Fix escaping of filenames in historyurl. (Thanks, aj)

[git.ikiwiki.info.git] / IkiWiki / Plugin / search.pm

diff --git a/IkiWiki/Plugin/search.pm b/IkiWiki/Plugin/search.pm
index 3ac4351979e7b96bd81c334d7d873cb40d612498..8fb9dff0ca082bb16b6332c052e848408db162b8 100644 (file)
--- a/IkiWiki/Plugin/search.pm
+++ b/IkiWiki/Plugin/search.pm
@@ -4,30 +4,52 @@ package IkiWiki::Plugin::search;
 
 use warnings;
 use strict;
 
 use warnings;
 use strict;

-use IkiWiki 2.00;
+use IkiWiki 3.00;

 
 

-sub import { #{{{
+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 => "checkconfig", id => "search", call => \&checkconfig);
        hook(type => "pagetemplate", id => "search", call => \&pagetemplate);

-       hook(type => "sanitize", id => "search", call => \&index);
+       hook(type => "indexhtml", id => "search", call => \&indexhtml);

        hook(type => "delete", id => "search", call => \&delete);
        hook(type => "cgi", id => "search", call => \&cgi);
        hook(type => "delete", id => "search", call => \&delete);
        hook(type => "cgi", id => "search", call => \&cgi);

-} # }}}
-
-sub checkconfig () { #{{{
+       hook(type => "disable", id => "search", call => \&disable);
+}
+
+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,
+               },
+}
+
+sub checkconfig () {

        foreach my $required (qw(url cgiurl)) {
                if (! length $config{$required}) {
        foreach my $required (qw(url cgiurl)) {
                if (! length $config{$required}) {

-                       error(sprintf(gettext("Must specify %s when using the search plugin"), $required));
+                       error(sprintf(gettext("Must specify %s when using the %s plugin"), $required, 'search'));

                }
        }
        
                }
        }
        

-       if (! exists $config{omega_cgi}) {
+       if (! defined $config{omega_cgi}) {

                $config{omega_cgi}="/usr/lib/cgi-bin/omega/omega";
        }
                $config{omega_cgi}="/usr/lib/cgi-bin/omega/omega";
        }

-} #}}}
+
+       # This is a mass dependency, so if the search form template
+       # changes, every page is rebuilt.
+       add_depends("", "templates/searchform.tmpl");
+}

 
 my $form;
 
 my $form;

-sub pagetemplate (@) { #{{{
+sub pagetemplate (@) {

        my %params=@_;
        my $page=$params{page};
        my $template=$params{template};
        my %params=@_;
        my $page=$params{page};
        my $template=$params{template};


@@ -37,29 +59,28 @@ sub pagetemplate (@) { #{{{
                if (! defined $form) {
                        my $searchform = template("searchform.tmpl", blind_cache => 1);
                        $searchform->param(searchaction => $config{cgiurl});
                if (! defined $form) {
                        my $searchform = template("searchform.tmpl", blind_cache => 1);
                        $searchform->param(searchaction => $config{cgiurl});

+                       $searchform->param(html5 => $config{html5});

                        $form=$searchform->output;
                }
 
                $template->param(searchform => $form);
        }
                        $form=$searchform->output;
                }
 
                $template->param(searchform => $form);
        }

-} #}}}
+}

 
 my $scrubber;
 my $stemmer;
 
 my $scrubber;
 my $stemmer;

-sub index (@) { #{{{
+sub indexhtml (@) {

        my %params=@_;
        my %params=@_;

-       
-       return $params{content} if $IkiWiki::preprocessing{$params{destpage}};

 
        setupfiles();
 
        # A unique pageterm is used to identify the document for a page.
        my $pageterm=pageterm($params{page});
 
        setupfiles();
 
        # A unique pageterm is used to identify the document for a page.
        my $pageterm=pageterm($params{page});

-       return $params{content} unless defined $pageterm;
+       return unless defined $pageterm;

        
        my $db=xapiandb();
        my $doc=Search::Xapian::Document->new();
        
        my $db=xapiandb();
        my $doc=Search::Xapian::Document->new();

-       my $caption=IkiWiki::pagetitle($params{page});
+       my $caption=pagetitle($params{page});

        my $title;
        if (exists $pagestate{$params{page}}{meta} &&
                exists $pagestate{$params{page}}{meta}{title}) {
        my $title;
        if (exists $pagestate{$params{page}}{meta} &&
                exists $pagestate{$params{page}}{meta}{title}) {


@@ -92,11 +113,17 @@ sub index (@) { #{{{
        }
        $sample=~s/\n/ /g;
        
        }
        $sample=~s/\n/ /g;
        

+       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};
        # data used by omega
        # Decode html entities in it, since omega re-encodes them.
        eval q{use HTML::Entities};

+       error $@ if $@;

        $doc->set_data(
        $doc->set_data(

-               "url=".urlto($params{page}, "")."\n".
+               "url=".$url."\n".

                "sample=".decode_entities($sample)."\n".
                "caption=".decode_entities($caption)."\n".
                "modtime=$IkiWiki::pagemtime{$params{page}}\n".
                "sample=".decode_entities($sample)."\n".
                "caption=".decode_entities($caption)."\n".
                "modtime=$IkiWiki::pagemtime{$params{page}}\n".


