CategoryTraverser::descend() - Code Metrics - Inspection of "Typing & PHP74 (#55)" - addwiki/mediawiki-api - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — main ( 2daa48...b5d932 )

unknown

created 2021-02-18 21:06 UTC

CategoryTraverser::descend() B

↳ Parent: Project

Complexity

Conditions	9
Paths	26

Size

Total Lines

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
dl	0
loc	57
rs	7.3826
c	0
b	0
f	0
cc	9
nc	26
nop	2

How to fix Long Method

<?php

namespace Addwiki\Mediawiki\Api\Service;

use Addwiki\Mediawiki\Api\CategoryLoopException;
use Addwiki\Mediawiki\Api\Client\MediawikiApi;
use Addwiki\Mediawiki\Api\Client\SimpleRequest;
use Addwiki\Mediawiki\DataModel\Page;
use Addwiki\Mediawiki\DataModel\Pages;
use Addwiki\Mediawiki\DataModel\Title;

/**
 * Category traverser.
 *
 * Note on spelling 'descendant' (from Wiktionary):
 * The adjective, "descending from a biological ancestor", may be spelt either
 * with an 'a' or with an 'e' in the final syllable. However the noun descendant,
 * "one who is the progeny of someone", may be spelt only with an 'a'.
 */
class CategoryTraverser extends Service {

	/**
	 * @var int
	 */
	public const CALLBACK_CATEGORY = 10;
	/**
	 * @var int
	 */
	public const CALLBACK_PAGE = 20;

	/**
	 * @var string[]|null
	 */
	protected $namespaces;

	/**
	 * @var callable[]
	 */
	protected array $callbacks = [];


	/**
	 * Used to remember the previously-visited categories when traversing.
	 * @var string[]
	 */
	protected array $alreadyVisited = [];

	/**
	 * @param MediawikiApi $api The API to connect to.
	 */
	public function __construct( MediawikiApi $api ) {
		parent::__construct( $api );
		$this->callbacks = [];
	}

	/**
	 * Query the remote site for the list of namespaces in use, so that later we can tell what's a
	 * category and what's not. This populates $this->namespaces, and will not re-request on
	 * repeated invocations.
	 */
	protected function retrieveNamespaces(): void {
		if ( is_array( $this->namespaces ) ) {
			return;
		}
		$params = [ 'meta' => 'siteinfo', 'siprop' => 'namespaces' ];
		$namespaces = $this->api->getRequest( new SimpleRequest( 'query', $params ) );
		if ( isset( $namespaces['query']['namespaces'] ) ) {
			$this->namespaces = $namespaces['query']['namespaces'];
		}
	}

	/**
	 * Register a callback that will be called for each page or category visited during the
	 * traversal.
	 * @param int $type One of the 'CALLBACK_' constants of this class.
	 * @param callable $callback A callable that takes two \Addwiki\Mediawiki\DataModel\Page parameters.
	 */
	public function addCallback( int $type, callable $callback ): void {
		if ( !isset( $this->callbacks[$type] ) ) {
			$this->callbacks[$type] = [];
		}
		$this->callbacks[$type][] = $callback;
	}

	/**
	 * Visit every descendant page of $rootCategoryName (which will be a Category
	 * page, because there are no desecendants of any other pages).
	 * @param Page $rootCat The full name of the page to start at.
	 * @param Pages|null $currentPath Used only when recursing into this method, to track each path
	 * through the category hierarchy in case of loops.
	 * @return Pages All descendants of the given category.
	 * @throws CategoryLoopException If a category loop is detected.
	 */
	public function descend( Page $rootCat, ?Pages $currentPath = null ): Pages {
		// Make sure we know the namespace IDs.
		$this->retrieveNamespaces();

		$rootCatName = $rootCat->getPageIdentifier()->getTitle()->getText();
		if ( $currentPath === null ) {
			$this->alreadyVisited = [];
			$currentPath = new Pages();
		}
		$this->alreadyVisited[] = $rootCatName;
		$currentPath->addPage( $rootCat );

		// Start a list of child pages.
		$descendants = new Pages();
		do {
			$pageListGetter = new PageListGetter( $this->api );
			$members = $pageListGetter->getPageListFromCategoryName( $rootCatName );
			foreach ( $members->toArray() as $member ) {
				/** @var Title */
				$memberTitle = $member->getPageIdentifier()->getTitle();

