I have a server program that listens on 9000 port. But I can't find a way to write a client program for that server that connects server at 9000 port. Here is the main part of server program:
use strict;
use warnings;
use Net::WebSocket::Server;
my $port = "9000";
my $msg_count = 0;
print "starting server on $port \n\n";
my $count = 2400;
Net::WebSocket::Server->new(
listen => $port,
silence_max => 5,
tick_period => 300,
on_tick => sub {
my ($serv) = #_;
print "connections >> " . $serv->connections . "\n";
print $_->ip() for( $serv->connections() ); print "\n";
print $_->port() for( $serv->connections() ); print "\n\n";
$count++;
},
on_connect => sub {
my ($serv, $conn) = #_;
$conn->on(
handshake => sub {
my ($conn, $handshake) = #_;
my $tmp = $handshake->req->origin;
print "here ... $tmp \n\n";
},
utf8 => sub {
my ($conn, $msg) = #_;
my $IP = $conn->ip();
my $PORT = $conn->port();
my $SERVER = $conn->server();
my $SOCKET = $conn->socket();
my $str = Dumper $SOCKET;
I searched internet and what that sounds understandable to me is the following client program:
use strict;
use warnings;
use IO::Socket::SSL;
my $cl=IO::Socket::SSL->new("http://localhost:9000") or die "error=$!, ssl_error=$SSL_ERROR";
if($cl) {
$cl->connect_SSL or die $#;
# Something about certificates?
$cl->syswrite("Command");
close($cl);
}
But its not working. The error client program generates is as follows:
Expected 'PeerService' at client2.pl line 5.
I am newbie in Socket programming and currently understanding websockets programming in Perl.
Note: I am on windows platform.
I ran the example code suggested https://stackoverflow.com/questions/37318581/simple-perl-websocket-client. It gives error "Can't use an undefined value as a subroutine reference at C:/Strawberry/perl/site/lib/Protocol/WebSocket/Client.pm line 103.":
use strict;
use warnings;
use Protocol::WebSocket::Client;
my $client = Protocol::WebSocket::Client->new(url => 'ws://localhost:9000') or die "$!";
my $reply = "Free\n";
# Sends a correct handshake header
$client->connect or die "$!";
# Register on connect handler
$client->on(
connect => sub {
$client->write('hi there');
}
) or die "$!";
# Parses incoming data and on every frame calls on_read
$client->read($reply);
print "$reply\n";
# Sends correct close header
$client->disconnect;
Please investigate following demo code snippets for WebSocket Server and Client.
Note: please do not forget to alter code to match your server origin (ip address and port)
use strict;
use warnings;
use feature 'say';
use Net::WebSocket::Server;
my $origin = 'http://192.168.1.160:8080'; # server origin
my $port = 8080;
$| = 1;
say "Starting server on $port";
Net::WebSocket::Server->new(
listen => $port,
tick_period => 60,
on_tick => sub {
my ($serv) = #_;
my $stamp = 'Server time: ' . scalar localtime;
$_->send_utf8($stamp) for $serv->connections;
},
on_connect => sub {
my ($serv, $conn) = #_;
$conn->on(
handshake => sub {
my ($conn, $handshake) = #_;
$conn->disconnect() unless $handshake->req->origin eq $origin;
},
ready => sub {
my ($conn) = #_;
say "Client: connect IP $conn->{ip} PORT $conn->{port}";
my $msg = 'Connected server time is ' . scalar localtime . "\n";
$_->send_utf8($msg) for $conn->server->connections;
},
utf8 => sub {
my ($conn, $msg) = #_;
say "Client message: $conn->{ip} $msg";
$_->send_utf8('Server reply: ' . $msg)
for $conn->server->connections;
$conn->disconnect() if $msg eq 'exit';
},
binary => sub {
my ($conn, $msg) = #_;
