pve-docs-mediawiki-import.in: include text version inside pvehide tag

So that wiki search returns useful content.

diff --git a/debian/control b/debian/control
index 46cd222ca4dac7ce1f3fd7b826da599d95612e5e..1f593c5f81802471b7fc48efdbf1b0b9d12070bd 100644 (file)
--- a/debian/control
+++ b/debian/control
@@ -23,7 +23,7 @@ Description: Proxmox VE Documentation
 Package: pve-docs-mediawiki
 Architecture: all
 Section: doc
-Depends: ${misc:Depends}, pve-docs
+Depends: ${misc:Depends}, pve-docs, libhtml-parser-perl
 Suggests: apache2
 Description: Proxmox VE Documentation - mediawiki plugin
  This package contains tools to view Proxmox VE Documentation with mediakiki.

diff --git a/pve-docs-mediawiki-import.in b/pve-docs-mediawiki-import.in
index 1a3f6b48bb974fd0f2a43878b0dd9ce8135181b4..851cbf9bdc6f031cafc7d7412e8ecbfc44fcf9d5 100755 (executable)
--- a/pve-docs-mediawiki-import.in
+++ b/pve-docs-mediawiki-import.in
@@ -7,6 +7,7 @@ use Data::Dumper;
 use IO::File;
 use File::Basename;
 use MediaWiki::API;
+use HTML::Parser;
 
 use JSON;
 
@@ -35,25 +36,61 @@ $mw->login({ lgname => $username, lgpassword => $passwd })
     || die $mw->{error}->{code} . ': ' . $mw->{error}->{details};
 
 sub update_page {
-    my ($pagename, $include, $category) = @_;
+    my ($pagename, $filename, $category) = @_;
 
     print "update mediawiki page: $pagename\n";
 
     my $ref = $mw->get_page( { title => $pagename } );
     my $page =  $ref->{'*'} || '';
 
-    if ($page !~ m/^\{\{#pvedocs:.*\}\}\s*$/m) {
-       $page = "{{#pvedocs:$include}}\n$page";
-    } else {
-       $page =~ s/^\{\{#pvedocs:.*\}\}\s*$/\{\{#pvedocs:$include\}\}\n/m;
-    }
+    my $pve_content = "<!-- Do not edit - this is autogenerated content -->\n";
+
+    $pve_content .= "{{#pvedocs:$filename}}\n";
+    $pve_content .= "[[Category:$category]]\n" if $category;
+
+    my $starttag = '<!--PVE_IMPORT_START_MARKER-->';
+    my $endtag = '<!--PVE_IMPORT_END_MARKER-->';
 
-    if ($category) {
-       my $catstr = "Category:$category";
+    $pve_content .= "<pvehide>\n";
 
-       if ($page !~ m/^\[\[$catstr\]\]\s*$/m) {
-           $page .= "\n[[$catstr]]\n";
-       }
+    my $parser_opts = {
+       api_version => 3,
+       text_h => [ sub { $pve_content .= shift }, "dtext" ],
+    };
+    my $parser = HTML::Parser->new(%$parser_opts);
+
+    my $fh = IO::File->new("/usr/share/pve-docs/$filename", "r") or
+       die "unable to open file '$filename' - $!\n";
+    while (defined(my $line = <$fh>)) {
+       $parser->parse($line);
+    }
+    $pve_content .= "</pvehide>\n";
+
+    $pve_content =~ s/\s+$//gm;
+
+    chomp $pve_content;
+
+    if ($page =~ m/^(.*)$starttag\n.*\n$endtag(.*)$/s) {
+       my ($top_content, $bottom_content) = ($1, $2);
+       chomp $top_content;
+       chomp $bottom_content;
+       $page = $top_content;
+       $page .= "$starttag\n";
+       $page .= $pve_content;
+       $page .= "\n$endtag\n";
+       $page .= $bottom_content;
+    } elsif ($page =~ m/(.*)\{\{#pvedocs:.*?\}\}(.*)$/) {
+       # old style
+       my ($top_content, $bottom_content) = ($1, $2);
+       chomp $top_content;
+       chomp $bottom_content;
+       $page = $top_content;
+       $page .= "$starttag\n";
+       $page .= $pve_content;
+       $page .= "\n$endtag\n";
+       $page .= $bottom_content;
+    } else {
+       $page = "$starttag\n$pve_content\n$endtag\n$page";
     }
 
     my $timestamp = $ref->{timestamp};