includes/api/ApiPageSet.php

   1 <?php
   2
   3
   4 /*
   5  * Created on Sep 24, 2006
   6  *
   7  * API for MediaWiki 1.8+
   8  *
   9  * Copyright (C) 2006 Yuri Astrakhan <FirstnameLastname@gmail.com>
  10  *
  11  * This program is free software; you can redistribute it and/or modify
  12  * it under the terms of the GNU General Public License as published by
  13  * the Free Software Foundation; either version 2 of the License, or
  14  * (at your option) any later version.
  15  *
  16  * This program is distributed in the hope that it will be useful,
  17  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  18  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  19  * GNU General Public License for more details.
  20  *
  21  * You should have received a copy of the GNU General Public License along
  22  * with this program; if not, write to the Free Software Foundation, Inc.,
  23  * 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
  24  * http://www.gnu.org/copyleft/gpl.html
  25  */
  26
  27 if (!defined('MEDIAWIKI')) {
  28         // Eclipse helper - will be ignored in production
  29         require_once ('ApiQueryBase.php');
  30 }
  31
  32 class ApiPageSet extends ApiQueryBase {
  33
  34         private $mAllPages; // [ns][dbkey] => page_id or 0 when missing
  35         private $mGoodTitles, $mMissingTitles, $mRedirectTitles, $mNormalizedTitles;
  36
  37         private $mRequestedFields;
  38
  39         public function __construct($query) {
  40                 parent :: __construct($query, __CLASS__);
  41
  42                 $this->mAllPages = array ();
  43                 $this->mGoodTitles = array ();
  44                 $this->mMissingTitles = array ();
  45                 $this->mRedirectTitles = array ();
  46                 $this->mNormalizedTitles = array ();
  47
  48                 $this->mRequestedFields = array ();
  49         }
  50
  51         public function requestField($fieldName) {
  52                 $this->mRequestedFields[$fieldName] = null;
  53         }
  54
  55         /**
  56          * Title objects that were found in the database.
  57          * @return array page_id (int) => Title (obj)
  58          */
  59         public function getGoodTitles() {
  60                 return $this->mGoodTitles;
  61         }
  62
  63         /**
  64          * Title objects that were NOT found in the database.
  65          * @return array of Title objects
  66          */
  67         public function getMissingTitles() {
  68                 return $this->mMissingTitles;
  69         }
  70
  71         /**
  72          * Get a list of redirects when doing redirect resolution
  73          * @return array prefixed_title (string) => prefixed_title (string)
  74          */
  75         public function getRedirectTitles() {
  76                 return $this->mRedirectTitles;
  77         }
  78
  79         /**
  80          * Get a list of title normalizations - maps the title given
  81          * with its normalized version.
  82          * @return array raw_prefixed_title (string) => prefixed_title (string)
  83          */
  84         public function getNormalizedTitles() {
  85                 return $this->mNormalizedTitles;
  86         }
  87
  88         /**
  89          * Returns the number of unique pages (not revisions) in the set.
  90          */
  91         public function getGoodTitleCount() {
  92                 return count($this->getGoodTitles());
  93         }
  94
  95         /**
  96          * Get the list of revision IDs (requested with revids= parameter)
  97          */
  98         public function getRevisionIDs() {
  99                 $this->dieUsage(__METHOD__ . ' is not implemented', 'notimplemented');
 100         }
 101
 102         /**
 103          * Returns the number of revisions (requested with revids= parameter)
 104          */
 105         public function getRevisionCount() {
 106                 return 0; // TODO: implement
 107         }
 108
 109         /**
 110          * This method populates internal variables with page information
 111          * based on the given array of title strings.
 112          *
 113          * Steps:
 114          * #1 For each title, get data from `page` table
 115          * #2 If page was not found in the DB, store it as missing
 116          *
 117          * Additionally, when resolving redirects:
 118          * #3 If no more redirects left, stop.
 119          * #4 For each redirect, get its links from `pagelinks` table.
