This simple plugin allows you to insert an alphabetical index of a specified namespace in your pages. It should be useful in DokuWiki sites where pages are organized by namespaces.

It is fully inspired and based on the indexmenu plugin by Samuele Tognini samuele [at] samuele [dot] netsons [dot] org.

• Thanks to Samuele Tognini samuele [at] samuele [dot] netsons [dot] org, you can try the plugin here.
• Tested with 2006-03-09 DokuWiki version.
• ACL works, only authorized pages will be shown.
• useheading option works.
• Tested with some languages based on other characters than roman ones (Japanese, Chinese, Korean, Arabic, Russian, Thai).
• A big thanks to Samuele Tognini samuele [at] samuele [dot] netsons [dot] org for his plugin.
• Feel free to improve it.
• Feel free to correct my poor English

{{alphaindex>[namespace][#n][|nons]}}

Example that displays only pages inside wiki:plugins and lower namespaces (max two levels):

{{alphaindex>wiki:plugins#2|nons}}

example: {{alphaindex>.#2|nons}} shows all files with the depth of 2; the namespaces are not shown

All the data displayed by the alphaindex plugin are inside a div (called “alphaindex”).

You can customize each part of the generated code.

For example if you have defined the begin_letter_tpl as this :

==== {{letter}} ====

the plugin will generate the following HTML code :

<h3><a name="a" id="a">{{letter}}</a></h3>

and you can create the following CSS rule (for example) :

#alphaindex h3{
    display: block;
    background-color: #dee7ec;
    padding: 2px 0 2px 10px;
    margin: 20px 0 5px 0;
  }

Text display in block letter title for the page title beginning with a numerical value.

Optional
Default : '0-9'
Example : 'Numerical'

List of article treated by the plugin. Each article is separated with a '|'.

Optional
Default : (empty)
Example : ^the +|^a +|^an + or ^le +|^la +|^les +|^l\' * (in french)

With the example above, the page 'The Rolling Stones' will be treated as 'Rolling Stones' instead of its real name.

Management of the articles when the article_deletion option is active.

Optional
Default : 1
Authorized values :

• 0 : No change
• 1 : The article is move to the end of the page name inside brackets
• 2 or more : The article is deleted

Example: with the pagename 'The Rolling Stones'

• 0 : 'The Rolling Stones'
• 1 : 'Rolling Stones (The)'
• 2 : 'Rolling Stones'

Message shown in your page when the index is empty.
Note : {{ns}} is an alias for the requested namespace.

Optional
Default : 'No index for <b>{{ns}}</b>'

Template (in DokuWiki syntax) for the index title.

Optional
Default : '===== Index ====='

Template (in DokuWiki syntax) at the beginning of each letter's block.

Note : the {{letter}} pattern is mandatory. It's an alias for the current letter.

Optional
Default : '==== {{letter}} ===='

Template (in DokuWiki syntax) for each letter's entry.
Note : the {{link}} and {{name}} patterns are mandatory.They are aliases for the page link and the page name.

Optional
Default : ' * [[{{link}}|{{name}}]]'

Template (in DokuWiki syntax) at the end of each letter's block.

Optional
Default : (empty)

Name of pages not displayed in the index. Each page name is separated with a '|'.

Optional
Default : (empty)
Example : 'start|sidebar'

There is also a zip updated version with patches by community users.

Copy the Github link and paste it into the plugin_manager.

• 2006-06-14 - version 1.2
  • Fixed bug in namespaces display (lack of first letter for some namespaces).
  • Use of utf8 functions instead of PHP ones

• 2006-05-22 - version 1.1
  • Deletion of a dependence to indexmenu plugin

• 2006-05-19 - version 1.0
  • First release

* To use the UTF8 functions instead of classic PHP ones for characters and strings management.

* What's about an update to do Configuration via the Admin-panel?

* And what's about a RegEx for hidepages?

Please report bugs or feature requests at the Github.com - Bug tracker.

• Some random title indexes are displayed incorrectly, they lacks the first letter (i.e: okuWiki instead of DokuWiki).

