update get harvester results for new site
authorDon Armstrong <don@donarmstrong.com>
Tue, 28 Aug 2007 01:01:57 +0000 (01:01 +0000)
committerDon Armstrong <don@donarmstrong.com>
Tue, 28 Aug 2007 01:01:57 +0000 (01:01 +0000)
git-svn-id: file:///srv/svn/function2gene/trunk@9 a0738b58-4706-0410-8799-fb830574a030

bin/get_harvester_results

index c6b4f0b..97734e9 100755 (executable)
@@ -10,6 +10,7 @@
 # $Id: ss,v 1.1 2004/06/29 05:26:35 don Exp $
 
 
+use threads;
 use warnings;
 use strict;
 
@@ -68,8 +69,9 @@ BEGIN{
 }
 
 use IO::File;
-use URI::ParamMunge;
-use LWP::UserAgent;
+use URI;
+use WWW::Mechanize;
+use Thread::Queue;
 
 # XXX parse config file
 
@@ -81,10 +83,10 @@ my %options = (debug    => 0,
               dir      => '.',
               name     => '${search}_results_harvester',
               terms    => '-',
-              harvester_site => 'http://www-db.embl.de',
-              harvester_search_url  => '/jss/servlet/de.embl.bk.htmlfind.HarvesterPageSearchOutput?search=GOLGI&chipsetID=-1&maxHits=10000&submit=search',
-             );
+              orgn     => 'human',
+              harvester_site => 'http://harvester.fzk.de',
 
+             );
 GetOptions(\%options,'format|f=s','database|b=s','name|n=s',
           'terms|t=s','dir|D=s','debug|d+','help|h|?','man|m');
 
@@ -97,6 +99,9 @@ if (not -d $options{dir}) {
      die "$options{dir} does not exist or is not a directory";
 }
 
+
+$options{harvester_search_url}  = '/cgi-bin/'.$options{orgn}.'/search.cgi?zoom_query=golgi&zoom_per_page=100&zoom_and=1&zoom_sort=0';
+
 #open search terms file
 my $terms;
 if ($options{terms} eq '-') {
@@ -106,34 +111,66 @@ else {
      $terms = new IO::File $options{terms}, 'r' or die "Unable to open file $options{terms}: $!";
 }
 
-my $ua = new LWP::UserAgent(agent=>"DA_get_harvester_results/$REVISION");
-
 #For every term
+my @threads;
 while (<$terms>) {
      # Get uids to retrieve
      chomp;
      my $search = $_;
-     my $url = uri_param_munge($options{harvester_site}.$options{harvester_search_url},
-                              {search => $search,
-                              },
-                             );
-     my $request = HTTP::Request->new('GET', $url);
-     my $response = $ua->request($request);
-     $response = $response->content;
-     my @result_urls = $response =~ m#<a\s+href=(http://harvester.embl.de/harvester/[^\/]+/[^.]+.htm)\s*>#g;
-
+     my $uri = URI->new($options{harvester_site}.$options{harvester_search_url});
+     $uri->query_form(zoom_query =>[],
+                    );
+     $uri->query_form(zoom_query => $search,
+                    );
+     my $url = $uri->as_string;
+     my $queue = Thread::Queue->new();
      my $dir_name = eval qq("$options{name}") or die $@;
      mkdir("$options{dir}/$dir_name") or die "Unable to make directory $options{dir}/$dir_name $!";
-     # Get XML file
-     my @current_urls;
-     while (@current_urls = splice(@result_urls,0,30)) {
-         system(q(wget),'-nd','-nH','-w','2','--random-wait','-P',qq($options{dir}/$dir_name),@current_urls) == 0 or warn "$!";
-     }
+     my $wget_thread = threads->new(\&get_url,"$options{dir}/$dir_name",$queue);
+     push @threads,$wget_thread;
+
+     my $mech = WWW::Mechanize->new(agent => "DA_get_harvester_results/$REVISION");
+
+     #HTTP::Request->new('GET', $url);
+     $mech->get($url);
+     my $next_link;
+     do {
+         my @links = $mech->links;
+         $next_link = undef;
+         for my $link (@links) {
+              if ($link->text() =~ /Next /) {
+                   $next_link = $link;
+              }
+              elsif ($link->url =~ m#http://harvester.fzk.de/harvester/human/[^\/]+/[^.]+.htm#) {
+                   $queue->enqueue($link->url());
+              }
+         }
+         $mech->follow_link(url=>$next_link->url) if defined $next_link;
+     } while ($next_link);
+     $queue->enqueue(undef);
+}
+for my $thread (@threads) {
+     $thread->join;
 }
 
+sub get_url{
+     my ($dir,$queue) = @_;
 
 
-
-
+     my @current_urls;
+     while (my $url = $queue->dequeue) {
+         push @current_urls,$url;
+         if (@current_urls >= 30) {
+              wget_urls($dir,@current_urls);
+              @current_urls = ();
+         }
+     }
+     wget_urls($dir,@current_urls) if @current_urls;
+}
+sub wget_urls{
+     my ($dir,@urls) = @_;
+     return unless @urls;
+     system(q(wget),'-nd','-nH','-w','2','--random-wait','-P',$dir,@urls) == 0 or warn "$!";
+}
 
 __END__