X-Git-Url: https://sipb.mit.edu/gitweb.cgi/ikiwiki.git/blobdiff_plain/68d7fbfabd3d1b69cf6b9efbfdba9a0f032b977a..38077033c11977a0c8f5a10103b394c5fbfd88d9:/IkiWiki/Plugin/aggregate.pm

diff --git a/IkiWiki/Plugin/aggregate.pm b/IkiWiki/Plugin/aggregate.pm
index 0ad24dcd0..6f9c78075 100644
--- a/IkiWiki/Plugin/aggregate.pm
+++ b/IkiWiki/Plugin/aggregate.pm
@@ -4,38 +4,48 @@ package IkiWiki::Plugin::aggregate;
 
 use warnings;
 use strict;
-use IkiWiki;
+use IkiWiki 2.00;
+use HTML::Entities;
+use HTML::Parser;
+use HTML::Tagset;
+use URI;
+use open qw{:utf8 :std};
 
 my %feeds;
 my %guids;
 
 sub import { #{{{
-	IkiWiki::hook(type => "getopt", id => "aggregate", 
-		call => \&getopt);
-	IkiWiki::hook(type => "checkconfig", id => "aggregate",
-		call => \&checkconfig);
-	IkiWiki::hook(type => "filter", id => "aggregate", 
-		call => \&filter);
-	IkiWiki::hook(type => "preprocess", id => "aggregate",
-		call => \&preprocess);
-        IkiWiki::hook(type => "delete", id => "aggregate",
-                call => \&delete);
-	IkiWiki::hook(type => "savestate", id => "aggregate",
-		call => \&savestate);
+	hook(type => "getopt", id => "aggregate", call => \&getopt);
+	hook(type => "checkconfig", id => "aggregate", call => \&checkconfig);
+	hook(type => "filter", id => "aggregate", call => \&filter);
+	hook(type => "preprocess", id => "aggregate", call => \&preprocess);
+        hook(type => "delete", id => "aggregate", call => \&delete);
+	hook(type => "savestate", id => "aggregate", call => \&savestate);
 } # }}}
 
 sub getopt () { #{{{
         eval q{use Getopt::Long};
+	error($@) if $@;
         Getopt::Long::Configure('pass_through');
-        GetOptions("aggregate" => \$IkiWiki::config{aggregate});
+        GetOptions("aggregate" => \$config{aggregate});
 } #}}}
 
 sub checkconfig () { #{{{
-	loadstate();
-	if ($IkiWiki::config{aggregate}) {
+	if ($config{aggregate} && ! ($config{post_commit} && 
+	                             IkiWiki::commit_hook_enabled())) {
+		if (! IkiWiki::lockwiki(0)) {
+			debug("wiki is locked by another process, not aggregating");
+			exit 1;
+		}
+	
+		loadstate();
 		IkiWiki::loadindex();
 		aggregate();
+		expire();
 		savestate();
+		clearstate();
+
+		IkiWiki::unlockwiki();
 	}
 } #}}}
 
@@ -43,6 +53,7 @@ sub filter (@) { #{{{
 	my %params=@_;
 	my $page=$params{page};
 
+	loadstate(); # if not already loaded
 	# Mark all feeds originating on this page as removable;
 	# preprocess will unmark those that still exist.
 	remove_feeds($page);
@@ -55,7 +66,7 @@ sub preprocess (@) { #{{{
 
 	foreach my $required (qw{name url}) {
 		if (! exists $params{$required}) {
-			return "[[aggregate plugin missing $required parameter]]";
+			return "[[aggregate ".sprintf(gettext("missing %s parameter"), $required)."]]";
 		}
 	}
 