— Samuele Tognini 2006-05-25 03:08

Corrected in 1.2 version — Takashi 2006-06-14 10:16

• It's need to replace all ucfirst calls to UTF8-safe code. For example, utf8_strtoupper(utf8_substr($string, 0, 1)).utf8_substr($string, 1). Otherwise non-Latin page titles (Cyrillic in my case) displays corrupted. Sorry for my bad English.

— David Mzareulyan 2007-03-08 18:30

The following patches are included in the alphaindex modded version in the how_to_install section.

2008-08-31

There is a patch which :

• Fix odd bug by replacing utf8_str_replace by str_replace by Juice.
• Fix the match(12 to match(13 issue rise by Thanos.
• Add the @NS@ support from Thanos.
• Apply the patch for tag/pagelist issue with useheading option, by Nicolas.

And…

• Fix a bug with hidepages and useheading option.
• Fix the TOC. Just use ~~NOTOC~~ to hide it.
• Hide editsection buttons, useless.
• Should not required anymore ~~NOCACHE~~ within the pages. I test it, that seems to work…
• I also did spaces cleaning.

I test it on the release 2008-05-05.

diff -Naur alphaindex_orig/syntax.php alphaindex/syntax.php
--- alphaindex_orig/syntax.php	2006-06-14 10:30:44.000000000 +0200
+++ alphaindex/syntax.php	2008-08-31 19:44:35.000000000 +0200
@@ -1,21 +1,21 @@
 <?php
 /**
- * Info Alphaindex: Displays the alphabetical index of a specified namespace. 
+ * Info Alphaindex: Displays the alphabetical index of a specified namespace.
  *
- * Version: 1.2
- * last modified: 2006-06-14 12:00:00
+ * Version: 1.3
+ * last modified: 2008-08-31
  * @license     GPL 2 (http://www.gnu.org/licenses/gpl.html)
- * @author      Hubert Moli?re  <hubert.moliere@alternet.fr>
+ * @author      Hubert Moli?re  <hubert.moliere@alternet.fr>
  */
- 
+
 /**
- * Configuration additionnelle sp?cifique au plugin
+ * Configuration additionnelle sp?cifique au plugin
  *
  * - $conf['plugin']['alphaindex']['numerical_index']
  * - $conf['plugin']['alphaindex']['articles_deletion']
- * - $conf['plugin']['alphaindex']['articles_moving'] 
+ * - $conf['plugin']['alphaindex']['articles_moving']
  * - $conf['plugin']['alphaindex']['empty_msg']
- * - $conf['plugin']['alphaindex']['title_tpl'] 
+ * - $conf['plugin']['alphaindex']['title_tpl']
  * - $conf['plugin']['alphaindex']['begin_letter_tpl']
  * - $conf['plugin']['alphaindex']['entry_tpl']
  * - $conf['plugin']['alphaindex']['end_letter_tpl']
@@ -27,13 +27,13 @@
 if(!defined('DOKU_PLUGIN')) define('DOKU_PLUGIN',DOKU_INC.'lib/plugins/');
 require_once(DOKU_PLUGIN.'syntax.php');
 require_once(DOKU_INC.'inc/search.php');
- 
+
 /**
  * All DokuWiki plugins to extend the parser/rendering mechanism
  * need to inherit from this class
  */
 class syntax_plugin_alphaindex extends DokuWiki_Syntax_Plugin {
-  
+
     /**
      * return some info
      */
@@ -41,61 +41,88 @@
         return array(
             'author' => 'Hubert Moliere',
             'email'  => 'takashi@natural-design.net',
-            'date'   => '2006-06-14',
+            'date'   => '2008-08-31',
             'name'   => 'Alphaindex',
             'desc'   => 'Insert the alphabetical index of a specified namespace.',
             'url'    => 'http://wiki.splitbrain.org/plugin:alphaindex'
             );
     }
-  
+
     function getType(){ return 'substition';}
     function getAllowedTypes() { return array('baseonly', 'substition', 'formatting', 'paragraphs', 'protected'); }
-  
+
     /**
      * Where to sort in?
      */
     function getSort(){
         return 139;
     }
- 
+
     /**
      * Connect pattern to lexer
      */
     function connectTo($mode) {
         $this->Lexer->addSpecialPattern('{{alphaindex>.+?}}',$mode,'plugin_alphaindex');
     }
-      
+
   /**
    * Handle the match
    */
     function handle($match, $state, $pos, &$handler){
+        global $ID;
+
         $level = 0;
         $nons = true;
-        $match = substr($match,12,-2);
+        $match = substr($match,13,-2);
         //split namespaces
         $match = preg_split('/\|/u', $match, 2);
         //split level
         $ns_opt = preg_split('/\#/u', $match[0], 2);
         //namespace name
         $ns = $ns_opt[0];
+        // add @NS@ option
+        if(empty($ns) || $ns == '@NS@') {
+         $pos = strrpos($ID,':');
+          if($pos!=false){
+            $ns = substr($ID,0,$pos);
+          }else{
+             $ns = '.';
+          }
+        }
         //level;
         if (is_numeric($ns_opt[1])) {
             $level=$ns_opt[1];
         }
         $match = explode(" ", $match[1]);
-        // namespaces option 
+        // namespaces option
         $nons = in_array('nons', $match);
         // multi-columns option
         $incol = in_array('incol', $match);
         return array($ns,array('level' => $level,'nons' => $nons,'incol' => $incol));
     }
- 
+
     /**
      * Render output
      */
     function render($mode, &$renderer, $data) {
+        global $ID;
         global $conf;
-        if($mode == 'xhtml'){ 
+
+        if($mode == 'xhtml'){
+            // Never cache to get fresh info
+            $renderer->info['cache'] = false;
+            // Remove Section Edit Buttons
+            $oldmaxecl = $conf['maxseclevel'];
+            $conf['maxseclevel'] = 0;
+            // To see the TOC, to not use ~~NOTOC~~
+            if ($renderer->info['toc']) {
+              $oldmaxtoc = $conf['maxtoclevel'];
+              $conf['maxtoclevel'] = 3;
+            }
+            else {
+              $oldmaxtoc = -1;
+            }
+
             $alpha_data = $this->_alpha_index($data, $renderer);
             if ((!@$n)) {
                 if (isset ($conf['plugin']['alphaindex']['empty_msg'])) {
@@ -106,31 +133,28 @@
             }
             $alpha_data = str_replace('{{ns}}',cleanID($data[0]),$alpha_data);
 
