includes/Categoryfinder.php

   1 <?php
   2 /**
   3  * Recent changes filtering by category.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  */
  22
  23 /**
  24  * The "Categoryfinder" class takes a list of articles, creates an internal
  25  * representation of all their parent categories (as well as parents of
  26  * parents etc.). From this representation, it determines which of these
  27  * articles are in one or all of a given subset of categories.
  28  *
  29  * Example use :
  30  * <code>
  31  *     # Determines whether the article with the page_id 12345 is in both
  32  *     # "Category 1" and "Category 2" or their subcategories, respectively
  33  *
  34  *     $cf = new Categoryfinder;
  35  *     $cf->seed(
  36  *         array( 12345 ),
  37  *         array( 'Category 1', 'Category 2' ),
  38  *         'AND'
  39  *     );
  40  *     $a = $cf->run();
  41  *     print implode( ',' , $a );
  42  * </code>
  43  *
  44  */
  45 class Categoryfinder {
  46         var $articles = array(); # The original article IDs passed to the seed function
  47         var $deadend = array(); # Array of DBKEY category names for categories that don't have a page
  48         var $parents = array(); # Array of [ID => array()]
  49         var $next = array(); # Array of article/category IDs
  50         var $targets = array(); # Array of DBKEY category names
  51         var $name2id = array();
  52         var $mode; # "AND" or "OR"
  53
  54         /**
  55          * @var DatabaseBase
  56          */
  57         var $dbr; # Read-DB slave
  58
  59         /**
  60          * Constructor (currently empty).
  61          */
  62         function __construct() {
  63         }
  64
  65         /**
  66          * Initializes the instance. Do this prior to calling run().
  67          * @param $article_ids Array of article IDs
  68          * @param $categories FIXME
  69          * @param string $mode FIXME, default 'AND'.
  70          * @todo FIXME: $categories/$mode
  71          */
  72         function seed( $article_ids, $categories, $mode = 'AND' ) {
  73                 $this->articles = $article_ids;
  74                 $this->next = $article_ids;
  75                 $this->mode = $mode;
  76
  77                 # Set the list of target categories; convert them to DBKEY form first
  78                 $this->targets = array();
  79                 foreach ( $categories as $c ) {
  80                         $ct = Title::makeTitleSafe( NS_CATEGORY, $c );
  81                         if ( $ct ) {
  82                                 $c = $ct->getDBkey();
  83                                 $this->targets[$c] = $c;
  84                         }
  85                 }
  86         }
  87
  88         /**
  89          * Iterates through the parent tree starting with the seed values,
  90          * then checks the articles if they match the conditions
  91          * @return array of page_ids (those given to seed() that match the conditions)
  92          */
  93         function run() {
  94                 $this->dbr = wfGetDB( DB_SLAVE );
  95                 while ( count( $this->next ) > 0 ) {
  96                         $this->scan_next_layer();
  97                 }
  98
  99                 # Now check if this applies to the individual articles
 100                 $ret = array();
 101
 102                 foreach ( $this->articles as $article ) {
 103                         $conds = $this->targets;
 104                         if ( $this->check( $article, $conds ) ) {
 105                                 # Matches the conditions
 106                                 $ret[] = $article;
 107                         }
 108                 }
 109                 return $ret;
 110         }
 111
 112         /**
 113          * This functions recurses through the parent representation, trying to match the conditions
 114          * @param int $id The article/category to check
 115          * @param array $conds The array of categories to match
 116          * @param array $path used to check for recursion loops
 117          * @return bool Does this match the conditions?
 118          */
 119         function check( $id, &$conds, $path = array() ) {
 120                 // Check for loops and stop!
 121                 if ( in_array( $id, $path ) ) {
 122                         return false;
 123                 }
 124
 125                 $path[] = $id;
 126
 127                 # Shortcut (runtime paranoia): No conditions=all matched
 128                 if ( count( $conds ) == 0 ) {
 129                         return true;
 130                 }
 131
 132                 if ( !isset( $this->parents[$id] ) ) {
 133                         return false;
 134                 }
 135
 136                 # iterate through the parents
 137                 foreach ( $this->parents[$id] as $p ) {
 138                         $pname = $p->cl_to;
 139
 140                         # Is this a condition?
 141                         if ( isset( $conds[$pname] ) ) {
 142                                 # This key is in the category list!
 143                                 if ( $this->mode == 'OR' ) {
 144                                         # One found, that's enough!
 145                                         $conds = array();
 146                                         return true;
 147                                 } else {
 148                                         # Assuming "AND" as default
 149                                         unset( $conds[$pname] );
 150                                         if ( count( $conds ) == 0 ) {
 151                                                 # All conditions met, done
 152                                                 return true;
 153                                         }
 154                                 }
 155                         }
 156
 157                         # Not done yet, try sub-parents
 158                         if ( !isset( $this->name2id[$pname] ) ) {
 159                                 # No sub-parent
 160                                 continue;
 161                         }
 162                         $done = $this->check( $this->name2id[$pname], $conds, $path );
 163                         if ( $done || count( $conds ) == 0 ) {
 164                                 # Subparents have done it!
 165                                 return true;
 166                         }
 167                 }
 168                 return false;
 169         }
 170
 171         /**
 172          * Scans a "parent layer" of the articles/categories in $this->next
 173          */
 174         function scan_next_layer() {
 175                 wfProfileIn( __METHOD__ );
 176
 177                 # Find all parents of the article currently in $this->next
 178                 $layer = array();
 179                 $res = $this->dbr->select(
 180                         /* FROM   */ 'categorylinks',
 181                         /* SELECT */ '*',
 182                         /* WHERE  */ array( 'cl_from' => $this->next ),
 183                         __METHOD__ . '-1'
 184                 );
 185                 foreach ( $res as $o ) {
 186                         $k = $o->cl_to;
 187
 188                         # Update parent tree
 189                         if ( !isset( $this->parents[$o->cl_from] ) ) {
 190                                 $this->parents[$o->cl_from] = array();
 191                         }
 192                         $this->parents[$o->cl_from][$k] = $o;
 193
 194                         # Ignore those we already have
 195                         if ( in_array( $k, $this->deadend ) ) {
 196                                 continue;
 197                         }
 198
 199                         if ( isset( $this->name2id[$k] ) ) {
 200                                 continue;
 201                         }
 202
 203                         # Hey, new category!
 204                         $layer[$k] = $k;
 205                 }
 206
 207                 $this->next = array();
 208
 209                 # Find the IDs of all category pages in $layer, if they exist
 210                 if ( count( $layer ) > 0 ) {
 211                         $res = $this->dbr->select(
 212                                 /* FROM   */ 'page',
 213                                 /* SELECT */ array( 'page_id', 'page_title' ),
 214                                 /* WHERE  */ array( 'page_namespace' => NS_CATEGORY, 'page_title' => $layer ),
 215                                 __METHOD__ . '-2'
 216                         );
 217                         foreach ( $res as $o ) {
 218                                 $id = $o->page_id;
 219                                 $name = $o->page_title;
 220                                 $this->name2id[$name] = $id;
 221                                 $this->next[] = $id;
 222                                 unset( $layer[$name] );
 223                         }
 224                 }
 225
 226                 # Mark dead ends
 227                 foreach ( $layer as $v ) {
 228                         $this->deadend[$v] = $v;
 229                 }
 230
 231                 wfProfileOut( __METHOD__ );
 232         }
 233 }