$_->send_binary($msg) for $conn->server->connections;
},
pong => sub {
my ($conn, $msg) = #_;
$_->send_utf8($msg) for $conn->server->connections;
},
disconnect => sub {
my ($conn, $code, $reason) = #_;
say "Client: disconnect IP $conn->{ip} PORT $conn->{port}";
},
);
},
)->start;
Client
use strict;
use warnings;
use feature 'say';
use IO::Async::Loop;
use Net::Async::WebSocket::Client;
my $HOST = '192.168.1.160';
my $PORT = 8080;
my $loop = IO::Async::Loop->new;
my $client = Net::Async::WebSocket::Client->new(
on_text_frame => sub {
my ( $self, $frame ) = #_;
say $frame;
},
);
my $input = IO::Async::Stream->new_for_stdin(
on_read => sub {
my ( $self, $buffref, $eof ) = #_;
my $msg;
$msg = $1 while $$buffref =~ s/^(.*)\n//;
$client->send_text_frame( $msg );
$loop->loop_stop if $msg eq 'exit';
return 0;
},
);
$loop->add( $client );
$loop->add( $input );
$client->connect(
url => "ws://$HOST:$PORT/"
)->then( sub {
say 'Successfully connected to server';
$client->send_text_frame( scalar localtime );
})->get;
$loop->run;
say 'Bye, until next time';
exit 0;
References:
Net::WebSocket::Server
Net::Async::WebSocket::Client
IO::Async::Loop
Anybody help me make a virtual file system in Perl.
Very simple, 2 depth level, as
/subdir
subdir-l2
file2.txt
/file1.txt
I try use Fuse.pm, but not understand how create subdir level. I create %files hash, and if go to subdir, recreate it with new records. It's for test only.
#!/usr/bin/env perl
use strict;
use warnings;
use utf8;
use Fuse;
use POSIX qw(ENOENT EISDIR EINVAL);
my (%files) = (
'.' => {
type => 0040,
mode => 0755,
ctime => 1490603721
},
subdir => {
type => 0040,
mode => 0755,
ctime => 1490603721
},
"file1.txt" => {
type => 0100,
mode => 0755,
ctime => 1490603721
}
);
sub filename_fixup {
my ($file) = shift;
$file =~ s,^/,,;
$file = '.' unless length($file);
return $file;
}
sub getdir {
my $tmp = shift;
if ($tmp eq '/') {
return (keys %files),0;
} else {
(%files) = (
'.' => {
type => 0040,
mode => 0755,
ctime => 1490603721
},
# /subdir/subdir-l2
"subdir-l2" => {
type => 0040,
mode => 0755,
ctime => 1490603721
} ,
# /subdir/a-l2.file
"file2.txt" => {
cont => "File 'al2'.\n",
type => 0100,
mode => 0755,
ctime => 1490603721
}
);
return (keys %files),0;
}
}
sub getattr {
my ($file) = filename_fixup(shift);
$file =~ s,^/,,;
$file = '.' unless length($file);
return -ENOENT() unless exists($files{$file});
my ($size) = exists($files{$file}{cont}) ? length($files{$file}{cont}) : 0;
$size = $files{$file}{size} if exists $files{$file}{size};
my ($modes) = ($files{$file}{type}<<9) + $files{$file}{mode};
my ($dev, $ino, $rdev, $blocks, $gid, $uid, $nlink, $blksize) = (0,0,0,1,0,0,1,1024);
my ($atime, $ctime, $mtime);
$atime = $ctime = $mtime = $files{$file}{ctime};
return ($dev,$ino,$modes,$nlink,$uid,$gid,$rdev,$size,$atime,$mtime,$ctime,$blksize,$blocks);
}
Fuse::main(
mountpoint => "/tmp/123",
getdir => \&getdir,
getattr => \&getattr,
);
one level mount fine, but if go to deeper i get
?????????? ? ? ? ? ? file2.txt
?????????? ? ? ? ? ? subdir-l2
I'm really not a regular user of the Fuse module, neither of FUSE system. Tinkered with this issue out of pure curiosity. Thus, although I can't explain in very much details how to use the plain Fuse module to achieve your goal, I have a working code that does create the wanted filesystem (at least on my system, and seems that it is capable of creating any arbitrary filesystem tree), and I can explain how I got this code working.
