X-Git-Url: http://git.vanrenterghem.biz/git.ikiwiki.info.git/blobdiff_plain/302bb836f298c462da8e8bff7b04f1b00ded4665..9b9ecda62f893a2109ccdb6bcdb2358f217f7fef:/IkiWiki/Plugin/aggregate.pm diff --git a/IkiWiki/Plugin/aggregate.pm b/IkiWiki/Plugin/aggregate.pm index e93492863..7789c4c2a 100644 --- a/IkiWiki/Plugin/aggregate.pm +++ b/IkiWiki/Plugin/aggregate.pm @@ -4,7 +4,7 @@ package IkiWiki::Plugin::aggregate; use warnings; use strict; -use IkiWiki 2.00; +use IkiWiki 3.00; use HTML::Parser; use HTML::Tagset; use HTML::Entities; @@ -14,8 +14,9 @@ use open qw{:utf8 :std}; my %feeds; my %guids; -sub import { #{{{ +sub import { hook(type => "getopt", id => "aggregate", call => \&getopt); + hook(type => "getsetup", id => "aggregate", call => \&getsetup); hook(type => "checkconfig", id => "aggregate", call => \&checkconfig); hook(type => "needsbuild", id => "aggregate", call => \&needsbuild); hook(type => "preprocess", id => "aggregate", call => \&preprocess); @@ -25,9 +26,9 @@ sub import { #{{{ if (exists $config{aggregate_webtrigger} && $config{aggregate_webtrigger}) { hook(type => "cgi", id => "aggregate", call => \&cgi); } -} # }}} +} -sub getopt () { #{{{ +sub getopt () { eval q{use Getopt::Long}; error($@) if $@; Getopt::Long::Configure('pass_through'); @@ -35,16 +36,42 @@ sub getopt () { #{{{ "aggregate" => \$config{aggregate}, "aggregateinternal!" => \$config{aggregateinternal}, ); -} #}}} +} + +sub getsetup () { + return + plugin => { + safe => 1, + rebuild => undef, + }, + aggregateinternal => { + type => "boolean", + example => 1, + description => "enable aggregation to internal pages?", + safe => 0, # enabling needs manual transition + rebuild => 0, + }, + aggregate_webtrigger => { + type => "boolean", + example => 0, + description => "allow aggregation to be triggered via the web?", + safe => 1, + rebuild => 0, + }, +} + +sub checkconfig () { + if (! defined $config{aggregateinternal}) { + $config{aggregateinternal}=1; + } -sub checkconfig () { #{{{ if ($config{aggregate} && ! ($config{post_commit} && IkiWiki::commit_hook_enabled())) { launchaggregation(); } -} #}}} +} -sub cgi ($) { #{{{ +sub cgi ($) { my $cgi=shift; if (defined $cgi->param('do') && @@ -67,9 +94,9 @@ sub cgi ($) { #{{{ } exit 0; } -} #}}} +} -sub launchaggregation () { #{{{ +sub launchaggregation () { # See if any feeds need aggregation. loadstate(); my @feeds=needsaggregate(); @@ -112,58 +139,64 @@ sub launchaggregation () { #{{{ unlockaggregate(); return 1; -} #}}} +} # Pages with extension _aggregated have plain html markup, pass through. -sub htmlize (@) { #{{{ +sub htmlize (@) { my %params=@_; return $params{content}; -} #}}} - -sub migrate_to_internal { #{{{ +} +# Used by ikiwiki-transition aggregateinternal. +sub migrate_to_internal { if (! lockaggregate()) { - error("an aggregation process is already running"); - return; + error("an aggregation process is currently running"); } IkiWiki::lockwiki(); loadstate(); + $config{verbose}=1; foreach my $data (values %guids) { next unless $data->{page}; - + next if $data->{expired}; + $config{aggregateinternal} = 0; - my $oldname = pagefile($data->{page}); - + my $oldname = "$config{srcdir}/".htmlfn($data->{page}); + my $oldoutput = $config{destdir}."/".IkiWiki::htmlpage($data->{page}); + $config{aggregateinternal} = 1; - my $newname = pagefile($data->{page}); - - print "I: $oldname -> $newname\n"; + my $newname = "$config{srcdir}/".htmlfn($data->{page}); + + debug "moving $oldname -> $newname"; if (-e $newname) { if (-e $oldname) { error("$newname already exists"); } else { - print STDERR - "W: already renamed to $newname?