--- CVSROOT/cvs2rss.pl	2005/10/11 19:45:18	1.1
+++ CVSROOT/cvs2rss.pl	2005/10/17 19:50:57	1.7
@@ -13,23 +13,18 @@
 #
 # 	AddType application/rss+xml .rss
 #
-# Important note: 
-# If you have html files (or files with html in them) in your CVS Repository, look in your RSS.pm for:
-#
-#	$text =~ s/</&lt;/g;
-#
-# I suggest commenting out this line because it will screw up this script (I think its a bug in the parser)
-# If you do not, when this script uses XML:RSS's parsefile(), it will mess up any <descriptions> that 
-# contain HTML.
-#
 # P.S. The "other" cvs2rss.pl isn't really cvs2rss, it is more like cvs2xml, boo!
 # P.P.S. Parts of this are stolen from commit2rss.rb and the XML::RSS examples.
 #
 # Version 1.0 - 08.24.04 RSS Compliant, no cvs diff capability yet..
 # Version 1.1 - 08.31.04 Adding CVS Diff capability
+# Edit by Guy Albertelli, -- corrected the packaging of Entities
+#                          - commits now <pre>ed
+#                          - removed html rants
 #
 
 use strict;
+use HTML::Entities();
 use XML::RSS;
 use POSIX;
 
@@ -37,10 +32,10 @@ use POSIX;
 my $rssFeed ="/home/loninst/public_html/loncapa.rss";
 my $emailDomain = "loncapa.org";
 my $channelTitle = "Lon-CAPA RSS Feed";
-my $channelLink = "http://install.loncapa.org/cvs.rss";
+my $channelLink = "http://install.loncapa.org/loncapa.rss";
 my $numEntries = 200;
 
-# Set this to 1 to enable cvs rdiff (is pretty broken if you are diffing html content)
+# Set this to 1 to enable cvs rdiff
 my $cvsDiff = 1;
 
 # Leave everything else alone
@@ -51,18 +46,21 @@ my $description;
 my @title=split(",",$ARGV[0]);
 
 my $rss = new XML::RSS(version => '2.0');
+
+# If rssFeed exists, parse it
+if (-r "$rssFeed") {
+	$rss->parsefile($rssFeed);
+}
+
 $rss->channel(
 	title=> $channelTitle,
 	link => $channelLink,
 	language => 'en',
 	description => $channelTitle,
-	copyright => '(c) 2005 MSU Board of Trustees.'
+	copyright => '(c) 2005 MSU Board of Trustees.',
+	pubDate => $pubDate 
 );
 
-# If rssFeed exists, parse it
-if (-r "$rssFeed") {
-	$rss->parsefile($rssFeed);
-}
 
 # Limit entries in the feed to $numEntries
 pop(@{$rss->{'items'}}) while (@{$rss->{'items'}} >= $numEntries);
@@ -75,9 +73,10 @@ $title[0] =~s/ /\//;
 while (<STDIN>) {
 	chomp($_);
 	if ($_=~/^[A-Z].*:\s*$/) {
-		$_ = "<br /><b>" . $_ . "</b><br />";
+		$_ = "<br /><b>" . &HTML::Entities::encode($_,'<>&"') . "</b><br />";
 	}
 	else {
+		$_ = &HTML::Entities::encode($_,'<>&"');
 		$_ .= "<br />";
 	}
 	$description .= $_;
@@ -87,17 +86,17 @@ if ($cvsDiff == 1) {
 	# If the old version of the file is not NONE (if it isn't a new file), and if it is a .txt or .java file (that has no html)
 	# This will rdiff it against the previous version, and include that diff in the rss feed
 
-	if (($title[1] != "NONE") && ($title[0]=~/(.*).(txt|java)$/)){
+	if (($title[1] != "NONE") && ($title[0]=~/(.*).(pm)$/)){
 		my $tmpFile = "/tmp/diff.$$";
-		my $cmdLine = "cvs -n rdiff -kk -r " . $title[1] .  " -r " . $title[2] . " " . $title[0] . ">" . $tmpFile;
+		my $cmdLine = "cvs -n rdiff -u -kk -r " . $title[1] .  " -r " . $title[2] . " " . $title[0] . ">" . $tmpFile;
 		system($cmdLine);
 
-		$description .= "<br /><b>Differences:</b><br />";
+		$description .= "<br /><b>Differences:</b><br /><pre>";
 		open CVSDIFF, "<" . $tmpFile;
 		foreach my $line (<CVSDIFF>) {
-			chomp($line);
-			$description .= $line . "<br />";
+			$description .= &HTML::Entities::encode($line,'<>&"');
 		}
+		$description .= "</pre>";
 		unlink($tmpFile);   
 	} 
 }
@@ -105,16 +104,14 @@ if ($cvsDiff == 1) {
 $rss->add_item(
 	title => "/" . $title[0] . " - " . $title[1] . "/" . $title[2],
 	author => $author,
-	pubDate => $pubDate, 
 	description=> $description,
-	mode => 'insert'
+	mode => 'insert',
+	pubDate => $pubDate,
+	link => 'http://install.loncapa.org/cgi-bin/cvsweb.cgi/'.$title[0].'.diff?r1='.$title[1].';r2='.$title[2].';f=h'
 );		
 
-# XML::RSS doesn't like HTML in the XML so Escape the description body with <![CDATA[ ]]>
-# This is messy and stupid, but both XML::RSS and XML are sort of dumb, and I can't figure out a way around it.
-# Certain RSS readers will try and render the HTML regardless of if it is wrapped in a CDATA tag or not so um, deal.
 foreach my $element (@{$rss->{'items'}}) {
-	$element->{'description'} = "<![CDATA[" . $element->{'description'} . "]]>";
+	$element->{'description'} = &HTML::Entities::encode($element->{'description'},'<>&"'); 
 }
 
 $rss->save($rssFeed);