@@ -108,8 +135,14 @@ sub index (@) { #{{{
        if (! $stemmer) {
                my $langcode=$ENV{LANG} || "en";
                $langcode=~s/_.*//;
        if (! $stemmer) {
                my $langcode=$ENV{LANG} || "en";
                $langcode=~s/_.*//;

-               eval { $stemmer=Search::Xapian::Stem->new($langcode) };
-               if ($@) {
+
+               # 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);
+               }
+               else {

                        $stemmer=Search::Xapian::Stem->new("english");
                }
        }
                        $stemmer=Search::Xapian::Stem->new("english");
                }
        }


@@ -119,26 +152,24 @@ sub index (@) { #{{{
        $tg->index_text($caption, 2);
        $tg->index_text($title, 2) if $title ne $caption;
        $tg->index_text($toindex);
        $tg->index_text($caption, 2);
        $tg->index_text($title, 2) if $title ne $caption;
        $tg->index_text($toindex);

-       $tg->index_text(lc($title), 1, "ZS"); # for title:foo
+       $tg->index_text(lc($title), 1, "S"); # for title:foo

        foreach my $link (@{$links{$params{page}}}) {
        foreach my $link (@{$links{$params{page}}}) {

-               $tg->index_text(lc($link), 1, "ZLINK"); # for link:bar
+               $tg->index_text(lc($link), 1, "XLINK"); # for link:bar

        }
 
        $doc->add_term($pageterm);
        $db->replace_document_by_term($pageterm, $doc);
        }
 
        $doc->add_term($pageterm);
        $db->replace_document_by_term($pageterm, $doc);

+}

 
 

-       return $params{content};
-} #}}}
-
-sub delete (@) { #{{{
+sub delete (@) {

        my $db=xapiandb();
        foreach my $page (@_) {
                my $pageterm=pageterm(pagename($page));
                $db->delete_document_by_term($pageterm) if defined $pageterm;
        }
        my $db=xapiandb();
        foreach my $page (@_) {
                my $pageterm=pageterm(pagename($page));
                $db->delete_document_by_term($pageterm) if defined $pageterm;
        }

-} #}}}
+}

 
 

-sub cgi ($) { #{{{
+sub cgi ($) {

        my $cgi=shift;
 
        if (defined $cgi->param('P')) {
        my $cgi=shift;
 
        if (defined $cgi->param('P')) {


@@ -151,31 +182,31 @@ sub cgi ($) { #{{{
                        noimageinline => 1, linktext => "Help");
                exec($config{omega_cgi}) || error("$config{omega_cgi} failed: $!");
        }
                        noimageinline => 1, linktext => "Help");
                exec($config{omega_cgi}) || error("$config{omega_cgi} failed: $!");
        }

-} #}}}
+}

 
 

-sub pageterm ($) { #{{{
+sub pageterm ($) {

        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.
        if (length $page > 240) {
        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.
        if (length $page > 240) {

-               eval q{use Digest::SHA1};
+               eval q{use Digest::SHA};

                if ($@) {
                if ($@) {

-                       debug("search: ".sprintf(gettext("need Digest::SHA1 to index %s"), $page)) 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 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::SHA1::sha1_hex($page));
+               return "U".lc(Digest::SHA::sha1_hex($page));

        }
        else {
                return "U:".$page;
        }
        }
        else {
                return "U:".$page;
        }

-} #}}}
+}

 
 my $db;
 
 my $db;

-sub xapiandb () { #{{{
+sub xapiandb () {

        if (! defined $db) {
                eval q{
                        use Search::Xapian;
        if (! defined $db) {
                eval q{
                        use Search::Xapian;


@@ -186,17 +217,39 @@ sub xapiandb () { #{{{
                        Search::Xapian::DB_CREATE_OR_OPEN());
        }
        return $db;
                        Search::Xapian::DB_CREATE_OR_OPEN());
        }
        return $db;

-} #}}}
+}

 
 

-sub setupfiles () { #{{{
-       if (! -e $config{wikistatedir}."/xapian" || $config{rebuild}) {
+{
+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");
                writefile("omega.conf", $config{wikistatedir}."/xapian",
                        "database_dir .\n".
                        "template_dir ./templates\n");

+               
+               # Avoid omega interpreting anything in the misctemplate
+               # as an omegascript command.
+               my $misctemplate=IkiWiki::misctemplate(gettext("search"), "\0",
+                       searchform => "", # avoid showing the small search form
+               );
+               eval q{use HTML::Entities};
+               error $@ if $@;
+               $misctemplate=encode_entities($misctemplate, '\$');
+
+               my $querytemplate=readfile(IkiWiki::template_file("searchquery.tmpl"));
+               $misctemplate=~s/\0/$querytemplate/;
+

                writefile("query", $config{wikistatedir}."/xapian/templates",
                writefile("query", $config{wikistatedir}."/xapian/templates",

-                       IkiWiki::misctemplate(gettext("search"),
-                               readfile(IkiWiki::template_file("searchquery.tmpl"))));
+                       $misctemplate);
+               $setup=1;
+       }
+}
+}
+
+sub disable () {
+       if (-d $config{wikistatedir}."/xapian") {
+               system("rm", "-rf", $config{wikistatedir}."/xapian");

        }
        }

-} #}}}
+}

 
 1
 
 1