\n"; + debug("already renamed to $newname?"); } } elsif (-e $oldname) { rename($oldname, $newname) || error("$!"); } else { - print "W: $oldname not found\n"; + debug("$oldname not found"); + } + if (-e $oldoutput) { + require IkiWiki::Render; + debug("removing output file $oldoutput"); + IkiWiki::prune($oldoutput); } } - + savestate(); IkiWiki::unlockwiki; - + unlockaggregate(); -} #}}} +} -sub needsbuild (@) { #{{{ +sub needsbuild (@) { my $needsbuild=shift; loadstate(); @@ -177,9 +210,9 @@ sub needsbuild (@) { #{{{ markunseen($feed->{sourcepage}); } } -} # }}} +} -sub preprocess (@) { #{{{ +sub preprocess (@) { my %params=@_; foreach my $required (qw{name url}) { @@ -199,7 +232,7 @@ sub preprocess (@) { #{{{ $feed->{name}=$name; $feed->{sourcepage}=$params{page}; $feed->{url}=$params{url}; - my $dir=exists $params{dir} ? $params{dir} : $params{page}."/".IkiWiki::titlepage($params{name}); + my $dir=exists $params{dir} ? $params{dir} : $params{page}."/".titlepage($params{name}); $dir=~s/^\/+//; ($dir)=$dir=~/$config{wiki_file_regexp}/; $feed->{dir}=$dir; @@ -216,6 +249,7 @@ sub preprocess (@) { #{{{ $feed->{template}=$params{template} . ".tmpl"; delete $feed->{unseen}; $feed->{lastupdate}=0 unless defined $feed->{lastupdate}; + $feed->{lasttry}=$feed->{lastupdate} unless defined $feed->{lasttry}; $feed->{numposts}=0 unless defined $feed->{numposts}; $feed->{newposts}=0 unless defined $feed->{newposts}; $feed->{message}=gettext("new feed") unless defined $feed->{message}; @@ -236,9 +270,9 @@ sub preprocess (@) { #{{{ ($feed->{newposts} ? "; ".$feed->{newposts}. " ".gettext("new") : ""). ")"; -} # }}} +} -sub delete (@) { #{{{ +sub delete (@) { my @files=@_; # Remove feed data for removed pages. @@ -246,9 +280,9 @@ sub delete (@) { #{{{ my $page=pagename($file); markunseen($page); } -} #}}} +} -sub markunseen ($) { #{{{ +sub markunseen ($) { my $page=shift; foreach my $id (keys %feeds) { @@ -256,15 +290,15 @@ sub markunseen ($) { #{{{ $feeds{$id}->{unseen}=1; } } -} #}}} +} my $state_loaded=0; -sub loadstate () { #{{{ +sub loadstate () { return if $state_loaded; $state_loaded=1; if (-e "$config{wikistatedir}/aggregate") { - open(IN, "$config{wikistatedir}/aggregate") || + open(IN, "<", "$config{wikistatedir}/aggregate") || die "$config{wikistatedir}/aggregate: $!"; while () { $_=IkiWiki::possibly_foolish_untaint($_); @@ -294,14 +328,14 @@ sub loadstate () { #{{{ close IN; } -} #}}} +} -sub savestate () { #{{{ +sub savestate () { return unless $state_loaded; garbage_collect(); my $newfile="$config{wikistatedir}/aggregate.new"; my $cleanup = sub { unlink($newfile) }; - open (OUT, ">$newfile") || error("open $newfile: $!", $cleanup); + open (OUT, ">", $newfile) || error("open $newfile: $!", $cleanup); foreach my $data (values %feeds, values %guids) { my @line; foreach my $field (keys %$data) { @@ -313,7 +347,8 @@ sub savestate () { #{{{ push @line, "tag=$_" foreach @{$data->{tags}}; } else { - push @line, "$field=".$data->{$field}; + push @line, "$field=".$data->{$field} + if defined $data->{$field}; } } print OUT join(" ", @line)."\n" || error("write $newfile: $!", $cleanup); @@ -321,9 +356,23 @@ sub savestate () { #{{{ close OUT || error("save $newfile: $!", $cleanup); rename($newfile, "$config{wikistatedir}/aggregate") || error("rename $newfile: $!", $cleanup); -} #}}} -sub garbage_collect () { #{{{ + my $timestamp=undef; + foreach my $feed (keys %feeds) { + my $t=$feeds{$feed}->{lastupdate}+$feeds{$feed}->{updateinterval}; + if (! defined $timestamp || $timestamp > $t) { + $timestamp=$t; + } + } + $newfile=~s/\.new$/time/; + open (OUT, ">", $newfile) || error("open $newfile: $!", $cleanup); + if (defined $timestamp) { + print OUT $timestamp."