MediaWiki  master
ApiOpenSearch.php
Go to the documentation of this file.
1 <?php
26 
30 class ApiOpenSearch extends ApiBase {
31  use SearchApi;
32 
33  private $format = null;
34  private $fm = null;
35 
37  private $allowedParams = null;
38 
44  protected function getFormat() {
45  if ( $this->format === null ) {
46  $params = $this->extractRequestParams();
47  $format = $params['format'];
48 
50  if ( !in_array( $format, $allowedParams['format'][ApiBase::PARAM_TYPE] ) ) {
52  }
53 
54  if ( substr( $format, -2 ) === 'fm' ) {
55  $this->format = substr( $format, 0, -2 );
56  $this->fm = 'fm';
57  } else {
58  $this->format = $format;
59  $this->fm = '';
60  }
61  }
62  return $this->format;
63  }
64 
65  public function getCustomPrinter() {
66  switch ( $this->getFormat() ) {
67  case 'json':
68  return new ApiOpenSearchFormatJson(
69  $this->getMain(), $this->fm, $this->getParameter( 'warningsaserror' )
70  );
71 
72  case 'xml':
73  $printer = $this->getMain()->createPrinterByName( 'xml' . $this->fm );
74  $printer->setRootElement( 'SearchSuggestion' );
75  return $printer;
76 
77  default:
78  ApiBase::dieDebug( __METHOD__, "Unsupported format '{$this->getFormat()}'" );
79  }
80  }
81 
82  public function execute() {
83  $params = $this->extractRequestParams();
84  $search = $params['search'];
85  $suggest = $params['suggest'];
86  $results = [];
87  if ( !$suggest || $this->getConfig()->get( 'EnableOpenSearchSuggest' ) ) {
88  // Open search results may be stored for a very long time
89  $this->getMain()->setCacheMaxAge( $this->getConfig()->get( 'SearchSuggestCacheExpiry' ) );
90  $this->getMain()->setCacheMode( 'public' );
91  $results = $this->search( $search, $params );
92 
93  // Allow hooks to populate extracts and images
94  Hooks::run( 'ApiOpenSearchSuggest', [ &$results ] );
95 
96  // Trim extracts, if necessary
97  $length = $this->getConfig()->get( 'OpenSearchDescriptionLength' );
98  foreach ( $results as &$r ) {
99  if ( is_string( $r['extract'] ) && !$r['extract trimmed'] ) {
100  $r['extract'] = self::trimExtract( $r['extract'], $length );
101  }
102  }
103  }
104 
105  // Populate result object
106  $this->populateResult( $search, $results );
107  }
108 
115  private function search( $search, array $params ) {
116  $searchEngine = $this->buildSearchEngine( $params );
117  $titles = $searchEngine->extractTitles( $searchEngine->completionSearchWithVariants( $search ) );
118  $results = [];
119 
120  if ( !$titles ) {
121  return $results;
122  }
123 
124  // Special pages need unique integer ids in the return list, so we just
125  // assign them negative numbers because those won't clash with the
126  // always positive articleIds that non-special pages get.
127  $nextSpecialPageId = -1;
128 
129  if ( $params['redirects'] === null ) {
130  // Backwards compatibility, don't resolve for JSON.
