From: Don Armstrong <don@donarmstrong.com>
Date: Wed, 30 Jun 2010 02:16:44 +0000 (+0000)
Subject:  * handle multiple full text sources
X-Git-Url: https://git.donarmstrong.com/?a=commitdiff_plain;h=1d30f0cb452e9c5b4026be5d074aee74cc3efc79;p=bin.git

 * handle multiple full text sources
 * properly deal with iframes
---

diff --git a/get_pdf b/get_pdf
index 0aadc60..3768e52 100755
--- a/get_pdf
+++ b/get_pdf
@@ -119,17 +119,34 @@ if ($options{pmid}) {
 	$url = "http://${url}";
 	eval {
 	    $m->get($url) or die "Unable to get $url";
-	    $m->follow_link(text_regex => qr/to\s*read/i) or
-		die "Unable to follow link";
-	    # try to find pdf link
-	    my $pdf_m = find_pdf_link($m) or
-		die "Unable to find pdf";
-	    my $fh = IO::File->new($pmid.'.pdf','w') or
-		die "Unable to open ${pmid}.pdf for writing: $!";
-	    print {$fh} $pdf_m->content or
-		die "Unable to write to ${pmid}.pdf: $!";
-	    close $fh or
-		die "Unable to close ${pmid}.pdf filehandle: $!";
+	    my $orig_mech = $m->clone();
+	    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();
+	    die "No links" unless @possible_links;
+	    do {
+		$m = $orig_mech;
+		eval {
+		    print "trying ".$possible_links[0]->url()."\n" if $DEBUG;
+		    $m->get($possible_links[0]->url()) or
+			die "Unable to follow link";
+		    # try to find pdf link
+		    my $pdf_m = find_pdf_link($m) or
+			die "Unable to find pdf";
+		    my $fh = IO::File->new($pmid.'.pdf','w') or
+			die "Unable to open ${pmid}.pdf for writing: $!";
+		    print {$fh} $pdf_m->content or
+			die "Unable to write to ${pmid}.pdf: $!";
+		    close $fh or
+			die "Unable to close ${pmid}.pdf filehandle: $!";
+		};
+		shift @possible_links;
+	    } while ($@ and @possible_links);
+	    if ($@) {
+		die "$@";
+	    }
 	};
 	if ($@) {
 	    print STDERR "$@\n" if $DEBUG;
@@ -181,7 +198,10 @@ sub find_pdf_link {
     }
     my @sub_frames = $m->find_all_links(tag_regex=>qr/^i?frame$/);
     for my $frame (@sub_frames) {
-	$m->get($frame->url_abs());
+	my $r = $m->get($frame->url_abs());
+	if ($r->header('Content-Type') =~ /pdf/) {
+	    return $m;
+	}
 	my $pdf_m = find_pdf_link($m,
 				  0,
 				  $call+1,