x0D\x0A//;
+
+ # End of data, there may be trailing headers
+ if ( my ($headers) = $self->{chunk_buffer} =~ m/(.*)\x0D\x0A/s ) {
+ if ( my $message = HTTP::Message->parse( $headers ) ) {
+ $self->{trailing_headers} = $message->headers;
+ }
+ }
+
+ $self->{chunk_buffer} = '';
+
+ # Set content_length equal to the amount of data we read,
+ # so the spin methods can finish up.
+ $self->{content_length} = $self->{length};
+ }
+ else {
+ # Make sure we have the whole chunk in the buffer (+CRLF)
+ if ( length( $self->{chunk_buffer} ) >= $chunk_len ) {
+ # Strip chunk len
+ $self->{chunk_buffer} =~ s/^([\da-fA-F]+).*\x0D\x0A//;
+
+ # Pull chunk data out of chunk buffer into real buffer
+ $self->{buffer} .= substr $self->{chunk_buffer}, 0, $chunk_len, '';
+
+ # Strip remaining CRLF
+ $self->{chunk_buffer} =~ s/^\x0D\x0A//;
+
+ $self->{length} += $chunk_len;
+ }
+ else {
+ # Not enough data for this chunk, wait for more calls to add()
+ return;
+ }
+ }
+
+ unless ( $self->{state} eq 'done' ) {
+ $self->spin;
+ }
+ }
+
+ return;
+ }
+
+ my $cl = $self->content_length;
if ( defined $_[0] ) {
- $self->{buffer} .= $_[0];
$self->{length} += length( $_[0] );
+
+ # Don't allow buffer data to exceed content-length
+ if ( $self->{length} > $cl ) {
+ $_[0] = substr $_[0], 0, $cl - $self->{length};
+ $self->{length} = $cl;
+ }
+
+ $self->{buffer} .= $_[0];
}
unless ( $self->state eq 'done' ) {
$self->spin;
}
- return ( $self->length - $self->content_length );
+ return ( $self->length - $cl );
}
=item body
+accessor for the body.
+
=cut
sub body {
return $self->{body};
}
-=item buffer
+=item chunked
+
+Returns 1 if the request is chunked.
=cut
-sub buffer {
- return shift->{buffer};
+sub chunked {
+ return shift->{chunked};
}
=item content_length
+Returns the content-length for the body data if known.
+Returns -1 if the request is chunked.
+
=cut
sub content_length {
=item content_type
+Returns the content-type of the body data.
+
=cut
sub content_type {
=item init
+return self.
+
=cut
sub init {
=item length
+Returns the total length of data we expect to read if known.
+In the case of a chunked request, returns the amount of data
+read so far.
+
=cut
sub length {
return shift->{length};
}
+=item trailing_headers
+
+If a chunked request body had trailing headers, trailing_headers will
+return an HTTP::Headers object populated with those headers.
+
+=cut
+
+sub trailing_headers {
+ return shift->{trailing_headers};
+}
+
=item spin
+Abstract method to spin the io handle.
+
=cut
sub spin {
=item state
+Returns the current state of the parser.
+
=cut
sub state {
=item param
+Get/set body parameters.
+
=cut
sub param {
=item upload
+Get/set file uploads.
+
=cut
sub upload {
Sebastian Riedel, C<sri@cpan.org>
+Andy Grundman, C<andy@hybridized.org>
+
=head1 LICENSE
This library is free software. You can redistribute it and/or modify