-            $alpha_data = p_render('xhtml', p_get_instructions($alpha_data), $info); 
-            
-            // remove toc, section edit buttons and category tags
-            $patterns = array('!<div class="toc">.*?(</div>\n</div>)!s',
-                            '#<!-- SECTION \[(\d*-\d*)\] -->#e',
-                            '!<div class="category">.*?</div>!s');
-            $replace  = array('','','');
-            $alpha_data = preg_replace($patterns, $replace, $alpha_data);      
+            $alpha_data = p_render('xhtml', p_get_instructions($alpha_data), $info);
+
             $renderer->doc .= '<div id="alphaindex">' ;
             $renderer->doc .= $ns_data;
             $renderer->doc .= '<hr />';
             $renderer->doc .= $alpha_data;
             $renderer->doc .= '</div>' ;
+
+            $conf['maxseclevel'] = $oldmaxecl;
+            if ( $oldmaxtoc == -1 ) $conf['maxtoclevel'] = $oldmaxtoc;
             return true;
         }
 
 
 
 
 
 
         return false;
     }
 
     /**
-     * Return the alphabetical index 
+     * Return the alphabetical index
      * @author Hubert MOLIERE <hubert.moliere@alternet.fr>
      *
      * This function is a hack of Indexmenu _tree_menu($ns)
      * @author Samuele Tognini <samuele@samuele.netsons.org>
-     * 
+     *
      * This function is a simple hack of DokuWiki html_index($ns)
      * @author Andreas Gohr <andi@splitbrain.org>
      */
@@ -138,9 +162,9 @@
         global $conf;
         global $ID;
 
