#!/usr/bin/perl use CGI qw(:standard); use LWP::UserAgent; use URI::Escape; use HTTP::Request; use HTTP::Response; use HTTP::Cookies; my $pagequery = param('link'); my $sessionid = param('phpsessid'); #$url_to_fetch = "http://www.pangaea.de/PangaVista?query=censor"; $url_to_fetch = "http://www.pangaea.de/search"; if($pagequery =~ /similar/){ $url_to_fetch = "http://www.pangaea.de/PangaVista".$pagequery."&phpsessid=".$sessionid; } $new_url_on_your_site = "http://www.censor.name/pagev2/cgi-bin/showdata.pl"; #$first_text_delimmiter = 'function showMap() {'; # wouldn't it be nice if it was something like ''; $first_text_delimmiter = ''; $last_text_delimiter = ''; #$last_text_delimiter = '? $last_text_delimiter_2 = ''; # Another option in case some pages are different %REPLACEMENTS = ( "/PICS/PanLogo.gif", "../../../uploads/pics/Logo_web_01.jpg", "/pangaea.css", "pangea.css", "", "", "/pvmap/", "http://www.pangaea.de/pvmap/", "/help/help.php/PangaVista/index.html", "http://www.pangaea.de/help/help.php/PangaVista/index.html", "/javaclient/windrose.gif", "http://www.pangaea.de/javaclient/windrose.gif", "form action=\"/search\"", "form action=\"http://www.pangaea.de/search\"", "http://doi.pangaea.de/", "showdatadetail.pl?dataurl=", "/shared/map/preparemap.php?", "http://www.pangaea.de/shared/map/preparemap.php" );# NO COMMA AFTER THE LAST ONE. print "Content-type: text/html\n\n"; # set up print type for browser print " Censor data search
\"\"
Mark datapoints on map

Please visit the official Pangaea site for advanced search! "; &parse_form; &process_the_url; &fetch_it; &fix_links; &replacements; print $content; print " "; exit; #------------------------- sub parse_form{ read(STDIN, $buffer, $ENV{'CONTENT_LENGTH'}); @input_chunks = split(/&/, $buffer); foreach $chunk (@input_chunks) { ($name, $value) = split(/=/, $chunk); $name =~ s/%([a-fA-F0-9][a-fA-F0-9])/pack("C", hex($1))/eg; $name =~ tr/+/ /; $value =~ s/%([a-fA-F0-9][a-fA-F0-9])/pack("C", hex($1))/eg; $value =~ tr/+/ /; $FROMWEB{$name} = $value; } # end of foreach @input_chunks = split(/&/, $ENV{'QUERY_STRING'}); foreach $chunk (@input_chunks) { ($name, $value) = split(/=/, $chunk); $name =~ s/%([a-fA-F0-9][a-fA-F0-9])/pack("C", hex($1))/eg; $name =~ tr/+/ /; $value =~ s/%([a-fA-F0-9][a-fA-F0-9])/pack("C", hex($1))/eg; $value =~ tr/+/ /; $FROMWEB{$name} = $value; } # end of foreach $chunk } #------------------------- sub process_the_url{ if ($FROMWEB{'link'}){ $processed_fetch_url = $url_to_fetch."$FROMWEB{'link'}?".$ENV{'QUERY_STRING'}; $url_to_fetch = $url_to_fetch."$FROMWEB{'link'}?".$ENV{'QUERY_STRING'}; } else{ $processed_fetch_url = $url_to_fetch."?".$ENV{'QUERY_STRING'}; } } #------------------------- sub choose_random_agent{ my @agent_options = ( "Mozilla/4.76 [en] (Windows NT 5.0; U)", "Mozilla/4.0 (compatible; MSIE 5.12; Mac_PowerPC)" ); # now pick one my $random_index = rand($#agent_options); return ("$agent_options[$random_index]¡"); } #------------------------- sub fetch_it{ my $ua = new LWP::UserAgent; $ua->cookie_jar(HTTP::Cookies->new(file=>'cookies.txt',autosave=>1,ignore_discard=>1)); $ua->agent("&choose_random_agent"); my $req = HTTP::Request->new(GET => "$url_to_fetch"); $response = $ua->request($req); $content = $response->content(); $line_break = "\n"; @content = split(/$line_break/, $content); $content = join(" ",@content); $content =~ s/ / /sg; chomp $content; $start_location = index($content,$first_text_delimmiter); $end_location = index($content,$last_text_delimiter); $end_location = index($content,$last_text_delimiter_2) if ($end_location < 1); $length = $end_location - $start_location; $article_only = substr($content, $start_location, $length); $content = $article_only; } #------------------------- sub fix_links{ # $content =~ s/href=\"$url_to_fetch/href=\"$new_url_on_your_site?link=/sgi; $content =~ s/href=\"\/search?/href=\"$new_url_on_your_site\?link=/sgi; # $content =~ s/$new_url_on_your_site\?link=http\:\/\//http\:\/\//sgi; # $content =~ s/$new_url_on_your_site\?link=mailto/mailto/sgi; # $content =~ s/\&\?/\&/sgi; } #------------------------- sub replacements{ foreach my $key (keys(%REPLACEMENTS)) { $content =~ s/$key/$REPLACEMENTS{$key}/sgi; } # print $content."\n\r"; }