@@ -70,19 +81,22 @@ sub preprocess (@) { #{{{
 	$feed->{name}=$name;
 	$feed->{sourcepage}=$params{page};
 	$feed->{url}=$params{url};
-	my $dir=exists $params{dir} ? $params{dir} : IkiWiki::titlepage($params{name});
+	my $dir=exists $params{dir} ? $params{dir} : $params{page}."/".IkiWiki::titlepage($params{name});
 	$dir=~s/^\/+//;
-	($dir)=$dir=~/$IkiWiki::config{wiki_file_regexp}/;
+	($dir)=$dir=~/$config{wiki_file_regexp}/;
 	$feed->{dir}=$dir;
-	$feed->{feedurl}=defined $params{feedurl} ? $params{feedurl} : $params{url};
+	$feed->{feedurl}=defined $params{feedurl} ? $params{feedurl} : "";
 	$feed->{updateinterval}=defined $params{updateinterval} ? $params{updateinterval} * 60 : 15 * 60;
 	$feed->{expireage}=defined $params{expireage} ? $params{expireage} : 0;
 	$feed->{expirecount}=defined $params{expirecount} ? $params{expirecount} : 0;
 	delete $feed->{remove};
+	delete $feed->{expired};
 	$feed->{lastupdate}=0 unless defined $feed->{lastupdate};
 	$feed->{numposts}=0 unless defined $feed->{numposts};
 	$feed->{newposts}=0 unless defined $feed->{newposts};
-	$feed->{message}="new feed" unless defined $feed->{message};
+	$feed->{message}=gettext("new feed") unless defined $feed->{message};
+	$feed->{error}=0 unless defined $feed->{error};
+	$feed->{tags}=[];
 	while (@_) {
 		my $key=shift;
 		my $value=shift;
@@ -92,8 +106,12 @@ sub preprocess (@) { #{{{
 	}
 
 	return "<a href=\"".$feed->{url}."\">".$feed->{name}."</a>: ".
-	       "<i>".$feed->{message}."</i> (".$feed->{numposts}.
-	       " stored posts; ".$feed->{newposts}." new)";
+	       ($feed->{error} ? "<em>" : "").$feed->{message}.
+	       ($feed->{error} ? "</em>" : "").
+	       " (".$feed->{numposts}." ".gettext("posts").
+	       ($feed->{newposts} ? "; ".$feed->{newposts}.
+	                            " ".gettext("new") : "").
+	       ")";
 } # }}}
 
 sub delete (@) { #{{{
@@ -101,23 +119,26 @@ sub delete (@) { #{{{
 
 	# Remove feed data for removed pages.
 	foreach my $file (@files) {
-		my $page=IkiWiki::pagename($file);
+		my $page=pagename($file);
 		remove_feeds($page);
 	}
 } #}}}
 
+my $state_loaded=0;
 sub loadstate () { #{{{
-	if (-e "$IkiWiki::config{wikistatedir}/aggregate") {
-		open (IN, "$IkiWiki::config{wikistatedir}/aggregate" ||
-			die "$IkiWiki::config{wikistatedir}/aggregate: $!");
+	return if $state_loaded;
+	if (-e "$config{wikistatedir}/aggregate") {
+		open(IN, "$config{wikistatedir}/aggregate") ||
+			die "$config{wikistatedir}/aggregate: $!";
 		while (<IN>) {
 			$_=IkiWiki::possibly_foolish_untaint($_);
 			chomp;
 			my $data={};
 			foreach my $i (split(/ /, $_)) {
 				my ($field, $val)=split(/=/, $i, 2);
-				if ($field eq "name" || $field eq "message") {
-					$data->{$field}=IkiWiki::pagetitle($val);
+				if ($field eq "name" || $field eq "feed" ||
+				    $field eq "guid" || $field eq "message") {
+					$data->{$field}=decode_entities($val, " \t\n");
 				}
 				elsif ($field eq "tag") {
 					push @{$data->{tags}}, $val;
@@ -136,12 +157,19 @@ sub loadstate () { #{{{
 		}
 
 		close IN;
+		
+		$state_loaded=1;
 	}
 } #}}}
 
 sub savestate () { #{{{
-	open (OUT, ">$IkiWiki::config{wikistatedir}/aggregate" ||
-		die "$IkiWiki::config{wikistatedir}/aggregate: $!");
+	eval q{use HTML::Entities};
+	error($@) 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 $data (values %feeds, values %guids) {
 		if ($data->{remove}) {
 			if ($data->{name}) {
@@ -156,11 +184,17 @@ sub savestate () { #{{{
 			}
 			next;
 		}
+		elsif ($data->{expired} && exists $data->{page}) {
+			unlink pagefile($data->{page});
+			delete $data->{page};
+			delete $data->{md5};
+		}
 