 120          * #5 Substitute the original LinkBatch object with the new list
 121          * #6 Repeat from step #1
 122          */
 123         private function populateTitles($titles, $redirects) {
 124
 125                 // Ensure we get minimum required fields
 126                 $pageFlds = array (
 127                         'page_id' => null,
 128                         'page_namespace' => null,
 129                         'page_title' => null
 130                 );
 131
 132                 // only store non-default fields
 133                 $this->mRequestedFields = array_diff_key($this->mRequestedFields, $pageFlds);
 134
 135                 if ($redirects)
 136                         $pageFlds['page_is_redirect'] = null;
 137
 138                 $pageFlds = array_keys(array_merge($this->mRequestedFields, $pageFlds));
 139
 140                 // Get validated and normalized title objects
 141                 $linkBatch = $this->processTitlesStrArray($titles);
 142
 143                 $db = $this->getDB();
 144
 145                 //
 146                 // Repeat until all redirects have been resolved
 147                 //
 148                 while (false !== ($set = $linkBatch->constructSet('page', $db))) {
 149
 150                         // Hack: get the ns:titles stored in array(ns => array(titles)) format
 151                         $remaining = $linkBatch->data;
 152
 153                         $redirectIds = array ();
 154
 155                         //
 156                         // Get data about $linkBatch from `page` table
 157                         //
 158                         $this->profileDBIn();
 159                         $res = $db->select('page', $pageFlds, $set, __METHOD__);
 160                         $this->profileDBOut();
 161                         while ($row = $db->fetchObject($res)) {
 162
 163                                 unset ($remaining[$row->page_namespace][$row->page_title]);
 164                                 $title = Title :: makeTitle($row->page_namespace, $row->page_title);
 165                                 $this->mAllPages[$row->page_namespace][$row->page_title] = $row->page_id;
 166
 167                                 if ($redirects && $row->page_is_redirect == '1') {
 168                                         $redirectIds[$row->page_id] = $title;
 169                                 } else {
 170                                         $this->mGoodTitles[$row->page_id] = $title;
 171                                 }
 172                         }
 173                         $db->freeResult($res);
 174
 175                         //
 176                         // The remaining titles in $remaining are non-existant pages
 177                         //
 178                         foreach ($remaining as $ns => $dbkeys) {
 179                                 foreach ($dbkeys as $dbkey => $nothing) {
 180                                         $this->mMissingTitles[] = Title :: makeTitle($ns, $dbkey);
 181                                         $this->mAllPages[$ns][$dbkey] = 0;
 182                                 }
 183                         }
 184
 185                         if (!$redirects || empty ($redirectIds))
 186                                 break;
 187
 188                         //
 189                         // Resolve redirects by querying the pagelinks table, and repeat the process
 190                         //
 191
 192                         // Create a new linkBatch object for the next pass
 193                         $linkBatch = new LinkBatch();
 194
 195                         // find redirect targets for all redirect pages
 196                         $this->profileDBIn();
 197                         $res = $db->select('pagelinks', array (
 198                                 'pl_from',
 199                                 'pl_namespace',
 200                                 'pl_title'
 201                         ), array (
 202                                 'pl_from' => array_keys($redirectIds
 203                         )), __METHOD__);
 204                         $this->profileDBOut();
 205
 206                         while ($row = $db->fetchObject($res)) {
 207
 208                                 // Bug 7304 workaround
 209                                 // ( http://bugzilla.wikipedia.org/show_bug.cgi?id=7304 )
 210                                 // A redirect page may have more than one link.
 211                                 // This code will only use the first link returned.
 212                                 if (isset ($redirectIds[$row->pl_from])) { // remove line when 7304 is fixed
 213
 214                                         $titleStrFrom = $redirectIds[$row->pl_from]->getPrefixedText();
 215                                         $titleStrTo = Title :: makeTitle($row->pl_namespace, $row->pl_title)->getPrefixedText();
 216                                         $this->mRedirectTitles[$titleStrFrom] = $titleStrTo;
 217
 218                                         unset ($redirectIds[$row->pl_from]); // remove line when 7304 is fixed
 219
 220                                         // Avoid an infinite loop by checking if we have already processed this target
 221                                         if (!isset ($this->mAllPages[$row->pl_namespace][$row->pl_title])) {
 222                                                 $linkBatch->add($row->pl_namespace, $row->pl_title);
 223                                         }
 224                                 }
 225                         }
 226                         $db->freeResult($res);
 227                 }
 228         }
 229
 230         /**
 231          * Given an array of title strings, convert them into Title objects.