131  $resolveRedir = $this->getFormat() !== 'json';
132  } else {
133  $resolveRedir = $params['redirects'] === 'resolve';
134  }
135 
136  if ( $resolveRedir ) {
137  // Query for redirects
138  $redirects = [];
139  $lb = new LinkBatch( $titles );
140  if ( !$lb->isEmpty() ) {
141  $db = $this->getDB();
142  $res = $db->select(
143  [ 'page', 'redirect' ],
144  [ 'page_namespace', 'page_title', 'rd_namespace', 'rd_title' ],
145  [
146  'rd_from = page_id',
147  'rd_interwiki IS NULL OR rd_interwiki = ' . $db->addQuotes( '' ),
148  $lb->constructSet( 'page', $db ),
149  ],
150  __METHOD__
151  );
152  foreach ( $res as $row ) {
153  $redirects[$row->page_namespace][$row->page_title] =
154  [ $row->rd_namespace, $row->rd_title ];
155  }
156  }
157 
158  // Bypass any redirects
159  $seen = [];
160  foreach ( $titles as $title ) {
161  $ns = $title->getNamespace();
162  $dbkey = $title->getDBkey();
163  $from = null;
164  if ( isset( $redirects[$ns][$dbkey] ) ) {
165  list( $ns, $dbkey ) = $redirects[$ns][$dbkey];
166  $from = $title;
167  $title = Title::makeTitle( $ns, $dbkey );
168  }
169  if ( !isset( $seen[$ns][$dbkey] ) ) {
170  $seen[$ns][$dbkey] = true;
171  $resultId = $title->getArticleID();
172  if ( $resultId === 0 ) {
173  $resultId = $nextSpecialPageId;
174  $nextSpecialPageId -= 1;
175  }
176  $results[$resultId] = [
177  'title' => $title,
178  'redirect from' => $from,
179  'extract' => false,
180  'extract trimmed' => false,
181  'image' => false,
182  'url' => wfExpandUrl( $title->getFullURL(), PROTO_CURRENT ),
183  ];
184  }
185  }
186  } else {
187  foreach ( $titles as $title ) {
188  $resultId = $title->getArticleID();
189  if ( $resultId === 0 ) {
190  $resultId = $nextSpecialPageId;
191  $nextSpecialPageId -= 1;
192  }
193  $results[$resultId] = [
194  'title' => $title,
195  'redirect from' => null,
196  'extract' => false,
197  'extract trimmed' => false,
198  'image' => false,
199  'url' => wfExpandUrl( $title->getFullURL(), PROTO_CURRENT ),
200  ];
201  }
202  }
203 
204  return $results;
205  }
206 
211  protected function populateResult( $search, &$results ) {
212  $result = $this->getResult();
213 
214  switch ( $this->getFormat() ) {
215  case 'json':
216  // http://www.opensearch.org/Specifications/OpenSearch/Extensions/Suggestions/1.1
217  $result->addArrayType( null, 'array' );
218  $result->addValue( null, 0, strval( $search ) );
219  $terms = [];
220  $descriptions = [];
221  $urls = [];
222  foreach ( $results as $r ) {
223  $terms[] = $r['title']->getPrefixedText();
224  $descriptions[] = strval( $r['extract'] );
225  $urls[] = $r['url'];
226  }
227  $result->addValue( null, 1, $terms );
228  $result->addValue( null, 2, $descriptions );
229  $result->addValue( null, 3, $urls );
230  break;
231 
232  case 'xml':
233  // https://msdn.microsoft.com/en-us/library/cc891508(v=vs.85).aspx
234  $imageKeys = [
235  'source' => true,
236  'alt' => true,
237  'width' => true,
238  'height' => true,
239  'align' => true,
240  ];
241  $items = [];
242  foreach ( $results as $r ) {
243  $item = [
244  'Text' => $r['title']->getPrefixedText(),
245  'Url' => $r['url'],
246  ];
247  if ( is_string( $r['extract'] ) && $r['extract'] !== '' ) {
248  $item['Description'] = $r['extract'];
249  }
250  if ( is_array( $r['image'] ) && isset( $r['image']['source'] ) ) {
251  $item['Image'] = array_intersect_key( $r['image'], $imageKeys );
252  }
253  ApiResult::setSubelementsList( $item, array_keys( $item ) );
254  $items[] = $item;
255  }
256  ApiResult::setIndexedTagName( $items, 'Item' );
257  $result->addValue( null, 'version', '2.0' );
258  $result->addValue( null, 'xmlns', 'http://opensearch.org/searchsuggest2' );
259  $result->addValue( null, 'Query', strval( $search ) );
260  $result->addSubelementsList( null, 'Query' );
261  $result->addValue( null, 'Section', $items );
262  break;
263 
264  default:
265  ApiBase::dieDebug( __METHOD__, "Unsupported format '{$this->getFormat()}'" );
266  }
267  }
268 
269  public function getAllowedParams() {
270  if ( $this->allowedParams !== null ) {
271  return $this->allowedParams;