				// See if this page is a Category page.
				$isCat = false;
				if ( isset( $this->namespaces[ $memberTitle->getNs() ] ) ) {
					$ns = $this->namespaces[ $memberTitle->getNs() ];
					$isCat = ( isset( $ns['canonical'] ) && $ns['canonical'] === 'Category' );
				}
				// If it's a category, descend into it.
				if ( $isCat ) {
					// If this member has already been visited on this branch of the traversal,
					// throw an Exception with information about which categories form the loop.
					if ( $currentPath->hasPage( $member ) ) {
						$currentPath->addPage( $member );
						$loop = new CategoryLoopException();
						$loop->setCategoryPath( $currentPath );
						throw $loop;
					}
					// Don't go any further if we've already visited this member
					// (does not indicate a loop, however; we've already caught that above).
					if ( in_array( $memberTitle->getText(), $this->alreadyVisited ) ) {
						continue;
					}
					// Call any registered callbacked, and carry on to the next branch.
					$this->call( self::CALLBACK_CATEGORY, [ $member, $rootCat ] );
					$newDescendants = $this->descend( $member, $currentPath );
					$descendants->addPages( $newDescendants );
					// Re-set the path.
					$currentPath = new Pages();
				} else {
					// If it's a page, add it to the list and carry on.
					$descendants->addPage( $member );
					$this->call( self::CALLBACK_PAGE, [ $member, $rootCat ] );
				}
			}
		} while ( isset( $result['continue'] ) );
		return $descendants;
	}

	/**
	 * Call all the registered callbacks of a particular type.
	 * @param int $type The callback type; should match one of the 'CALLBACK_' constants.
	 * @param mixed[] $params The parameters to pass to the callback function.
	 */
	protected function call( int $type, array $params ): void {
		if ( !isset( $this->callbacks[$type] ) ) {
			return;
		}
		foreach ( $this->callbacks[$type] as $callback ) {
			if ( is_callable( $callback ) ) {
				call_user_func_array( $callback, $params );
			}
		}
	}