\n"; + } + close OUT || error("save $newfile: $!", $cleanup); +} + +sub garbage_collect () { foreach my $name (keys %feeds) { # remove any feeds that were not seen while building the pages # that used to contain them @@ -335,20 +384,20 @@ sub garbage_collect () { #{{{ foreach my $guid (values %guids) { # any guid whose feed is gone should be removed if (! exists $feeds{$guid->{feed}}) { - unlink pagefile($guid->{page}) + unlink "$config{srcdir}/".htmlfn($guid->{page}) if exists $guid->{page}; delete $guids{$guid->{guid}}; } # handle expired guids elsif ($guid->{expired} && exists $guid->{page}) { - unlink pagefile($guid->{page}); + unlink "$config{srcdir}/".htmlfn($guid->{page}); delete $guid->{page}; delete $guid->{md5}; } } -} #}}} +} -sub mergestate () { #{{{ +sub mergestate () { # Load the current state in from disk, and merge into it # values from the state in memory that might have changed # during aggregation. @@ -361,14 +410,15 @@ sub mergestate () { #{{{ # fields. foreach my $name (keys %myfeeds) { if (exists $feeds{$name}) { - foreach my $field (qw{message lastupdate numposts - newposts error}) { + foreach my $field (qw{message lastupdate lasttry + numposts newposts error}) { $feeds{$name}->{$field}=$myfeeds{$name}->{$field}; } } } # New guids can be created during aggregation. + # Guids have a few fields that may be updated during aggregation. # It's also possible that guids were removed from the on-disk state # while the aggregation was in process. That would only happen if # their feed was also removed, so any removed guids added back here @@ -377,25 +427,30 @@ sub mergestate () { #{{{ if (! exists $guids{$guid}) { $guids{$guid}=$myguids{$guid}; } + else { + foreach my $field (qw{md5}) { + $guids{$guid}->{$field}=$myguids{$guid}->{$field}; + } + } } -} #}}} +} -sub clearstate () { #{{{ +sub clearstate () { %feeds=(); %guids=(); $state_loaded=0; -} #}}} +} -sub expire () { #{{{ +sub expire () { foreach my $feed (values %feeds) { next unless $feed->{expireage} || $feed->{expirecount}; my $count=0; my %seen; - foreach my $item (sort { $IkiWiki::pagectime{$b->{page}} <=> $IkiWiki::pagectime{$a->{page}} } - grep { exists $_->{page} && $_->{feed} eq $feed->{name} && $IkiWiki::pagectime{$_->{page}} } + foreach my $item (sort { ($IkiWiki::pagectime{$b->{page}} || 0) <=> ($IkiWiki::pagectime{$a->{page}} || 0) } + grep { exists $_->{page} && $_->{feed} eq $feed->{name} } values %guids) { if ($feed->{expireage}) { - my $days_old = (time - $IkiWiki::pagectime{$item->{page}}) / 60 / 60 / 24; + my $days_old = (time - ($IkiWiki::pagectime{$item->{page}} || 0)) / 60 / 60 / 24; if ($days_old > $feed->{expireage}) { debug(sprintf(gettext("expiring %s (%s days old)"), $item->{page}, int($days_old))); @@ -415,24 +470,24 @@ sub expire () { #{{{ } } } -} #}}} +} -sub needsaggregate () { #{{{ +sub needsaggregate () { return values %feeds if $config{rebuild}; return grep { time - $_->{lastupdate} >= $_->{updateinterval} } values %feeds; -} #}}} +} -sub aggregate (@) { #{{{ +sub aggregate (@) { eval q{use XML::Feed}; error($@) if $@; eval q{use URI::Fetch}; error($@) if $@; foreach my $feed (@_) { - $feed->{lastupdate}=time; + $feed->{lasttry}=time; $feed->{newposts}=0; - $feed->{message}=sprintf(gettext("processed ok at %s"), - displaytime($feed->{lastupdate})); + $feed->{message}=sprintf(gettext("last checked %s"), + displaytime($feed->{lasttry})); $feed->{error}=0; debug(sprintf(gettext("checking feed %s ..."), $feed->{name})); @@ -454,6 +509,10 @@ sub aggregate (@) { #{{{ debug($feed->{message}); next; } + + # lastupdate is only set if we were able to contact the server + $feed->{lastupdate}=$feed->{lasttry}; + if ($res->status == URI::Fetch::URI_GONE()) { $feed->{message}=gettext("feed not found"); $feed->{error}=1; @@ -467,15 +526,19 @@ sub aggregate (@) { #{{{ # 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, 0); - $f=eval{XML::Feed->parse(\$content)}; + $f=eval { + $content=Encode::decode_utf8($content, 0); + 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, 0); - $f=eval{XML::Feed->parse(\$content)}; + $f=eval { + $content=Encode::decode_utf8($content, 0); + XML::Feed->parse(\$content) + }; } if ($@) { $feed->{message}=gettext("feed crashed XML::Feed!")