So first of all I discovered the Fuse::Simple module on CPAN.
Its SYNOPSIS shows that it provides a really simple API to the Fuse module for creating arbitrary filesystems from a hash structure. Its source code isn't that huge, so I just created 'listing.pl' script file and copied there most of the functions (except fserr that caused a Modification of a read-only value exception), put the main sub contents out, so they will be the main script's flow, hardcoded the filesystem structure ($fs var), and made some little adjustments here and there (like declare vars with my to prevent exceptions), and finally got the filesystem mounted, with all directories listed and files readable. So this is the code I got at last:
#!/usr/bin/env perl
use strict;
use warnings;
use diagnostics;
use Carp;
use Fuse;
use Errno qw(:POSIX); # ENOENT EISDIR etc
use Fcntl qw(:DEFAULT :mode); # S_IFREG S_IFDIR, O_SYNC O_LARGEFILE etc.
use Switch;
my $debug = 0;
my %codecache = ();
my $ctime = time();
my $uid = $>;
my $gid = $) + 0;
my $fs = {
"file1.txt" => "File 1 contents",
"subdir" => {
"subdir-l2" => {},
"file2.txt" => "File 2 contents"
}
};
# some default args
my %args = (
"mountpoint" => "listing",
"debug" => $debug,
"fuse_debug" => 0,
"threaded" => 0,
"/" => $fs
);
# the default subs
my %fs_subs = (
"chmod" => \&fs_not_imp,
"chown" => \&fs_not_imp,
"flush" => \&fs_flush,
"fsync" => \&fs_not_imp,
"getattr" => \&fs_getattr,
"getdir" => \&fs_getdir,
"getxattr" => \&fs_not_imp,
"link" => \&fs_not_imp,
"listxattr" => \&fs_not_imp,
"mkdir" => \&fs_not_imp,
"mknod" => \&fs_not_imp,
"open" => \&fs_open,
"read" => \&fs_read,
"readlink" => \&fs_readlink,
"release" => \&fs_release,
"removexattr" => \&fs_not_imp,
"rmdir" => \&fs_not_imp,
"rename" => \&fs_not_imp,
"setxattr" => \&fs_not_imp,
"statfs" => \&fs_statfs,
"symlink" => \&fs_not_imp,
"truncate" => \&fs_truncate,
"unlink" => \&fs_not_imp,
"utime" => sub{return 0},
"write" => \&fs_write,
);
# except extract these ones back out.
$debug = delete $args{"debug"};
$args{"debug"} = delete( $args{"fuse_debug"} ) || 0;
delete $args{"/"};
# add the functions, if not already defined.
# wrap in debugger if debug is set.
for my $name (keys %fs_subs) {
my $sub = $fs_subs{$name};
# $sub = wrap($sub, $name) if $debug;
$args{$name} ||= $sub;
}
Fuse::main(%args);
sub fetch {
my ($path, #args) = #_;
my $obj = $fs;
for my $elem (split '/', $path) {
next if $elem eq ""; # skip empty // and before first /
$obj = runcode($obj); # if there's anything to run
# the dir we're changing into must be a hash (dir)
return ENOTDIR() unless ref($obj) eq "HASH";