 		my @line;
 		foreach my $field (keys %$data) {
-			if ($field eq "name" || $field eq "message") {
-				push @line, "$field=".IkiWiki::titlepage($data->{$field});
+			if ($field eq "name" || $field eq "feed" ||
+			    $field eq "guid" || $field eq "message") {
+				push @line, "$field=".encode_entities($data->{$field}, " \t\n");
 			}
 			elsif ($field eq "tags") {
 				push @line, "tag=$_" foreach @{$data->{tags}};
@@ -169,64 +203,135 @@ sub savestate () { #{{{
 				push @line, "$field=".$data->{$field};
 			}
 		}
-		print OUT join(" ", @line)."\n";
+		print OUT join(" ", @line)."\n" || error("write $newfile: $!", $cleanup);
+	}
+	close OUT || error("save $newfile: $!", $cleanup);
+	rename($newfile, "$config{wikistatedir}/aggregate") ||
+		error("rename $newfile: $!", $cleanup);
+} #}}}
+
+sub clearstate () { #{{{
+	%feeds=();
+	%guids=();
+	$state_loaded=0;
+} #}}}
+
+sub expire () { #{{{
+	foreach my $feed (values %feeds) {
+		next unless $feed->{expireage} || $feed->{expirecount};
+		my $count=0;
+		foreach my $item (sort { $IkiWiki::pagectime{$b->{page}} <=> $IkiWiki::pagectime{$a->{page}} }
+		                  grep { exists $_->{page} && $_->{feed} eq $feed->{name} && $IkiWiki::pagectime{$_->{page}} }
+		                  values %guids) {
+			if ($feed->{expireage}) {
+				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));
+					$item->{expired}=1;
+				}
+			}
+			elsif ($feed->{expirecount} &&
+			       $count >= $feed->{expirecount}) {
+				debug(sprintf(gettext("expiring %s"), $item->{page}));
+				$item->{expired}=1;
+			}
+			else {
+				$count++;
+			}
+		}
 	}
-	close OUT;
 } #}}}
 
 sub aggregate () { #{{{
 	eval q{use XML::Feed};
-	die $@ if $@;
+	error($@) if $@;
+	eval q{use URI::Fetch};
+	error($@) if $@;
 	eval q{use HTML::Entities};
-	die $@ if $@;
+	error($@) if $@;
 
-FEED:	foreach my $feed (values %feeds) {
-		next unless time - $feed->{lastupdate} >= $feed->{updateinterval};
+	foreach my $feed (values %feeds) {
+		next unless $config{rebuild} || 
+			time - $feed->{lastupdate} >= $feed->{updateinterval};
 		$feed->{lastupdate}=time;
 		$feed->{newposts}=0;
+		$feed->{message}=sprintf(gettext("processed ok at %s"),
+			displaytime($feed->{lastupdate}));
+		$feed->{error}=0;
 		$IkiWiki::forcerebuild{$feed->{sourcepage}}=1;
 
-		IkiWiki::debug("checking feed ".$feed->{name}." ...");
+		debug(sprintf(gettext("checking feed %s ..."), $feed->{name}));
 
-		my @urls=XML::Feed->find_feeds($feed->{feedurl});
-		if (! @urls) {
-			$feed->{message}="could not find feed at ".$feed->{feedurl};
-			IkiWiki::debug($feed->{message});
-		}
-		foreach my $url (@urls) {
-			my $f=XML::Feed->parse(URI->new($url));
-			if (! $f) {
-				$feed->{message}=XML::Feed->errstr;
-				IkiWiki::debug($feed->{message});
-				next FEED;
+		if (! length $feed->{feedurl}) {
+			my @urls=XML::Feed->find_feeds($feed->{url});
+			if (! @urls) {
+				$feed->{message}=sprintf(gettext("could not find feed at %s"), $feed->{url});
+				$feed->{error}=1;
+				debug($feed->{message});
+				next;
 			}
+			$feed->{feedurl}=pop @urls;
+		}
+		my $res=URI::Fetch->fetch($feed->{feedurl});
+		if (! $res) {
+			$feed->{message}=URI::Fetch->errstr;
+			$feed->{error}=1;
+			debug($feed->{message});
+			next;
+		}
+		if ($res->status == URI::Fetch::URI_GONE()) {
+			$feed->{message}=gettext("feed not found");
+			$feed->{error}=1;
+			debug($feed->{message});
+			next;
+		}
+		my $content=$res->content;
+		my $f=eval{XML::Feed->parse(\$content)};
+		if ($@) {
+			# One common cause of XML::Feed crashing is a feed
+			# that contains invalid UTF-8 sequences. Convert
+			# feed to ascii to try to work around.
+			$feed->{message}.=" ".sprintf(gettext("(invalid UTF-8 stripped from feed)"));
+			$content=Encode::decode_utf8($content);
+			$f=eval{XML::Feed->parse(\$content)};
+		}
+		if ($@) {
+			$feed->{message}=gettext("feed crashed XML::Feed!")." ($@)";
+			$feed->{error}=1;
+			debug($feed->{message});
+			next;
+		}
+		if (! $f) {
+			$feed->{message}=XML::Feed->errstr;
+			$feed->{error}=1;
+			debug($feed->{message});
+			next;
+		}
 