." ($@)"; @@ -491,20 +554,32 @@ sub aggregate (@) { #{{{ } foreach my $entry ($f->entries) { + # XML::Feed doesn't work around XML::Atom's bizarre + # API, so we will. Real unicode strings? Yes please. + # See [[bugs/Aggregated_Atom_feeds_are_double-encoded]] + local $XML::Atom::ForceUnicode = 1; + + my $c=$entry->content; + # atom feeds may have no content, only a summary + if (! defined $c && ref $entry->summary) { + $c=$entry->summary; + } + 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 : "", + content => (defined $c && defined $c->body) ? $c->body : "", guid => defined $entry->id ? $entry->id : time."_".$feed->{name}, ctime => $entry->issued ? ($entry->issued->epoch || time) : time, + base => (defined $c && $c->can("base")) ? $c->base : undef, ); } } -} #}}} +} -sub add_page (@) { #{{{ +sub add_page (@) { my %params=@_; my $feed=$params{feed}; @@ -524,7 +599,7 @@ sub add_page (@) { #{{{ $feed->{newposts}++; # assign it an unused page - my $page=IkiWiki::titlepage($params{title}); + my $page=titlepage($params{title}); # escape slashes and periods in title so it doesn't specify # directory name or trigger ".." disallowing code. $page=~s!([/.])!"__".ord($1)."__"!eg; @@ -535,7 +610,7 @@ sub add_page (@) { #{{{ } my $c=""; while (exists $IkiWiki::pagecase{lc $page.$c} || - -e pagefile($page.$c)) { + -e "$config{srcdir}/".htmlfn($page.$c)) { $c++ } @@ -546,7 +621,7 @@ sub add_page (@) { #{{{ $c=""; $page=$feed->{dir}."/item"; while (exists $IkiWiki::pagecase{lc $page.$c} || - -e pagefile($page.$c)) { + -e "$config{srcdir}/".htmlfn($page.$c)) { $c++ } } @@ -570,7 +645,8 @@ sub add_page (@) { #{{{ my $template=template($feed->{template}, blind_cache => 1); $template->param(title => $params{title}) if defined $params{title} && length($params{title}); - $template->param(content => htmlescape(htmlabs($params{content}, $feed->{feedurl}))); + $template->param(content => wikiescape(htmlabs($params{content}, + defined $params{base} ? $params{base} : $feed->{feedurl}))); $template->param(name => $feed->{name}); $template->param(url => $feed->{url}); $template->param(copyright => $params{copyright}) @@ -583,27 +659,34 @@ sub add_page (@) { #{{{ writefile(htmlfn($guid->{page}), $config{srcdir}, $template->output); - # 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 && $mtime <= time; -} #}}} + if (defined $mtime && $mtime <= time) { + # Set the mtime, this lets the build process get the right + # creation time on record for the new page. + utime $mtime, $mtime, "$config{srcdir}/".htmlfn($guid->{page}); + # Store it in pagectime for expiry code to use also. + $IkiWiki::pagectime{$guid->{page}}=$mtime + unless exists $IkiWiki::pagectime{$guid->{page}}; + } + else { + # Dummy value for expiry code. + $IkiWiki::pagectime{$guid->{page}}=time + unless exists $IkiWiki::pagectime{$guid->{page}}; + } +} -sub htmlescape ($) { #{{{ +sub wikiescape ($) { # escape accidental wikilinks and preprocessor stuff - my $html=shift; - $html=~s/(?new_abs($url, $urlbase)->as_string; -} #}}} +} -sub htmlabs ($$) { #{{{ +sub htmlabs ($$) { # 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 @@ -639,21 +722,15 @@ sub htmlabs ($$) { #{{{ $p->eof; return $ret; -} #}}} - -sub pagefile ($) { #{{{ - my $page=shift; - - return "$config{srcdir}/".htmlfn($page); -} #}}} +} -sub htmlfn ($) { #{{{ +sub htmlfn ($) { return shift().".".($config{aggregateinternal} ? "_aggregated" : $config{htmlext}); -} #}}} +} my $aggregatelock; -sub lockaggregate () { #{{{ +sub lockaggregate () { # Take an exclusive lock to prevent multiple concurrent aggregators. # Returns true if the lock was aquired. if (! -d $config{wikistatedir}) { @@ -666,11 +743,11 @@ sub lockaggregate () { #{{{ return 0; } return 1; -} #}}} +} -sub unlockaggregate () { #{{{ +sub unlockaggregate () { return close($aggregatelock) if $aggregatelock; return; -} #}}} +} 1