#!/usr/bin/perl
use CGI qw(:standard);
use HTML::PullParser;
use Data::Dumper;
my $url;
if ($ENV{'PATH_INFO'} eq '') {
$url = param("proxgo");
if ($url ne '') {
$url =~ s|^http://||i;
$url =~ s|^/||;
print "Location: $ENV{SCRIPT_NAME}/$url\n\n";
exit 0;
}
} else {
$url = "$ENV{'PATH_INFO'}?$ENV{'QUERY_STRING'}";
}
if ($url eq '') {
print header("text/html");
print <
Phone Proxy
EOF
;
exit 0;
}
$url =~ s/\?$//;
$url =~ s|^/([^/])|$1|;
$url = "http://$url" unless $url =~ m|http://|i;
$url =~ s/\"//g;
my $base = $url;
$base =~ s|^http://||i;
$base =~ s|/[^/]*$||;
$base = "$base/" unless substr($base, -1, 1) eq '/';
my $data = qx|wget -U "$ENV{'HTTP_USER_AGENT'}" -O - "$url"|;
print STDERR "$0: Got " . length($data) . " bytes for $url\n";
print header("text/html");
print "\n";
print "\n";
#print "$0: Got " . length($data) . " bytes for $url\n";
my $p = HTML::PullParser->new( doc => $data,
start => 'event,tagname,attr,text',
end => 'event,tagname',
text => '@{dtext}',
unbroken_text => 1,
ignore_elements => ['script', 'iframe', 'style', 'noframes']);
my $token;
my $output;
while (defined($token = $p->get_token())) {
if (my @token = @{$token}) { # text is a scalar, tags are an array ref
if (my $start = shift(@token)) {
if ($start eq 'start') {
my $tag = shift(@token);
if ($tag eq 'a') { # handle links
my $attr = shift(@token);
my $linktext;
while (my $text = $p->get_token()) {
if (my @t = @{$text}) {
last if $t[1] eq 'a'; # scan until close tag (if we're nesting a tags we have a problem)
} else {
$linktext .= $text;
}
}
my $href = $attr->{href};
if ($href !~ /^http:/i) {
$href = "$base$href";
}
$href =~ s|^http://||;
$href = "$ENV{'SCRIPT_NAME'}/$href";
$linktext =~ s/(\S{20})/$1 /g;
$output .= qq|$linktext\n|;
} elsif ($tag =~ /^tr/) { # table row
$output .= "
\n";
} elsif ($tag =~ /^t[hd]/) { # table cell
$output .= "|\n";
} elsif ($tag =~ /(^form$)/) { # handle forms specially (action specifically)
my $action = $token->[2]->{action};
my $text = $token->[3];
if ($action !~ /^http:/) {
$action = "$base$action";
}
$action =~ s|^http://||;
$text =~ s|action=\".+?\"|action=\"$ENV{'SCRIPT_NAME'}/$action\"|;
$text =~ s/method\s*=\s*"?post"?/method=GET/i;
$output .= $text;
} elsif ($tag =~ /(^h.$)|(^b$)|(^i$)|(^br$)|(^p$)|(^input$)|(^textarea$)|(^select$)|(^.l$)|(^li$)|(^title$)|(^option$)/i) {
$output .= $token->[3]; # print raw text of tag
} elsif ($tag =~ /^html$/) { # do simple html tag
$output .= "";
} elsif ($tag =~ /^body$/) { # do simple body tag
$output .= "";
} elsif ($tag =~ /^frame/) {
my $frameurl = fixurl($token->[2]->{src});
if ($tag eq 'frameset') {
$output .= "";
} else {
$output .= qq|Frame: $token->[2]->{name}
\n|;
}
}
} else {
my $tag = shift(@token);
if ($tag =~ /(^h.$)|(^b$)|(^i$)|(^form$)|(^textarea$)|(^select$)|(^.l$)|(^li$)|(^title$)|(^option$)|(^html$)|(^body$)/) {
$output .= "";
}
}
}
} else { # nope, it's just text
my $text = $token;
$text =~ s/(\S{20})/$1 /g;
$output .= $text;
}
}
$output =~ s/(
\s*|\s*)+/
/ig;
$output =~ s/(\s*
\s*)+/
/ig;
$output =~ s/
\s*
/
/ig;
$output =~ s|()\s*
|$1|ig;
$output =~ s/\|(
|\s+)\|/\|/ig;
$output =~ s/\n\s*/\n/g;
$output =~ s/\n+/\n/g;
$output =~ s/
\n*\|/
/gi;
$output =~ s/\s+/ /g;
print $output;
sub fixurl {
my $url = shift;
if ($url !~ /^http:/i) {
$url = "$base$url";
}
$url =~ s|^http://||;
$url = "$ENV{'SCRIPT_NAME'}/$url";
return $url;
$output .= "\n";
}