X-Git-Url: https://sipb.mit.edu/gitweb.cgi/ikiwiki.git/blobdiff_plain/2a6e353c205a6c2c8b8e2eaf85fe9c585c1af0cd..fafb2edaa7aeb1293e716fa96f087cb713f4a70a:/IkiWiki/Plugin/aggregate.pm

diff --git a/IkiWiki/Plugin/aggregate.pm b/IkiWiki/Plugin/aggregate.pm
index ea27fc199..71368e254 100644
--- a/IkiWiki/Plugin/aggregate.pm
+++ b/IkiWiki/Plugin/aggregate.pm
@@ -17,7 +17,7 @@ my %guids;
 sub import { #{{{
 	hook(type => "getopt", id => "aggregate", call => \&getopt);
 	hook(type => "checkconfig", id => "aggregate", call => \&checkconfig);
-	hook(type => "filter", id => "aggregate", call => \&filter);
+	hook(type => "needsbuild", id => "aggregate", call => \&needsbuild);
 	hook(type => "preprocess", id => "aggregate", call => \&preprocess);
         hook(type => "delete", id => "aggregate", call => \&delete);
 	hook(type => "savestate", id => "aggregate", call => \&savestate);
@@ -37,28 +37,41 @@ sub checkconfig () { #{{{
 			debug("wiki is locked by another process, not aggregating");
 			exit 1;
 		}
-	
-		loadstate();
-		IkiWiki::loadindex();
-		aggregate();
-		expire();
-		savestate();
-		clearstate();
-
+		
+		# Fork a child process to handle the aggregation.
+		# The parent process will then handle building the result.
+		# This avoids messy code to clear state accumulated while
+		# aggregating.
+		defined(my $pid = fork) or error("Can't fork: $!");
+		if (! $pid) {
+			loadstate();
+			IkiWiki::loadindex();
+			aggregate();
+			expire();
+			savestate();
+			exit 0;
+		}
+		waitpid($pid,0);
+		if ($?) {
+			error "aggregation failed with code $?";
+		}
+		
 		IkiWiki::unlockwiki();
 	}
 } #}}}
 
-sub filter (@) { #{{{
-	my %params=@_;
-	my $page=$params{page};
-
+sub needsbuild (@) { #{{{
+	my $needsbuild=shift;
+	
 	loadstate(); # if not already loaded
-	# Mark all feeds originating on this page as removable;
-	# preprocess will unmark those that still exist.
-	remove_feeds($page);
 
-	return $params{content};
+	foreach my $feed (values %feeds) {
+		if (grep { $_ eq $pagesources{$feed->{sourcepage}} } @$needsbuild) {
+			# Mark all feeds originating on this page as removable;
+			# preprocess will unmark those that still exist.
+			remove_feeds($feed->{sourcepage});
+		}
+	}
 } # }}}
 
 sub preprocess (@) { #{{{
@@ -127,6 +140,7 @@ sub delete (@) { #{{{
 my $state_loaded=0;
 sub loadstate () { #{{{
 	return if $state_loaded;
+	$state_loaded=1;
 	if (-e "$config{wikistatedir}/aggregate") {
 		open(IN, "$config{wikistatedir}/aggregate") ||
 			die "$config{wikistatedir}/aggregate: $!";
@@ -157,17 +171,14 @@ sub loadstate () { #{{{
 		}
 
 		close IN;
-		
-		$state_loaded=1;
 	}
 } #}}}
 
 sub savestate () { #{{{
+	return unless $state_loaded;
 	eval q{use HTML::Entities};
 	error($@) if $@;
 	my $newfile="$config{wikistatedir}/aggregate.new";
-	# TODO: This cleanup function could use improvement. Any newly
-	# aggregated files are left behind unrecorded, and should be deleted.
 	my $cleanup = sub { unlink($newfile) };
 	open (OUT, ">$newfile") || error("open $newfile: $!", $cleanup);
 	foreach my $data (values %feeds, values %guids) {
@@ -180,7 +191,8 @@ sub savestate () { #{{{
 				}
 			}
 			else {
-				unlink pagefile($data->{page});
+				unlink pagefile($data->{page})
+					if exists $data->{page};
 			}
 			next;
 		}
@@ -210,12 +222,6 @@ sub savestate () { #{{{
 		error("rename $newfile: $!", $cleanup);
 } #}}}
 
-sub clearstate () { #{{{
-	%feeds=();
-	%guids=();
-	$state_loaded=0;
-} #}}}
-
 sub expire () { #{{{
 	foreach my $feed (values %feeds) {
 		next unless $feed->{expireage} || $feed->{expirecount};
@@ -228,7 +234,7 @@ sub expire () { #{{{
 				my $days_old = (time - $IkiWiki::pagectime{$item->{page}}) / 60 / 60 / 24;
 				if ($days_old > $feed->{expireage}) {
 					debug(sprintf(gettext("expiring %s (%s days old)"),
-						$item->{page}, $days_old));
+						$item->{page}, int($days_old)));
 					$item->{expired}=1;
 				}
 			}
@@ -300,6 +306,13 @@ sub aggregate () { #{{{
 			$content=Encode::decode_utf8($content);
 			$f=eval{XML::Feed->parse(\$content)};
 		}
+		if ($@) {
+			# Another possibility is badly escaped entities.
+			$feed->{message}.=" ".sprintf(gettext("(feed entities escaped)"));
+			$content=~s/\&(?!amp)(\w+);/&$1;/g;
+			$content=Encode::decode_utf8($content);
+			$f=eval{XML::Feed->parse(\$content)};
+		}
 		if ($@) {
 			$feed->{message}=gettext("feed crashed XML::Feed!")." ($@)";
 			$feed->{error}=1;
@@ -316,6 +329,7 @@ sub aggregate () { #{{{
 		foreach my $entry ($f->entries) {
 			add_page(
 				feed => $feed,
+				copyright => $f->copyright,
 				title => defined $entry->title ? decode_entities($entry->title) : "untitled",
 				link => $entry->link,
 				content => defined $entry->content->body ? $entry->content->body : "",
@@ -395,6 +409,8 @@ sub add_page (@) { #{{{
 	$template->param(content => htmlescape(htmlabs($params{content}, $feed->{feedurl})));
 	$template->param(name => $feed->{name});
 	$template->param(url => $feed->{url});
+	$template->param(copyright => $params{copyright})
+		if defined $params{copyright} && length $params{copyright};
 	$template->param(permalink => urlabs($params{link}, $feed->{feedurl}))
 		if defined $params{link};
 	if (ref $feed->{tags}) {
@@ -405,7 +421,8 @@ sub add_page (@) { #{{{
 
 	# Set the mtime, this lets the build process get the right creation
 	# time on record for the new page.
-	utime $mtime, $mtime, pagefile($guid->{page}) if defined $mtime;
+	utime $mtime, $mtime, pagefile($guid->{page})
+		if defined $mtime && $mtime <= time;
 } #}}}
 
 sub htmlescape ($) { #{{{