-        $ns = $myns[0];                                                
+        $ns = $myns[0];
         $opts = $myns[1];
-    
+
         // Articles deletion configuration
         $articlesDeletionPatterns = array();
         if(isset($conf['plugin']['alphaindex']['articles_deletion'])) {
@@ -149,13 +173,13 @@
         } else {
             $articlesDeletion = false;
         }
-        
+
         // Hide pages configuration
         $hidepages = array();
         if(isset($conf['plugin']['alphaindex']['hidepages'])) {
             $hidepages = explode('|', $conf['plugin']['alphaindex']['hidepages']);
         }
-    
+
         // template configuration
         if(isset($conf['plugin']['alphaindex']['title_tpl'])) {
             $titleTpl = $conf['plugin']['alphaindex']['title_tpl'];
@@ -177,25 +201,25 @@
         } else {
             $endLetterTpl = '';
         }
-        
+
         if($ns == '.') {
           $ns = dirname(str_replace(':','/',$ID));
           if ($ns == '.') $ns = '';
         } else {
           $ns = cleanID($ns);
         }
-        
+
         $ns  = utf8_encodeFN(str_replace(':','/',$ns));
         $data = array();
         search($data,$conf['datadir'],'alphaindex_search_index',$opts,"/".$ns);
-        
+
         $nb_data = count($data);
         $alpha_data = array();
-    
+
         // alphabetical ordering
         for($cpt=0; $cpt<$nb_data; $cpt++) {
             $tmpData = $data[$cpt]['id'];
-            
+
             $pos = strrpos(utf8_decode($tmpData), ':');
             if($conf['useheading']) {
                 $pageName = p_get_first_heading($tmpData);
@@ -205,7 +229,7 @@
                     } else {
                         $pageName = $tmpData;
                     }
-                    $pageName = utf8_str_replace('_', ' ', $pageName);
+                    $pageName = str_replace('_', ' ', $pageName);
                 }
             } else {
                 if($pos != FALSE) {
@@ -213,14 +237,21 @@
                 } else {
                     $pageName = $tmpData;
                 }
-                $pageName = utf8_str_replace('_', ' ', $pageName);
+                $pageName = str_replace('_', ' ', $pageName);
             }
             $pageNameArticle = '';
-            
-            // if the current page is not a page to hide 
-            if(!in_array($pageName, $hidepages)) { 
+
+            // Otherwise,hhidepages doesn't work with useheading
+            if($pos != FALSE) {
+              $pageNameNS = utf8_substr($tmpData, $pos+1, utf8_strlen($tmpData));
+            } else {
+              $pageNameNS = $tmpData;
+            }
+
+            // if the current page is not a page to hide
+            if(!in_array($pageNameNS, $hidepages)) {
                 // Articles deletion
-                if($articlesDeletion) { 
+                if($articlesDeletion) {
                     foreach($articlesDeletionPatterns as $pattern) {
                         if(eregi($pattern, $pageName, $result)) {
                             $pageName = eregi_replace($pattern, '', $pageName);
@@ -228,7 +259,14 @@
                         }
                     }
                 }
-                // R?cup?ration de la premi?re lettre du mot et classement 
+
+                if(isset($conf['plugin']['alphaindex']['metadata_title'])) {
+                  $tmp = p_get_metadata($data[$cpt]['id']);
+                  if(isset($tmp['title']))
+                    $pageName = $tmp['title'];
+                }
+
+                // R?cup?ration de la premi?re lettre du mot et classement
                 $firstLetter = utf8_deaccent(utf8_strtolower(utf8_substr($pageName, 0, 1)));
                 if(is_numeric($firstLetter)) {
                     if(isset($conf['plugin']['alphaindex']['numerical_index'])) {
@@ -237,7 +275,7 @@
                         $firstLetter = '0-9';
                     }
                 }
-                
+
                 if(isset($conf['plugin']['alphaindex']['article_moving'])) {
                     $articleMoving = $conf['plugin']['alphaindex']['article_moving'];
                 } else {
@@ -248,38 +286,38 @@
                 } else if (($articleMoving == 1)&&($pageNameArticle != '')) {
                     $pageName = $pageName.' ('.$pageNameArticle.')';
                 }
-        
+
                 $data[$cpt]['id2'] = ucfirst($pageName);
                 $alpha_data[$firstLetter][] = $data[$cpt];
             }
         }
-    
+
         // array sorting by key
         ksort($alpha_data);
-    
+
         // Display of results
-    
+
         // alphabetical index
         $alphaOutput .= $titleTpl."\n";
         $nb_data = count($alpha_data);
         foreach($alpha_data as $key => $currentLetter) {
             // Sorting of $currentLetter array
             usort($currentLetter, create_function('$a, $b', "return strnatcasecmp(\$a['id2'], \$b['id2']);"));
-            
-            $begin = utf8_str_replace("{{letter}}" ,utf8_strtoupper($key), $beginLetterTpl);
+
+            $begin = str_replace("{{letter}}" ,utf8_strtoupper($key), $beginLetterTpl);
             $alphaOutput .= $begin."\n";
             foreach($currentLetter as $currentLetterEntry) {
-                $link = utf8_str_replace("{{link}}" ,$currentLetterEntry['id'], $entryTpl);
-                $alphaOutput .= utf8_str_replace("{{name}}" ,$currentLetterEntry['id2'], $link);
+                $link = str_replace("{{link}}" ,$currentLetterEntry['id'], $entryTpl);
+                $alphaOutput .= str_replace("{{name}}" ,$currentLetterEntry['id2'], $link);
                 $alphaOutput .= "\n";
             }
-            $end = utf8_str_replace("{{letter}}" ,utf8_strtoupper($key), $endLetterTpl); 
+            $end = str_replace("{{letter}}" ,utf8_strtoupper($key), $endLetterTpl);
             $alphaOutput .= $end."\n";
         }
-        
+
         return $alphaOutput;
     }
-    
+
 } //Alphaindex class end
 
