X-Git-Url: http://git.vanrenterghem.biz/git.ikiwiki.info.git/blobdiff_plain/5466a1daf99e8e4c67a19f871aaf558312569652..bd8198fb77ed5cf77ddeb860ca1ac983bfaacbbd:/IkiWiki/Plugin/search.pm

diff --git a/IkiWiki/Plugin/search.pm b/IkiWiki/Plugin/search.pm
index c94974ac4..24b16fe2a 100644
--- a/IkiWiki/Plugin/search.pm
+++ b/IkiWiki/Plugin/search.pm
@@ -1,130 +1,289 @@
 #!/usr/bin/perl
-# hyperestraier search engine plugin
+# xapian-omega search engine plugin
 package IkiWiki::Plugin::search;
 
 use warnings;
 use strict;
-use IkiWiki;
-
-sub import { #{{{
-	IkiWiki::hook(type => "checkconfig", id => "hyperestraier",
-		call => \&checkconfig);
-	IkiWiki::hook(type => "pagetemplate", id => "hyperestraier",
-		call => \&pagetemplate);
-	IkiWiki::hook(type => "delete", id => "hyperestraier",
-		call => \&delete);
-	IkiWiki::hook(type => "change", id => "hyperestraier",
-		call => \&change);
-	IkiWiki::hook(type => "cgi", id => "hyperestraier",
-		call => \&cgi);
-} # }}}
-
-sub checkconfig () { #{{{
+use IkiWiki 3.00;
+
+sub import {
+	hook(type => "getsetup", id => "search", call => \&getsetup);
+	hook(type => "checkconfig", id => "search", call => \&checkconfig);
+	hook(type => "pagetemplate", id => "search", call => \&pagetemplate);
+	hook(type => "indexhtml", id => "search", call => \&indexhtml);
+	hook(type => "delete", id => "search", call => \&delete);
+	hook(type => "cgi", id => "search", call => \&cgi);
+	hook(type => "disable", id => "search", call => \&disable);
+	hook(type => "needsbuild", id => "search", call => \&needsbuild);
+		
+	eval q{ use Search::Xapian }; # load early to work around #622591
+}
+
+sub getsetup () {
+	return
+		plugin => {
+			safe => 1,
+			rebuild => 1,
+			section => "web",
+		},
+		omega_cgi => {
+			type => "string",
+			example => "/usr/lib/cgi-bin/omega/omega",
+			description => "path to the omega cgi program",
+			safe => 0, # external program
+			rebuild => 0,
+		},
+		google_search => {
+			type => "booblean",
+			example => 1,
+			description => "use google site search rather than internal xapian index?",
+			safe => 1,
+			rebuild => 0,
+		},
+}
+
+sub checkconfig () {
 	foreach my $required (qw(url cgiurl)) {
-		if (! length $IkiWiki::config{$required}) {
-			IkiWiki::error("Must specify $required when using the search plugin\n");
+		if (! length $config{$required}) {
+			error(sprintf(gettext("Must specify %s when using the %s plugin"), $required, 'search'));
 		}
 	}
-} #}}}
+	
+	if (! defined $config{omega_cgi}) {
+		$config{omega_cgi}="/usr/lib/cgi-bin/omega/omega";
+	}
 
-sub pagetemplate ($$) { #{{{
-	my $page=shift;
-	my $template=shift;
+	# This is a mass dependency, so if the search form template
+	# changes, every page is rebuilt.
+	add_depends("", "templates/searchform.tmpl");
+}
 
-	# Add search box to page header.
-	$template->param(searchform => qq{
-<form method="get" action="$IkiWiki::config{cgiurl}" id="searchform">
-<div>
-<input type="text" name="phrase" value="" size="16" />
-<input type="hidden" name="enc" value="UTF-8" />
-<input type="hidden" name="do" value="hyperestraier" />
-</div>
-</form>
-});
-} #}}}
-
-sub delete (@) { #{{{
-	IkiWiki::debug("cleaning hyperestraier search index");
-	IkiWiki::estcmd("purge -cl");
-	IkiWiki::estcfg();
-} #}}}
-
-sub change (@) { #{{{
-	IkiWiki::debug("updating hyperestraier search index");
-	IkiWiki::estcmd("gather -cm -bc -cl -sd",
-		map {
-			$IkiWiki::config{destdir}."/".$IkiWiki::renderedfiles{IkiWiki::pagename($_)}
-		} @_
-	);
-	IkiWiki::estcfg();
-} #}}}
+my $form;
+sub pagetemplate (@) {
+	my %params=@_;
+	my $page=$params{page};
+	my $template=$params{template};
 
-sub cgi ($) { #{{{
-	my $cgi=shift;
+	# Add search box to page header.
+	if ($template->query(name => "searchform")) {
+		if (! defined $form) {
+			my $searchform = template("searchform.tmpl", blind_cache => 1);
+			$searchform->param(searchaction => IkiWiki::cgiurl());
+			$searchform->param(html5 => $config{html5});
+			$form=$searchform->output;
+		}
 