 232          * This method validates access rights for the title,
 233          * and appends normalization values to the output.
 234          *
 235          * @return LinkBatch of title objects.
 236          */
 237         private function processTitlesStrArray($titles) {
 238
 239                 $linkBatch = new LinkBatch();
 240
 241                 foreach ($titles as $titleString) {
 242                         $titleObj = Title :: newFromText($titleString);
 243
 244                         // Validation
 245                         if (!$titleObj)
 246                                 $this->dieUsage("bad title $titleString", 'invalidtitle');
 247                         if ($titleObj->getNamespace() < 0)
 248                                 $this->dieUsage("No support for special page $titleString has been implemented", 'unsupportednamespace');
 249                         if (!$titleObj->userCanRead())
 250                                 $this->dieUsage("No read permission for $titleString", 'titleaccessdenied');
 251
 252                         $linkBatch->addObj($titleObj);
 253
 254                         // Make sure we remember the original title that was given to us
 255                         // This way the caller can correlate new titles with the originally requested,
 256                         // i.e. namespace is localized or capitalization is different
 257                         if ($titleString !== $titleObj->getPrefixedText()) {
 258                                 $this->mNormalizedTitles[$titleString] = $titleObj->getPrefixedText();
 259                         }
 260                 }
 261
 262                 return $linkBatch;
 263         }
 264
 265         private function populatePageIDs($pageids) {
 266                 $this->dieUsage(__METHOD__ . ' is not implemented', 'notimplemented');
 267         }
 268
 269         public function execute() {
 270                 $titles = $pageids = $revids = $redirects = null;
 271                 extract($this->extractRequestParams());
 272
 273                 // Only one of the titles/pageids/revids is allowed at the same time
 274                 $dataSource = null;
 275                 if (isset ($titles))
 276                         $dataSource = 'titles';
 277                 if (isset ($pageids)) {
 278                         if (isset ($dataSource))
 279                                 $this->dieUsage("Cannot use 'pageids' at the same time as '$dataSource'", 'multisource');
 280                         $dataSource = 'pageids';
 281                 }
 282                 if (isset ($revids)) {
 283                         if (isset ($dataSource))
 284                                 $this->dieUsage("Cannot use 'revids' at the same time as '$dataSource'", 'multisource');
 285                         $dataSource = 'revids';
 286                 }
 287
 288                 switch ($dataSource) {
 289                         case 'titles' :
 290                                 $this->populateTitles($titles, $redirects);
 291                                 break;
 292                         case 'pageids' :
 293                                 $this->populatePageIDs($pageids, $redirects);
 294                                 break;
 295                         case 'revids' :
 296                                 $this->populateRevIDs($revids);
 297                                 break;
 298                         default :
 299                                 // Do nothing - some queries do not need any of the data sources.
 300                                 break;
 301                 }
 302         }
 303
 304         protected function getAllowedParams() {
 305                 return array (
 306                         'titles' => array (
 307                                 ApiBase::PARAM_ISMULTI => true
 308                         ),
 309                         'pageids' => array (
 310                                 ApiBase::PARAM_TYPE => 'integer',
 311                                 ApiBase::PARAM_ISMULTI => true
 312                         ),
 313                         'revids' => array (
 314                                 ApiBase::PARAM_TYPE => 'integer',
 315                                 ApiBase::PARAM_ISMULTI => true
 316                         ),
 317                         'redirects' => false
 318                 );
 319         }
 320
 321         protected function getParamDescription() {
 322                 return array (
 323                         'titles' => 'A list of titles to work on',
 324                         'pageids' => 'A list of page IDs to work on',
 325                         'revids' => 'A list of revision IDs to work on',
 326                         'redirects' => 'Automatically resolve redirects'
 327                 );
 328         }
 329 }
 330 ?>