}


1			<?php
2
3			namespace Addwiki\Mediawiki\Api\Service;
4
5			use Addwiki\Mediawiki\Api\CategoryLoopException;
6			use Addwiki\Mediawiki\Api\Client\MediawikiApi;
7			use Addwiki\Mediawiki\Api\Client\SimpleRequest;
8			use Addwiki\Mediawiki\DataModel\Page;
9			use Addwiki\Mediawiki\DataModel\Pages;
10			use Addwiki\Mediawiki\DataModel\Title;
11
12			/**
13			* Category traverser.
14			*
15			* Note on spelling 'descendant' (from Wiktionary):
16			* The adjective, "descending from a biological ancestor", may be spelt either
17			* with an 'a' or with an 'e' in the final syllable. However the noun descendant,
18			* "one who is the progeny of someone", may be spelt only with an 'a'.
19			*/
20			class CategoryTraverser extends Service {
21
22			/**
23			* @var int
24			*/
25			public const CALLBACK_CATEGORY = 10;
26			/**
27			* @var int
28			*/
29			public const CALLBACK_PAGE = 20;
30
31			/**
32			* @var string[]\|null
33			*/
34			protected $namespaces;
35
36			/**
37			* @var callable[]
38			*/
39			protected array $callbacks = [];
			0 ignored issues – show Bug introduced 2021-02-18 21:15 UTC by Report Bug Copy Issue Report This code did not parse for me. Apparently, there is an error somewhere around this line: Syntax error, unexpected T_ARRAY, expecting T_FUNCTION or T_CONST Loading history...
40
41			/**
42			* Used to remember the previously-visited categories when traversing.
43			* @var string[]
44			*/
45			protected array $alreadyVisited = [];
46
47			/**
48			* @param MediawikiApi $api The API to connect to.
49			*/
50			public function __construct( MediawikiApi $api ) {
51			parent::__construct( $api );
52			$this->callbacks = [];
53			}
54
55			/**
56			* Query the remote site for the list of namespaces in use, so that later we can tell what's a
57			* category and what's not. This populates $this->namespaces, and will not re-request on
58			* repeated invocations.
59			*/
60			protected function retrieveNamespaces(): void {
61			if ( is_array( $this->namespaces ) ) {
62			return;
63			}
64			$params = [ 'meta' => 'siteinfo', 'siprop' => 'namespaces' ];
65			$namespaces = $this->api->getRequest( new SimpleRequest( 'query', $params ) );
66			if ( isset( $namespaces['query']['namespaces'] ) ) {
67			$this->namespaces = $namespaces['query']['namespaces'];
68			}
69			}
70
71			/**
72			* Register a callback that will be called for each page or category visited during the
73			* traversal.
74			* @param int $type One of the 'CALLBACK_' constants of this class.
75			* @param callable $callback A callable that takes two \Addwiki\Mediawiki\DataModel\Page parameters.
76			*/
77			public function addCallback( int $type, callable $callback ): void {
78			if ( !isset( $this->callbacks[$type] ) ) {
79			$this->callbacks[$type] = [];
80			}
81			$this->callbacks[$type][] = $callback;
82			}
83
84			/**
85			* Visit every descendant page of $rootCategoryName (which will be a Category
86			* page, because there are no desecendants of any other pages).
87			* @param Page $rootCat The full name of the page to start at.
88			* @param Pages\|null $currentPath Used only when recursing into this method, to track each path
89			* through the category hierarchy in case of loops.
90			* @return Pages All descendants of the given category.
91			* @throws CategoryLoopException If a category loop is detected.
92			*/
93			public function descend( Page $rootCat, ?Pages $currentPath = null ): Pages {
94			// Make sure we know the namespace IDs.
95			$this->retrieveNamespaces();
96
97			$rootCatName = $rootCat->getPageIdentifier()->getTitle()->getText();
98			if ( $currentPath === null ) {
99			$this->alreadyVisited = [];
100			$currentPath = new Pages();
101			}
102			$this->alreadyVisited[] = $rootCatName;
103			$currentPath->addPage( $rootCat );
104
105			// Start a list of child pages.
106			$descendants = new Pages();
107			do {
108			$pageListGetter = new PageListGetter( $this->api );
109			$members = $pageListGetter->getPageListFromCategoryName( $rootCatName );
110			foreach ( $members->toArray() as $member ) {
111			/** @var Title */
112			$memberTitle = $member->getPageIdentifier()->getTitle();
113
114			// See if this page is a Category page.
115			$isCat = false;
116			if ( isset( $this->namespaces[ $memberTitle->getNs() ] ) ) {
117			$ns = $this->namespaces[ $memberTitle->getNs() ];
118			$isCat = ( isset( $ns['canonical'] ) && $ns['canonical'] === 'Category' );
119			}
120			// If it's a category, descend into it.
121			if ( $isCat ) {
122			// If this member has already been visited on this branch of the traversal,
123			// throw an Exception with information about which categories form the loop.
124			if ( $currentPath->hasPage( $member ) ) {
125			$currentPath->addPage( $member );
126			$loop = new CategoryLoopException();
127			$loop->setCategoryPath( $currentPath );
128			throw $loop;
129			}
130			// Don't go any further if we've already visited this member
131			// (does not indicate a loop, however; we've already caught that above).
132			if ( in_array( $memberTitle->getText(), $this->alreadyVisited ) ) {
133			continue;
134			}
135			// Call any registered callbacked, and carry on to the next branch.
136			$this->call( self::CALLBACK_CATEGORY, [ $member, $rootCat ] );
137			$newDescendants = $this->descend( $member, $currentPath );
138			$descendants->addPages( $newDescendants );
139			// Re-set the path.
140			$currentPath = new Pages();
141			} else {
142			// If it's a page, add it to the list and carry on.
143			$descendants->addPage( $member );
144			$this->call( self::CALLBACK_PAGE, [ $member, $rootCat ] );
145			}
146			}
147			} while ( isset( $result['continue'] ) );
148			return $descendants;
149			}
150
151			/**
152			* Call all the registered callbacks of a particular type.
153			* @param int $type The callback type; should match one of the 'CALLBACK_' constants.
154			* @param mixed[] $params The parameters to pass to the callback function.
155			*/
156			protected function call( int $type, array $params ): void {
157			if ( !isset( $this->callbacks[$type] ) ) {
158			return;
159			}
160			foreach ( $this->callbacks[$type] as $callback ) {
161			if ( is_callable( $callback ) ) {
162			call_user_func_array( $callback, $params );
163			}
164			}
165			}
166
167			}
168

addwiki / mediawiki-api

Push — main ( 2daa48...b5d932 )

CategoryTraverser::descend() B

Complexity

Size

Duplication

Importance

How to fix Long Method

Long Method

Duplication Side-by-Side

Filter issues like