-	if (defined $cgi->param('phrase')) {
-		# only works for GET requests
-		chdir("$IkiWiki::config{wikistatedir}/hyperestraier") || IkiWiki::error("chdir: $!");
-		exec("./".IkiWiki::basename($IkiWiki::config{cgiurl})) || IkiWiki::error("estseek.cgi failed");
+		$template->param(searchform => $form);
 	}
-} #}}}
+}
+
+my $scrubber;
+my $stemmer;
+sub indexhtml (@) {
+	my %params=@_;
 
-# Easier to keep these in the IkiWiki namespace.
-package IkiWiki;
+	return if $config{google_search};
 
-my $configured=0;
-sub estcfg () { #{{{
-	return if $configured;
-	$configured=1;
+	setupfiles();
+
+	# A unique pageterm is used to identify the document for a page.
+	my $pageterm=pageterm($params{page});
+	return unless defined $pageterm;
 	
-	my $estdir="$config{wikistatedir}/hyperestraier";
-	my $cgi=basename($config{cgiurl});
-	$cgi=~s/\..*$//;
-	open(TEMPLATE, ">$estdir/$cgi.tmpl") ||
-		error("write $estdir/$cgi.tmpl: $!");
-	print TEMPLATE misctemplate("search", 
-		"<!--ESTFORM-->\n\n<!--ESTRESULT-->\n\n<!--ESTINFO-->\n\n");
-	close TEMPLATE;
-	open(TEMPLATE, ">$estdir/$cgi.conf") ||
-		error("write $estdir/$cgi.conf: $!");
-	my $template=HTML::Template->new(
-		filename => "$config{templatedir}/estseek.conf"
-	);
-	eval q{use Cwd 'abs_path'};
-	$template->param(
-		index => $estdir,
-		tmplfile => "$estdir/$cgi.tmpl",
-		destdir => abs_path($config{destdir}),
-		url => $config{url},
-	);
-	print TEMPLATE $template->output;
-	close TEMPLATE;
-	$cgi="$estdir/".basename($config{cgiurl});
-	unlink($cgi);
-	symlink("/usr/lib/estraier/estseek.cgi", $cgi) ||
-		error("symlink $cgi: $!");
-} # }}}
-
-sub estcmd ($;@) { #{{{
-	my @params=split(' ', shift);
-	push @params, "-cl", "$config{wikistatedir}/hyperestraier";
-	if (@_) {
-		push @params, "-";
+	my $db=xapiandb();
+	my $doc=Search::Xapian::Document->new();
+	my $caption=pagetitle($params{page});
+	my $title;
+	if (exists $pagestate{$params{page}}{meta} &&
+		exists $pagestate{$params{page}}{meta}{title}) {
+		$title=$pagestate{$params{page}}{meta}{title};
 	}
+	else {
+		$title=$caption;
+	}
+
+	# Remove html from text to be indexed.
+	if (! defined $scrubber) {
+		eval q{use HTML::Scrubber};
+		if (! $@) {
+			$scrubber=HTML::Scrubber->new(allow => []);
+		}
+	}
+	my $toindex = defined $scrubber ? $scrubber->scrub($params{content}) : $params{content};
+	
+	# Take 512 characters for a sample, then extend it out
+	# if it stopped in the middle of a word.
+	my $size=512;
+	my ($sample)=substr($toindex, 0, $size);
+	if (length($sample) == $size) {
+		my $max=length($toindex);
+		my $next;
+		while ($size < $max &&
+		       ($next=substr($toindex, $size++, 1)) !~ /\s/) {
+			$sample.=$next;
+		}
+	}
+	$sample=~s/\n/ /g;
 	