272  }
273  $this->allowedParams = $this->buildCommonApiParams( false ) + [
274  'suggest' => false,
275  'redirects' => [
276  ApiBase::PARAM_TYPE => [ 'return', 'resolve' ],
277  ],
278  'format' => [
279  ApiBase::PARAM_DFLT => 'json',
280  ApiBase::PARAM_TYPE => [ 'json', 'jsonfm', 'xml', 'xmlfm' ],
281  ],
282  'warningsaserror' => false,
283  ];
284 
285  // Use open search specific default limit
286  $this->allowedParams['limit'][ApiBase::PARAM_DFLT] = $this->getConfig()->get(
287  'OpenSearchDefaultLimit'
288  );
289 
290  return $this->allowedParams;
291  }
292 
293  public function getSearchProfileParams() {
294  return [
295  'profile' => [
296  'profile-type' => SearchEngine::COMPLETION_PROFILE_TYPE,
297  'help-message' => 'apihelp-query+prefixsearch-param-profile'
298  ],
299  ];
300  }
301 
302  protected function getExamplesMessages() {
303  return [
304  'action=opensearch&search=Te'
305  => 'apihelp-opensearch-example-te',
306  ];
307  }
308 
309  public function getHelpUrls() {
310  return 'https://www.mediawiki.org/wiki/Special:MyLanguage/API:Opensearch';
311  }
312 
323  public static function trimExtract( $text, $length ) {
324  static $regex = null;
325 
326  if ( $regex === null ) {
327  $endchars = [
328  '([^\d])\.\s', '\!\s', '\?\s', // regular ASCII
329  '。', // full-width ideographic full-stop
330  '.', '!', '?', // double-width roman forms
331  '。', // half-width ideographic full stop
332  ];
333  $endgroup = implode( '|', $endchars );
334  $end = "(?:$endgroup)";
335  $sentence = ".{{$length},}?$end+";
336  $regex = "/^($sentence)/u";
337  }
338 
339  $matches = [];
340  if ( preg_match( $regex, $text, $matches ) ) {
341  return trim( $matches[1] );
342  } else {
343  // Just return the first line
344  return trim( explode( "\n", $text )[0] );
345  }
346  }
347 
355  public static function getOpenSearchTemplate( $type ) {
356  $config = MediaWikiServices::getInstance()->getSearchEngineConfig();
357  $template = $config->getConfig()->get( 'OpenSearchTemplate' );
358 
359  if ( $template && $type === 'application/x-suggestions+json' ) {
360  return $template;
361  }
362 
363  $ns = implode( '|', $config->defaultNamespaces() );
364  if ( !$ns ) {
365  $ns = '0';
366  }
367 
368  switch ( $type ) {
369  case 'application/x-suggestions+json':
370  return $config->getConfig()->get( 'CanonicalServer' ) . wfScript( 'api' )
371  . '?action=opensearch&search={searchTerms}&namespace=' . $ns;
372 
373  case 'application/x-suggestions+xml':
374  return $config->getConfig()->get( 'CanonicalServer' ) . wfScript( 'api' )
375  . '?action=opensearch&format=xml&search={searchTerms}&namespace=' . $ns;
376 
377  default:
378  throw new MWException( __METHOD__ . ": Unknown type '$type'" );
379  }
380  }
381 }
382 
387  private $warningsAsError = false;
388 
389  public function __construct( ApiMain $main, $fm, $warningsAsError ) {
390  parent::__construct( $main, "json$fm" );
391  $this->warningsAsError = $warningsAsError;
392  }
393 
394  public function execute() {
395  $result = $this->getResult();
396  if ( !$result->getResultData( 'error' ) && !$result->getResultData( 'errors' ) ) {
397  // Ignore warnings or treat as errors, as requested
398  $warnings = $result->removeValue( 'warnings', null );
399  if ( $this->warningsAsError && $warnings ) {
400  $this->dieWithError(
401  'apierror-opensearch-json-warnings',
402  'warnings',
403  [ 'warnings' => $warnings ]
404  );
405  }
406 
407  // Ignore any other unexpected keys (e.g. from $wgDebugToolbar)
408  $remove = array_keys( array_diff_key(
409  $result->getResultData(),
410  [ 0 => 'search', 1 => 'terms', 2 => 'descriptions', 3 => 'urls' ]
411  ) );
412  foreach ( $remove as $key ) {
413  $result->removeValue( $key, null );
414  }
415  }
416 
417  parent::execute();
418  }
419 }
The wiki should then use memcached to cache various data To use multiple just add more items to the array To increase the weight of a make its entry a array("192.168.0.1:11211", 2))
const PARAM_TYPE
(string|string[]) Either an array of allowed value strings, or a string type as described below...