-			foreach my $entry ($f->entries) {
-				add_page(
-					feed => $feed,
-					title => decode_entities($entry->title),
-					link => $entry->link,
-					content => $entry->content->body,
-					guid => defined $entry->id ? $entry->id : time."_".$feed->name,
-					ctime => $entry->issued ? ($entry->issued->epoch || time) : time,
-				);
-			}
+		foreach my $entry ($f->entries) {
+			add_page(
+				feed => $feed,
+				title => defined $entry->title ? decode_entities($entry->title) : "untitled",
+				link => $entry->link,
+				content => defined $entry->content->body ? $entry->content->body : "",
+				guid => defined $entry->id ? $entry->id : time."_".$feed->name,
+				ctime => $entry->issued ? ($entry->issued->epoch || time) : time,
+			);
 		}
-		$feed->{message}="processed ok";
 	}
-
-	# TODO: expiry
 } #}}}
 
 sub add_page (@) { #{{{
 	my %params=@_;
-
+	
 	my $feed=$params{feed};
 	my $guid={};
 	my $mtime;
 	if (exists $guids{$params{guid}}) {
 		# updating an existing post
 		$guid=$guids{$params{guid}};
+		return if $guid->{expired};
 	}
 	else {
 		# new post
@@ -237,19 +342,35 @@ sub add_page (@) { #{{{
 		$feed->{newposts}++;
 
 		# assign it an unused page
-		my $page=$feed->{dir}."/".IkiWiki::titlepage($params{title});
-		($page)=$page=~/$IkiWiki::config{wiki_file_regexp}/;
+		my $page=IkiWiki::titlepage($params{title});
+		# escape slashes and periods in title so it doesn't specify
+		# directory name or trigger ".." disallowing code.
+		$page=~s!([/.])!"__".ord($1)."__"!eg;
+		$page=$feed->{dir}."/".$page;
+		($page)=$page=~/$config{wiki_file_regexp}/;
 		if (! defined $page || ! length $page) {
 			$page=$feed->{dir}."/item";
 		}
-		$page=~s/\.\.//g; # avoid ".." directory tricks
 		my $c="";
-		while (exists $IkiWiki::pagesources{$page.$c} ||
+		while (exists $IkiWiki::pagecase{lc $page.$c} ||
 		       -e pagefile($page.$c)) {
 			$c++
 		}
+
+		# Make sure that the file name isn't too long. 
+		# NB: This doesn't check for path length limits.
+		my $max=POSIX::pathconf($config{srcdir}, &POSIX::_PC_NAME_MAX);
+		if (defined $max && length(htmlfn($page)) >= $max) {
+			$c="";
+			$page=$feed->{dir}."/item";
+			while (exists $IkiWiki::pagecase{lc $page.$c} ||
+			       -e pagefile($page.$c)) {
+				$c++
+			}
+		}
+
 		$guid->{page}=$page;
-		IkiWiki::debug("creating new page $page");
+		debug(sprintf(gettext("creating new page %s"), $page));
 	}
 	$guid->{feed}=$feed->{name};
 	
@@ -257,24 +378,25 @@ sub add_page (@) { #{{{
 	# to avoid unneccessary rebuilding. The mtime from rss cannot be
 	# trusted; let's use a digest.
 	eval q{use Digest::MD5 'md5_hex'};
+	error($@) if $@;
 	require Encode;
 	my $digest=md5_hex(Encode::encode_utf8($params{content}));
-	return unless ! exists $guid->{md5} || $guid->{md5} ne $digest;
+	return unless ! exists $guid->{md5} || $guid->{md5} ne $digest || $config{rebuild};
 	$guid->{md5}=$digest;
 