   /**
@@ -292,9 +330,9 @@
    */
 function alphaindex_search_index(&$data,$base,$file,$type,$lvl,$opts){
   $return = true;
- 
+
   $item = array();
- 
+
   if($type == 'd'){
     if ($opts['level'] == $lvl) $return=false;
     if ($opts['nons']) return $return;
@@ -302,24 +340,24 @@
     //don't add
     return false;
   }
- 
+
   $id = pathID($file);
- 
+
   //check hidden
   if($type=='f' && isHiddenPage($id)){
     return false;
   }
- 
+
   //check ACL
   if($type=='f' && auth_quickaclcheck($id) < AUTH_READ){
     return false;
   }
- 
+
   //Set all pages at first level
   if ($opts['nons']) {
-    $lvl=1;    
+    $lvl=1;
   }
- 
+
   $data[]=array( 'id'    => $id,
 		 'type'  => $type,
 		 'level' => $lvl,
@@ -327,4 +365,4 @@
   return $return;
 }
 
-?>
\ Pas de fin de ligne ? la fin du fichier.
+?>
diff -Naur alphaindex_orig/VERSION alphaindex/VERSION
--- alphaindex_orig/VERSION	1970-01-01 01:00:00.000000000 +0100
+++ alphaindex/VERSION	2008-08-31 18:32:47.000000000 +0200
@@ -0,0 +1 @@
+20080831

– Jonesy, have fun !

Since it seems that this plugin is abandoned by the original author, I've built an alphaindex darcs containing your patches (I had to manually correct some hunks) . You can fetch it with:

darcs get http://samuele.netsons.org/darcs/alphaindex

I'll apply any new working patch published here, sent with the darcs mechanism or created by me.
Next step could be to implement a better cache mechanism according for example to the indexmenu action plugin.
It's also possible to install it via plugin manager using this link.