# note that ENOENT and undef are NOT the same thing!
return ENOENT() unless exists $obj->{$elem};
$obj = $obj->{$elem};
}
return runcode($obj, #args);
}
sub runcode {
my ($obj, #args) = #_;
while (ref($obj) eq "CODE") {
my $old = $obj;
if (#args) { # run with these args. don't cache
delete $codecache{$old};
print "running $obj(",quoted(#args),") NO CACHE\n" if $debug;
$obj = saferun($obj, #args);
} elsif (exists $codecache{$obj}) { # found in cache
print "got cached $obj\n" if $debug;
$obj = $codecache{$obj}; # could be undef, or an error, BTW
} else {
print "running $obj() to cache\n" if $debug;
$obj = $codecache{$old} = saferun($obj);
}
if (ref($obj) eq "NOCACHE") {
print "returned a nocache() value - flushing\n" if $debug;
delete $codecache{$old};
$obj = $$obj;
}
print "returning ",ref($obj)," ",
defined($obj) ? $obj : "undef",
"\n" if $debug;
}
return $obj;
}
sub saferun {
my ($sub, #args) = #_;
my $ret = eval { &$sub(#args) };
my $died = $#;
if (ref($died)) {
print "+++ Error $$died\n" if ref($died) eq "ERROR";
return $died;
} elsif ($died) {
print "+++ $died\n";
# stale file handle? moreorless?
return ESTALE();
}
return $ret;
}
sub nocache {
return bless(\ shift, "NOCACHE"); # yup, utter abuse of bless :-)
}
sub dump_open_flags {
my $flags = shift;
printf " flags: 0%o = (", $flags;
for my $bits (
[ O_ACCMODE(), O_RDONLY(), "O_RDONLY" ],
[ O_ACCMODE(), O_WRONLY(), "O_WRONLY" ],
[ O_ACCMODE(), O_RDWR(), "O_RDWR" ],
[ O_APPEND(), O_APPEND(), "|O_APPEND" ],
[ O_NONBLOCK(), O_NONBLOCK(), "|O_NONBLOCK" ],
[ O_SYNC(), O_SYNC(), "|O_SYNC" ],
[ O_DIRECT(), O_DIRECT(), "|O_DIRECT" ],
[ O_LARGEFILE(), O_LARGEFILE(), "|O_LARGEFILE" ],
[ O_NOFOLLOW(), O_NOFOLLOW(), "|O_NOFOLLOW" ],
) {
my ($mask, $flag, $name) = #$bits;
if (($flags & $mask) == $flag) {
$flags -= $flag;
print $name;
}
}
printf "| 0%o !!!", $flags if $flags;
print ")\n";
}
sub accessor {
my $var_ref = shift;
croak "accessor() requires a reference to a scalar var\n"
unless defined($var_ref) && ref($var_ref) eq "SCALAR";
return sub {
my $new = shift;
$$var_ref = $new if defined($new);
return $$var_ref;
}
}
sub fs_not_imp { return -ENOSYS() }
sub fs_flush {
# we're passed a path, but finding my coderef stuff from a path
# is a bit of a 'mare. flush the lot, won't hurt TOO much.
print "Flushing\n" if $debug;
%codecache = ();
return 0;
}
sub easy_getattr {
my ($mode, $size) = #_;
return (
0, 0, # $dev, $ino,
$mode,
1, # $nlink, see fuse.sourceforge.net/wiki/index.php/FAQ
$uid, $gid, # $uid, $gid,
0, # $rdev,
$size, # $size,
$ctime, $ctime, $ctime, # actually $atime, $mtime, $ctime,
1024, 1, # $blksize, $blocks,
);
}
sub fs_getattr {
my $path = shift;
my $obj = fetch($path);
# undef doesn't actually mean "file not found", it could be a coderef
# file-sub which has returned undef.
return easy_getattr(S_IFREG | 0200, 0) unless defined($obj);
switch (ref($obj)) {
case "ERROR" { # this is an error to be returned.
return -$$obj;
}
case "" { # this isn't a ref, it's a real string "file"
return easy_getattr(S_IFREG | 0644, length($obj));
}
# case "CODE" should never happen - already been run by fetch()
case "HASH" { # this is a directory hash
return easy_getattr(S_IFDIR | 0755, 1);
}
case "SCALAR" { # this is a scalar ref. we use these for symlinks.
return easy_getattr(S_IFLNK | 0777, 1);
}
else { # what the hell is this file?!?
print "+++ What on earth is ",ref($obj)," $path ?\n";
return easy_getattr(S_IFREG | 0000, 0);
}
}
}
sub fs_getdir {
my $obj = fetch(shift);
return -$$obj if ref($obj) eq "ERROR"; # THINK this is a good idea.