Definition: ApiBase.php:87
populateResult( $search, &$results)
deferred txt A few of the database updates required by various functions here can be deferred until after the result page is displayed to the user For updating the view updating the linked to tables after a etc PHP does not yet have any way to tell the server to actually return and disconnect while still running these but it might have such a feature in the future We handle these by creating a deferred update object and putting those objects on a global list
Definition: deferred.txt:11
getResult()
Get the result object.
Definition: ApiBase.php:659
$batch execute()
Apache License January AND DISTRIBUTION Definitions License shall mean the terms and conditions for use
getMain()
Get the main module.
Definition: ApiBase.php:555
const PARAM_DFLT
(null|boolean|integer|string) Default value of the parameter.
Definition: ApiBase.php:48
trait SearchApi
Traits for API components that use a SearchEngine.
Definition: SearchApi.php:28
wfExpandUrl( $url, $defaultProto=PROTO_CURRENT)
Expand a potentially local URL to a fully-qualified URL.
__construct(ApiMain $main, $fm, $warningsAsError)
buildCommonApiParams( $isScrollable=true)
The set of api parameters that are shared between api calls that call the SearchEngine.
Definition: SearchApi.php:46
getDB()
Gets a default replica DB connection object.
Definition: ApiBase.php:687
search( $search, array $params)
Perform the search.
dieWithError( $msg, $code=null, $data=null, $httpCode=null)
Abort execution with an error.
Definition: ApiBase.php:1972
const PROTO_CURRENT
Definition: Defines.php:222
extractRequestParams( $options=[])
Using getAllowedParams(), this function makes an array of the values provided by the user...
Definition: ApiBase.php:770
static trimExtract( $text, $length)
Trim an extract to a sensible length.
injection txt This is an overview of how MediaWiki makes use of dependency injection The design described here grew from the discussion of RFC T384 The term dependency this means that anything an object needs to operate should be injected from the the object itself should only know narrow no concrete implementation of the logic it relies on The requirement to inject everything typically results in an architecture that based on two main types of and essentially stateless service objects that use other service objects to operate on the value objects As of the beginning MediaWiki is only starting to use the DI approach Much of the code still relies on global state or direct resulting in a highly cyclical dependency MediaWikiServices
Definition: injection.txt:23
static setIndexedTagName(array &$arr, $tag)
Set the tag name for numeric-keyed values in XML format.
Definition: ApiResult.php:616
getParameter( $paramName, $parseLimit=true)
Get a value for the given parameter.