 	# Create the page.
-	my $template=IkiWiki::template("aggregatepost.tmpl", blind_cache => 1);
-	my $content=$params{content};
-	$params{content}=~s/(?<!\\)\[\[/\\\[\[/g; # escape accidental wikilinks
-	                                          # and preprocessor stuff
-	$template->param(content => $params{content});
-	$template->param(url => $feed->{url});
+	my $template=template("aggregatepost.tmpl", blind_cache => 1);
+	$template->param(title => $params{title})
+		if defined $params{title} && length($params{title});
+	$template->param(content => htmlescape(htmlabs($params{content}, $feed->{feedurl})));
 	$template->param(name => $feed->{name});
-	$template->param(link => $params{link}) if defined $params{link};
+	$template->param(url => $feed->{url});
+	$template->param(permalink => urlabs($params{link}, $feed->{feedurl}))
+		if defined $params{link};
 	if (ref $feed->{tags}) {
 		$template->param(tags => [map { tag => $_ }, @{$feed->{tags}}]);
 	}
-	IkiWiki::writefile($guid->{page}.".html", $IkiWiki::config{srcdir},
+	writefile(htmlfn($guid->{page}), $config{srcdir},
 		$template->output);
 
 	# Set the mtime, this lets the build process get the right creation
@@ -282,6 +404,58 @@ sub add_page (@) { #{{{
 	utime $mtime, $mtime, pagefile($guid->{page}) if defined $mtime;
 } #}}}
 
+sub htmlescape ($) { #{{{
+	# escape accidental wikilinks and preprocessor stuff
+	my $html=shift;
+	$html=~s/(?<!\\)\[\[/\\\[\[/g;
+	return $html;
+} #}}}
+
+sub urlabs ($$) { #{{{
+	my $url=shift;
+	my $urlbase=shift;
+
+	URI->new_abs($url, $urlbase)->as_string;
+} #}}}
+
+sub htmlabs ($$) { #{{{
+	# Convert links in html from relative to absolute.
+	# Note that this is a heuristic, which is not specified by the rss
+	# spec and may not be right for all feeds. Also, see Debian
+	# bug #381359.
+	my $html=shift;
+	my $urlbase=shift;
+
+	my $ret="";
+	my $p = HTML::Parser->new(api_version => 3);
+	$p->handler(default => sub { $ret.=join("", @_) }, "text");
+	$p->handler(start => sub {
+		my ($tagname, $pos, $text) = @_;
+		if (ref $HTML::Tagset::linkElements{$tagname}) {
+			while (4 <= @$pos) {
+				# use attribute sets from right to left
+				# to avoid invalidating the offsets
+				# when replacing the values
+				my($k_offset, $k_len, $v_offset, $v_len) =
+					splice(@$pos, -4);
+				my $attrname = lc(substr($text, $k_offset, $k_len));
+				next unless grep { $_ eq $attrname } @{$HTML::Tagset::linkElements{$tagname}};
+				next unless $v_offset; # 0 v_offset means no value
+				my $v = substr($text, $v_offset, $v_len);
+				$v =~ s/^([\'\"])(.*)\1$/$2/;
+				my $new_v=urlabs($v, $urlbase);
+				$new_v =~ s/\"/&quot;/g; # since we quote with ""
+				substr($text, $v_offset, $v_len) = qq("$new_v");
+			}
+		}
+		$ret.=$text;
+	}, "tagname, tokenpos, text");
+	$p->parse($html);
+	$p->eof;
+
+	return $ret;
+} #}}}
+
 sub remove_feeds () { #{{{
 	my $page=shift;
 
@@ -297,7 +471,11 @@ sub remove_feeds () { #{{{
 sub pagefile ($) { #{{{
 	my $page=shift;
 
-	return "$IkiWiki::config{srcdir}/$page.html";
+	return "$config{srcdir}/".htmlfn($page);
+} #}}}
+
+sub htmlfn ($) { #{{{
+	return shift().".".$config{htmlext};
 } #}}}
 
 1