This version only has 1 error. When activate the articles_moving when the article_deletion option is active, look at this example:

$conf['articles_deletion'] = '^the +|^a +|^an +'
$conf['articles_moving']   = 1
$conf['metadata_title']    = 1

If I have a page which header is 'The Rolling Stones', this page will be treated as 'Rolling Stones' instead of its real name. But the page remains in the letter “T”, and not in the letter “R”.

I see in the code and I found 2 minimal errors:
Line 221

if($this->getConf('metadata_title')) {
	  $tmp = p_get_metadata($data[$cpt]['id']);
	  if(isset($tmp['title']))
	    $pageName = $tmp['title'];
	}

Move the entire block, just before this lines (line 211)

PUT THE ENTIRE BLOCK JUST HERE
// Articles deletion
	if(is_array($articlesDeletionPatterns)) {

This will correct the letter index, so 'The Rolling Stones' will place in the letter “R”

It doesn't seem to update when new pages are created. I assume it's a caching problem, but I don't know enough about DokuWiki's caching to edit my maintenance script and fix it. Can anyone suggest a fix? Thanks! — Matt Cooper 2007-05-06 18:00

Yes, it is. Better solution is to implement the caching fix into alphaindex code (you should take look at developer_note for info and to some plugins examples, like indexmenu action plugin).
Or use the DokuWiki internal core, this is from indexmenu page:

To purge cache either wait for the automatic purge cache time expiration or force it through a purge=true request on the page that contains the indexmenu tree.

You also might try the ~~NOCACHE~~ instruction to completely disable a specified page cache, so that you have no more to worry about purging cache. But pay attention that it could slow down the overall loading process of the page depending on the number of tree nodes.

I've installed it in my site, so that, if the author doesn't find a better way, you can test this useful plugin here. — Samuele Tognini 2006-05-25 03:58

I needed listing of namespaces with index pages only, so I make a new option indexes (goes instead of nons) which causes the plugin to generate listing from starting pages only (those are conf['start'] and page with same name as namespace). I'm attaching very simple patch. Not fully tested, just a quick hack. — Tomáš Valenta 2006-11-18 18:35

73a74
>       $indexes = false;
90c91,93
<         return array($ns,array('level' => $level,'nons' => $nons,'incol' => $incol));
---
>       // indexes option
>       $indexes = in_array('indexes', $match);
>         return array($ns,array('level' => $level,'nons' => $nons,'incol' => $incol,'indexes' => $indexes));
293a297,298
>   global $conf;
> 
307c312,317
<  
---
>   // show just index files
>   if ($opts['indexes']) {
>     $myns = explode(":",getNS($id));
>     if (end($myns)!=noNS($id) and noNS($id)!=$conf['start']) return true;
>   }
> 

Somehow the plugin list lots of red links in my installation, might this be because it supposes that there is always a file ns1:ns2 present if there is content in the namespage ns1:ns2 (e.g. ns1:ns2:home)? If that is the case, this plugin does not support wikis that make use of default namespace linking, which is a pity. – regards, bernhard

I would like to suggest a feature that prints the namespace beside the page if there are several pages with the same name, e.g.

> Files, (foo/)
> Files (The), (bar/)
> Files, (foobar/)

or with : too, of course

Edit: maybe even an argument to print each namespace (alphabetically sorted) own its own above the root namespace:

> ====== /Files ======
> ====== /Wiki ======
> (Root)

Hello, I use the alphaindex on http://jurtenrunde.de/lieder/doku.php?id=uebersicht:uebersicht/. I have a Namespace structure which looks like this:

liedtexte:t:testlied , liedtexte:b:test2 and so on. I get the index of this way:

{{alphaindex>[liedtexte#10][#n][|nons]}}

. You can see on http://jurtenrunde.de/lieder/doku.php?id=uebersicht:uebersicht/ that it posts pages where they do not exist - I would like it to show only existing pages. How can I solve this problem?

Is there a way to override global 'title_tpl' for some particular index? — Stoyan 2008-01-06

I am wondering if there is a way to make this plugin work with the XBR render plugin? What it does is repeat the index twice and omitting any other text in the site. Anyone have ideas?