Definition: ApiBase.php:884
The index of the header message $result[1]=The index of the body text message $result[2 through n]=Parameters passed to body text message. Please note the header message cannot receive/use parameters. 'ImportHandleLogItemXMLTag':When parsing a XML tag in a log item. Return false to stop further processing of the tag $reader:XMLReader object $logInfo:Array of information 'ImportHandlePageXMLTag':When parsing a XML tag in a page. Return false to stop further processing of the tag $reader:XMLReader object & $pageInfo:Array of information 'ImportHandleRevisionXMLTag':When parsing a XML tag in a page revision. Return false to stop further processing of the tag $reader:XMLReader object $pageInfo:Array of page information $revisionInfo:Array of revision information 'ImportHandleToplevelXMLTag':When parsing a top level XML tag. Return false to stop further processing of the tag $reader:XMLReader object 'ImportHandleUnknownUser':When a user doesn 't exist locally, this hook is called to give extensions an opportunity to auto-create it. If the auto-creation is successful, return false. $name:User name 'ImportHandleUploadXMLTag':When parsing a XML tag in a file upload. Return false to stop further processing of the tag $reader:XMLReader object $revisionInfo:Array of information 'ImportLogInterwikiLink':Hook to change the interwiki link used in log entries and edit summaries for transwiki imports. & $fullInterwikiPrefix:Interwiki prefix, may contain colons. & $pageTitle:String that contains page title. 'ImportSources':Called when reading from the $wgImportSources configuration variable. Can be used to lazy-load the import sources list. & $importSources:The value of $wgImportSources. Modify as necessary. See the comment in DefaultSettings.php for the detail of how to structure this array. 'InfoAction':When building information to display on the action=info page. $context:IContextSource object & $pageInfo:Array of information 'InitializeArticleMaybeRedirect':MediaWiki check to see if title is a redirect. & $title:Title object for the current page & $request:WebRequest & $ignoreRedirect:boolean to skip redirect check & $target:Title/string of redirect target & $article:Article object 'InternalParseBeforeLinks':during Parser 's internalParse method before links but after nowiki/noinclude/includeonly/onlyinclude and other processings. & $parser:Parser object & $text:string containing partially parsed text & $stripState:Parser 's internal StripState object 'InternalParseBeforeSanitize':during Parser 's internalParse method just before the parser removes unwanted/dangerous HTML tags and after nowiki/noinclude/includeonly/onlyinclude and other processings. Ideal for syntax-extensions after template/parser function execution which respect nowiki and HTML-comments. & $parser:Parser object & $text:string containing partially parsed text & $stripState:Parser 's internal StripState object 'InterwikiLoadPrefix':When resolving if a given prefix is an interwiki or not. Return true without providing an interwiki to continue interwiki search. $prefix:interwiki prefix we are looking for. & $iwData:output array describing the interwiki with keys iw_url, iw_local, iw_trans and optionally iw_api and iw_wikiid. 'InvalidateEmailComplete':Called after a user 's email has been invalidated successfully. $user:user(object) whose email is being invalidated 'IRCLineURL':When constructing the URL to use in an IRC notification. Callee may modify $url and $query, URL will be constructed as $url . $query & $url:URL to index.php & $query:Query string $rc:RecentChange object that triggered url generation 'IsFileCacheable':Override the result of Article::isFileCacheable()(if true) & $article:article(object) being checked 'IsTrustedProxy':Override the result of IP::isTrustedProxy() & $ip:IP being check & $result:Change this value to override the result of IP::isTrustedProxy() 'IsUploadAllowedFromUrl':Override the result of UploadFromUrl::isAllowedUrl() $url:URL used to upload from & $allowed:Boolean indicating if uploading is allowed for given URL 'isValidEmailAddr':Override the result of Sanitizer::validateEmail(), for instance to return false if the domain name doesn 't match your organization. $addr:The e-mail address entered by the user & $result:Set this and return false to override the internal checks 'isValidPassword':Override the result of User::isValidPassword() $password:The password entered by the user & $result:Set this and return false to override the internal checks $user:User the password is being validated for 'Language::getMessagesFileName':$code:The language code or the language we 're looking for a messages file for & $file:The messages file path, you can override this to change the location. 'LanguageGetMagic':DEPRECATED since 1.16! Use $magicWords in a file listed in $wgExtensionMessagesFiles instead. Use this to define synonyms of magic words depending of the language & $magicExtensions:associative array of magic words synonyms $lang:language code(string) 'LanguageGetNamespaces':Provide custom ordering for namespaces or remove namespaces. Do not use this hook to add namespaces. Use CanonicalNamespaces for that. & $namespaces:Array of namespaces indexed by their numbers 'LanguageGetSpecialPageAliases':DEPRECATED! Use $specialPageAliases in a file listed in $wgExtensionMessagesFiles instead. Use to define aliases of special pages names depending of the language & $specialPageAliases:associative array of magic words synonyms $lang:language code(string) 'LanguageGetTranslatedLanguageNames':Provide translated language names. & $names:array of language code=> language name $code:language of the preferred translations 'LanguageLinks':Manipulate a page 's language links. This is called in various places to allow extensions to define the effective language links for a page. $title:The page 's Title. & $links:Array with elements of the form "language:title" in the order that they will be output. & $linkFlags:Associative array mapping prefixed links to arrays of flags. Currently unused, but planned to provide support for marking individual language links in the UI, e.g. for featured articles. 'LanguageSelector':Hook to change the language selector available on a page. $out:The output page. $cssClassName:CSS class name of the language selector. 'LinkBegin':DEPRECATED since 1.28! Use HtmlPageLinkRendererBegin instead. Used when generating internal and interwiki links in Linker::link(), before processing starts. Return false to skip default processing and return $ret. See documentation for Linker::link() for details on the expected meanings of parameters. $skin:the Skin object $target:the Title that the link is pointing to & $html:the contents that the< a > tag should have(raw HTML) $result
Definition: hooks.txt:2039
this hook is for auditing only or null if authentication failed before getting that far or null if we can t even determine that probably a stub it is not rendered in wiki pages or galleries in category pages allow injecting custom HTML after the section Any uses of the hook need to handle escaping $template
Definition: hooks.txt:798
wfScript( $script='index')
Get the path to a specified script file, respecting file extensions; this is a wrapper around $wgScri...