return -ENOENT() unless ref($obj) eq "HASH";
return (".", "..", sort(keys %$obj), 0);
}
sub fs_open {
# doesn't really need to open, just needs to check.
my $obj = fetch(shift);
my $flags = shift;
dump_open_flags($flags) if $debug;
# if it's undefined, and we're not writing to it, return an error
return -EBADF() unless defined($obj) or ($flags & O_ACCMODE());
switch (ref($obj)) {
case "ERROR" { return -$$obj; }
case "" { return 0 } # this is a real string "file"
case "HASH" { return -EISDIR(); } # this is a directory hash
else { return -ENOSYS(); } # what the hell is this file?!?
}
}
sub fs_read {
my $obj = fetch(shift);
my $size = shift;
my $off = shift;
return -ENOENT() unless defined($obj);
return -$$obj if ref($obj) eq "ERROR";
# any other types of refs are probably bad
return -ENOENT() if ref($obj);
if ($off > length($obj)) {
return -EINVAL();
} elsif ($off == length($obj)) {
return 0; # EOF
}
return substr($obj, $off, $size);
}
sub fs_readlink {
my $obj = fetch(shift);
return -$$obj if ref($obj) eq "ERROR";
return -EINVAL() unless ref($obj) eq "SCALAR";
return $$obj;
}
sub fs_release {
my ($path, $flags) = #_;
dump_open_flags($flags) if $debug;
return 0;
}
sub fs_statfs {
return (
255, # $namelen,
1,1, # $files, $files_free,
1,1, # $blocks, $blocks_avail, # 0,0 seems to hide it from df?
2, # $blocksize,
);
}
sub fs_truncate {
my $obj = fetch(shift, ""); # run anything to set it to ""
return -$$obj if ref($obj) eq "ERROR";
return 0;
}
sub fs_write {
my ($path, $buf, $off) = #_;
my $obj = fetch($path, $buf, $off); # this runs the coderefs!
return -$$obj if ref($obj) eq "ERROR";
return length($buf);
}
Final word: I didn't try to use the module itself (it's not listed in my distro package repository, and I was too lazy (sorry) to install it by cpanm or other way). But I think that if I'll have to just use FUSE with Perl, I'll probably just use Fuse::Simple instead of Fuse, maybe forking it. I'd use plain Fuse only for my academic research, I think...
Hope this helps.
I'm trying to debug a weird warning that is showing up in server logs when a Plack::Request is being parsed. In some cases, a broken UserAgent will send a Content-Length header that looks something like "6375, 6375", which is obviously wrong.
To fix this properly, I need to be able to reproduce the warning. I'd like to include this in a unit test so that I can ensure there are no regressions after the warning is silenced. However, I'm having trouble doing this with Perl. I know this can be done using netcat and socat, but I don't want the unit test to have to rely on other binaries to be installed.
Here is what I've tried:
#!/usr/bin/env perl
use strict;
use warnings;
use JSON::XS qw( encode_json );
use WWW::Mechanize;
my $mech = WWW::Mechanize->new;
$mech->add_handler(
request_prepare => sub {
my ( $req, $ua, $h ) = #_;
$req->headers->header( 'Content-Length' => 9999 );
return;
}
);
my $json = encode_json( { foo => 'bar' } );
$mech->post(
'http://example.com'/url,
'Content-Length' => 999,
Content => $json
);
Output is:
Content-Length header value was wrong, fixed at /opt/perl5.16.3/lib/site_perl/5.16.3/LWP/Protocol/http.pm line 260.
200
That's entirely too helpful for me. :)
If I use HTTP::Request and LWP::UserAgent, it's the same end result.
So, I tried HTTP::Tiny.