-	my $pid=open(CHILD, "|-");
-	if ($pid) {
-		# parent
-		foreach (@_) {
-			print CHILD "$_\n";
+	my $url=urlto($params{destpage}, "");
+	if (defined $pagestate{$params{page}}{meta}{permalink}) {
+		$url=$pagestate{$params{page}}{meta}{permalink}
+	}
+
+	# data used by omega
+	# Decode html entities in it, since omega re-encodes them.
+	eval q{use HTML::Entities};
+	error $@ if $@;
+	$doc->set_data(
+		"url=".$url."\n".
+		"sample=".decode_entities($sample)."\n".
+		"caption=".decode_entities($caption)."\n".
+		"modtime=$IkiWiki::pagemtime{$params{page}}\n".
+		"size=".length($params{content})."\n"
+	);
+
+	# Index document and add terms for other metadata.
+	my $tg = Search::Xapian::TermGenerator->new();
+	if (! $stemmer) {
+		my $langcode=$ENV{LANG} || "en";
+		$langcode=~s/_.*//;
+
+		# This whitelist is here to work around a xapian bug (#486138)
+		my @whitelist=qw{da de en es fi fr hu it no pt ru ro sv tr};
+
+		if (grep { $_ eq $langcode } @whitelist) {
+			$stemmer=Search::Xapian::Stem->new($langcode);
 		}
-		close(CHILD) || error("estcmd @params exited nonzero: $?");
+		else {
+			$stemmer=Search::Xapian::Stem->new("english");
+		}
+	}
+	$tg->set_stemmer($stemmer);
+	$tg->set_document($doc);
+	$tg->index_text($params{page}, 2);
+	$tg->index_text($caption, 2);
+	$tg->index_text($title, 2) if $title ne $caption;
+	$tg->index_text($toindex);
+	$tg->index_text(lc($title), 1, "S"); # for title:foo
+	foreach my $link (@{$links{$params{page}}}) {
+		$tg->index_text(lc($link), 1, "XLINK"); # for link:bar
+	}
+
+	$doc->add_term($pageterm);
+	$db->replace_document_by_term($pageterm, $doc);
+}
+
+sub delete (@) {
+	return if $config{google_search};
+
+	my $db=xapiandb();
+	foreach my $page (@_) {
+		my $pageterm=pageterm(pagename($page));
+		$db->delete_document_by_term($pageterm) if defined $pageterm;
+	}
+}
+
+sub cgi ($) {
+	my $cgi=shift;
+
+	if (defined $cgi->param('P')) {
+		if ($config{google_search}) {
+			print $cgi->redirect("https://www.google.com/search?sitesearch=$config{url}&q=".$cgi->param('P'));
+			exit 0;
+		}
+		else {
+			# only works for GET requests
+			chdir("$config{wikistatedir}/xapian") || error("chdir: $!");
+			$ENV{OMEGA_CONFIG_FILE}="./omega.conf";
+			$ENV{CGIURL}=IkiWiki::cgiurl();
+			IkiWiki::loadindex();
+			$ENV{HELPLINK}=htmllink("", "", "ikiwiki/searching",
+				noimageinline => 1, linktext => "Help");
+			exec($config{omega_cgi}) || error("$config{omega_cgi} failed: $!");
+		}
+	}
+}
+
+sub pageterm ($) {
+	my $page=shift;
+
+	# 240 is the number used by omindex to decide when to hash an
+	# overlong term. This does not use a compatible hash method though.
+	eval q{use Encode};
+	if (length encode_utf8($page) > 240) {
+		eval q{use Digest::SHA};
+		if ($@) {
+			debug("search: ".sprintf(gettext("need Digest::SHA to index %s"), $page)) if $@;
+			return undef;
+		}
+
+		# Note no colon, therefore it's guaranteed to not overlap
+		# with a page with the same name as the hash..
+		return "U".lc(Digest::SHA::sha1_hex($page));
 	}
 	else {
-		# child
-		open(STDOUT, "/dev/null"); # shut it up (closing won't work)
-		exec("estcmd", @params) || error("can't run estcmd");
+		return "U:".$page;
+	}
+}
+
+my $db;
+sub xapiandb () {
+	if (! defined $db) {
+		eval q{
+			use Search::Xapian;
+			use Search::Xapian::WritableDatabase;
+		};
+		error($@) if $@;
+		$db=Search::Xapian::WritableDatabase->new($config{wikistatedir}."/xapian/default",
+			Search::Xapian::DB_CREATE_OR_OPEN());
+	}
+	return $db;
+}
+
+{
+my $setup=0;
+sub setupfiles () {
+	if (! $setup and (! -e $config{wikistatedir}."/xapian" || $config{rebuild})) {
+		writefile("omega.conf", $config{wikistatedir}."/xapian",
+			"database_dir .\n".
+			"template_dir ./templates\n");
+		omega_template();	
+		$setup=1;
+	}
+}
+}
+
+sub needsbuild {
+	my $list=shift;
+	if (grep {
+		$_ eq "templates/page.tmpl" ||
+		$_ eq "templates/searchquery.tmpl"
+	} @$list) {
+		omega_template();
+	}
+}
+
+sub omega_template {
+	# Avoid omega interpreting anything in the cgitemplate
+	# as an omegascript command.
+	eval q{use IkiWiki::CGI};
+	my $template=IkiWiki::cgitemplate(undef, gettext("search"), "\0",
+		searchform => "", # avoid showing the small search form
+	);
+	eval q{use HTML::Entities};
+	error $@ if $@;
+	$template=encode_entities($template, '\$');
+
+	my $querytemplate=readfile(IkiWiki::template_file("searchquery.tmpl"));
+	$template=~s/\0/$querytemplate/;
+	writefile("query", $config{wikistatedir}."/xapian/templates",
+		$template);
+}
+
+sub disable () {
+	if (-d $config{wikistatedir}."/xapian") {
+		system("rm", "-rf", $config{wikistatedir}."/xapian");
 	}
-} #}}}
+}
 
 1