Class representing a list of titles The execute() method checks them all for existence and adds them ...
Definition: LinkBatch.php:34
static getOpenSearchTemplate( $type)
Fetch the template for a type.
$res
Definition: database.txt:21
getFormat()
Get the output format.
$params
This is the main API class, used for both external and internal processing.
Definition: ApiMain.php:41
const COMPLETION_PROFILE_TYPE
string profile type for completionSearch
namespace and then decline to actually register it file or subcat img or subcat $title
Definition: hooks.txt:949
This document is intended to provide useful advice for parties seeking to redistribute MediaWiki to end users It s targeted particularly at maintainers for Linux since it s been observed that distribution packages of MediaWiki often break We ve consistently had to recommend that users seeking support use official tarballs instead of their distribution s and this often solves whatever problem the user is having It would be nice if this could such as
Definition: distributors.txt:9
static setSubelementsList(array &$arr, $names)
Causes the elements with the specified names to be output as subelements rather than attributes...
Definition: ApiResult.php:565
static makeTitle( $ns, $title, $fragment='', $interwiki='')
Create a new Title from a namespace index and a DB key.
Definition: Title.php:538
injection txt This is an overview of how MediaWiki makes use of dependency injection The design described here grew from the discussion of RFC T384 The term dependency this means that anything an object needs to operate should be injected from the the object itself should only know narrow no concrete implementation of the logic it relies on The requirement to inject everything typically results in an architecture that based on two main types of and essentially stateless service objects that use other service objects to operate on the value objects As of the beginning MediaWiki is only starting to use the DI approach Much of the code still relies on global state or direct resulting in a highly cyclical dependency which acts as the top level factory for services in MediaWiki which can be used to gain access to default instances of various services MediaWikiServices however also allows new services to be defined and default services to be redefined Services are defined or redefined by providing a callback the instantiator that will return a new instance of the service When it will create an instance of MediaWikiServices and populate it with the services defined in the files listed by thereby bootstrapping the DI framework Per $wgServiceWiringFiles lists includes ServiceWiring php
Definition: injection.txt:35
buildSearchEngine(array $params=null)
Build the search engine to use.
Definition: SearchApi.php:150
linkcache txt The LinkCache class maintains a list of article titles and the information about whether or not the article exists in the database This is used to mark up links when displaying a page If the same link appears more than once on any page then it only has to be looked up once In most cases link lookups are done in batches with the LinkBatch class or the equivalent in so the link cache is mostly useful for short snippets of parsed and for links in the navigation areas of the skin The link cache was formerly used to track links used in a document for the purposes of updating the link tables This application is now deprecated To create a you can use the following $titles
Definition: linkcache.txt:17
static dieDebug( $method, $message)
Internal code errors should be reported with this method.
Definition: ApiBase.php:2152
This abstract class implements many basic API functions, and is the base of all API classes...
Definition: ApiBase.php:37
API JSON output formatter.
array $allowedParams
list of api allowed params
if the prop value should be in the metadata multi language array format
Definition: hooks.txt:1694
static run( $event, array $args=[], $deprecatedVersion=null)
Call hook functions defined in Hooks::register and $wgHooks.
Definition: Hooks.php:200
$matches