#!/usr/bin/env perl
use strict;
use warnings;
use DDP;
use HTTP::Tiny;
use JSON::XS qw( encode_json );
my $http = HTTP::Tiny->new;
my $json = encode_json( { foo => 'bar' } );
my $response = $http->request(
'POST',
'http://example.com'/url',
{ headers => { 'Content-Length' => 999, },
content => $json,
}
);
p $response;
The output is:
{ content => "Content-Length missmatch (got: 13 expected: 999)
",
headers => {
content
-length => 49,
content-type => "text/plain",
},
reason => "Internal Exception",
status => 599,
success => "",
url => "http://example.com'/url",
}
Again, too helpful. At this point, I could use a few suggestions.
Seems like the higher level API's are fixing your error; Here's an example using raw sockets that overcomes this;
#!/usr/bin/env perl
use strict 'vars';
use warnings;
use Socket;
# initialize host and port
my $host = 'www.example.com';
my $port = 80;
# contact the server
open_tcp(F, $host, $port)
or die 'Could not connect to server';
# Send request data
while ( my $request = <DATA> ) {
print F $request;
}
# Get Response
while ( my $response = <F> ) {
print "Response:> $response";
}
close(F);
# TCP Helper
sub open_tcp
{
# get parameters
my ($FS, $dest, $port) = #_;
my $proto = getprotobyname('tcp');
socket($FS, PF_INET, SOCK_STREAM, $proto);
my $sin = sockaddr_in($port,inet_aton($dest));
connect($FS,$sin);
my $old_fh = select($FS);
$| = 1; # don't buffer output
select($old_fh);
}
__DATA__
GET / HTTP/1.1
Host: example.com
Content-Length: 999
-END-
I have the following Perl code that I found on this SO Q&A titled: Perl HTTP server. Specifically this answer. Here's my modified code:
httpserver.pl
#!/usr/bin/perl
use strict;
use warnings;
use CGI qw/ :standard /;
use Data::Dumper;
use HTTP::Daemon;
use HTTP::Response;
use HTTP::Status;
use POSIX qw/ WNOHANG /;
use constant HOSTNAME => qx{hostname};
my %O = (
'listen-host' => '127.0.0.1',
'listen-port' => 8080,
'listen-clients' => 30,
'listen-max-req-per-child' => 100,
);
my $d = HTTP::Daemon->new(
LocalAddr => $O{'listen-host'},
LocalPort => $O{'listen-port'},
Reuse => 1,
) or die "Can't start http listener at $O{'listen-host'}:$O{'listen-port'}";
print "Started HTTP listener at " . $d->url . "\n";
my %chld;
if ($O{'listen-clients'}) {
$SIG{CHLD} = sub {
# checkout finished children
while ((my $kid = waitpid(-1, WNOHANG)) > 0) {
delete $chld{$kid};
}
};
}
while (1) {
if ($O{'listen-clients'}) {
# prefork all at once
for (scalar(keys %chld) .. $O{'listen-clients'} - 1 ) {
my $pid = fork;
if (!defined $pid) { # error
die "Can't fork for http child $_: $!";
}
if ($pid) { # parent
$chld{$pid} = 1;
}
else { # child
$_ = 'DEFAULT' for #SIG{qw/ INT TERM CHLD /};
http_child($d);
exit;
}
}
sleep 1;
}
else {
http_child($d);
}
}
sub http_child {
my $d = shift;
my $i;
my $css = <<CSS;
form { display: inline; }
CSS
while (++$i < $O{'listen-max-req-per-child'}) {
my $c = $d->accept or last;
my $r = $c->get_request(1) or last;
$c->autoflush(1);
print sprintf("[%s] %s %s\n", $c->peerhost, $r->method, $r->uri->as_string);
my %FORM = $r->uri->query_form();
if ($r->uri->path eq '/') {
_http_response($c, { content_type => 'text/html' },
start_html(
-title => HOSTNAME,
-encoding => 'utf-8',
-style => { -code => $css },
),
p('Here are all input parameters:'),
pre(Data::Dumper->Dump([\%FORM],['FORM'])),
(map { p(a({ href => $_->[0] }, $_->[1])) }
['/', 'Home'],
['/ping', 'Ping the simple text/plain content'],
['/error', 'Sample error page'],
['/other', 'Sample not found page'],
),
end_html(),
)
}
elsif ($r->uri->path eq '/ping') {
