* fix get_pdf again

diff --git a/get_pdf b/get_pdf
index 745ea9aedec558b4ddea088c70784c2013f5dd15..74f62dc6998e1f6ec494b4411eac95f9495320d8 100755 (executable)
--- a/get_pdf
+++ b/get_pdf
@@ -120,11 +120,13 @@ if ($options{pmid}) {
        eval {
            $m->get($url) or die "Unable to get $url";
            my $orig_mech = $m->clone();
+           use Data::Dumper;
            my @possible_links = $m->find_all_links(text_regex => qr/to\s*read/i);
            # try to find the other links
            push @possible_links,
                grep {my $attr = $_->attrs(); exists $attr->{title} and $attr->{title} =~ qr/Full\s*Text/i}
                    $m->links();
+           print STDERR map {"article link: ".$_->url_abs()."\n"} @possible_links if $DEBUG;
            die "No links" unless @possible_links;
            do {
                $m = $orig_mech;
@@ -179,7 +181,10 @@ sub find_pdf_link {
     }
     my @possible_links;
     # this is to prioritize the real link at science direct
-    push @possible_links, grep { $_->url_abs() !~ /_orig=article/} $m->find_all_links(text_regex => qr/PDF/i);
+    push @possible_links, grep {my $temp = $_->attrs();
+                               exists $temp->{title} and $temp->{title} =~ qr/Download\s*PDF/i}
+       $m->find_all_links(text_regex => qr/PDF/i);
+    push @possible_links, grep { $_->url_abs() !~ /_orig(?:in)?=article/} $m->find_all_links(text_regex => qr/PDF/i);
     push @possible_links, $m->find_all_links(tag_regex => qr/meta/,
                                             url_regex  => qr/(reprint|\.pdf)/i,
                                            );
@@ -187,16 +192,19 @@ sub find_pdf_link {
     push @possible_links,$m->find_all_links(text_regex => qr/manual\s*download/i);
     print STDERR $m->uri() if $DEBUG;
     print STDERR $m->content() if $DEBUG;
-    print STDERR map{$_->url_abs(),qq(\n)} @possible_links if $DEBUG;
+    print STDERR map{"possible pdf link: ".$_->url_abs().qq(\n)} @possible_links if $DEBUG;
     if (not @possible_links and $DEBUG) {
        print STDERR $m->content();
     }
     my $best_guess = $possible_links[0] if @possible_links;
     for my $link (@possible_links) {
+       print STDERR "trying ".$link->url_abs()."..." if $DEBUG;
        my $r = $m->get($link->url_abs());
        if ($r->header('Content-Type') =~ /pdf/) {
+           print STDERR "success\n" if $DEBUG;
            return $m;
        }
+       print STDERR "failure; content type ".$r->header('Content-Type')."\n" if $DEBUG;
     }
     my @sub_frames = $m->find_all_links(tag_regex=>qr/^i?frame$/);
     for my $frame (@sub_frames) {