_http_response($c, { content_type => 'text/plain' }, 1);
}
elsif ($r->uri->path eq '/error') {a
my $error = 'AAAAAAAAA! My server error!';
_http_error($c, RC_INTERNAL_SERVER_ERROR, $error);
die $error;
}
elsif ($r->method eq 'POST' and $r->uri->path eq '/formdata') {
#_http_response($c, { content_type => 'text/plain' }, 1);
print "--> begin form data <--\n";
_http_response($c, { content_type => 'text/html' },
start_html(
-title => HOSTNAME,
-encoding => 'utf-8',
-style => { -code => $css },
),
p('Here are all the input parameters:'),
pre(Data::Dumper->Dump([\%FORM],['FORM'])),
end_html(),
);
print Data::Dumper->Dump([$r], [qw(r)]);
print "--> end form data <--\n";
}
else {
_http_error($c, RC_NOT_FOUND);
}
$c->close();
undef $c;
}
}
sub _http_error {
my ($c, $code, $msg) = #_;
$c->send_error($code, $msg);
}
sub _http_response {
my $c = shift;
my $options = shift;
$c->send_response(
HTTP::Response->new(
RC_OK,
undef,
[
'Content-Type' => $options->{content_type},
'Cache-Control' => 'no-store, no-cache, must-revalidate, post-check=0, pre-check=0',
'Pragma' => 'no-cache',
'Expires' => 'Thu, 01 Dec 1994 16:00:00 GMT',
],
join("\n", #_),
)
);
}
curl command
I'm using this curl command to connect to the server.
$ curl -X POST -H "Content-Type: multipart/form-data; \
boundary=----------------------------4ebf00fbcf09" \
--data-binary #test.txt \
http://localhost:8080/formdata?arg1=blah1\&arg2=blah2
test.txt data file
Along with this test file.
$ cat test.txt
This is some test text in a file.
Which returns the following when I run it:
<!DOCTYPE html
PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" lang="en-US" xml:lang="en-US">
<head>
<title>greeneggs.bubba.net
</title>
<style type="text/css">
<!--/* <![CDATA[ */
form { display: inline; }
/* ]]> */-->
</style>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
</head>
<body>
<p>Here are all the input parameters:</p>
<pre>$FORM = {
'arg2' => 'blah2',
'arg1' => 'blah1'
};
</pre>
</body>
</html>
My Question
How do I process the data from the multi-data form on the server side? I thought the data would be accessible through the request ($r) but when I use Data::Dumper to analyze it I don't see anything resembling the data.
Output from the http server after connecting to it via curl command:
[127.0.0.1] POST /formdata?arg1=blah1&arg2=blah2
--> begin form data <--
$r = bless( {
'_protocol' => 'HTTP/1.1',
'_content' => '',
'_uri' => bless( do{\(my $o = '/formdata?arg1=blah1&arg2=blah2')}, 'URI::http' ),
'_headers' => bless( {
'user-agent' => 'curl/7.29.0',
'content-type' => 'multipart/form-data; boundary=----------------------------4ebf00fbcf09',
'accept' => '*/*',
'content-length' => '34',
'host' => 'localhost:8080'
}, 'HTTP::Headers' ),
'_method' => 'POST'
}, 'HTTP::Request' );
--> end form data <--
What am I missing?
$c->get_request( $headers_only )
The get_request() method will normally not return until the whole request has been received from the client. This might not be what you want if the request is an upload of a large file (and with chunked transfer encoding HTTP can even support infinite request messages - uploading live audio for instance). If you pass a TRUE value as the $headers_only argument, then get_request() will return immediately after parsing the request headers and you are responsible for reading the rest of the request content. If you are going to call $c->get_request again on the same connection you better read the correct number of bytes.
try changing (inside httpsserver.pl above),
my $r = $c->get_request(1) or last;
to
my $r = $c->get_request() or last;