StaticPagesQueue::remove_duplicates() - Code Metrics - Inspection of "Merge pull request #55 from johannesx75/patch-2" - silverstripe/silverstripe-staticpublishqueue - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( bb3e3e...12ae4c )

by Ed

created 2016-07-23 05:04 UTC

StaticPagesQueue::remove_duplicates() A

↳ Parent: StaticPagesQueue

Complexity

Conditions	2
Paths	2

Size

Total Lines	7
Code Lines	5

Duplication

Lines	0
Ratio	0 %

Importance

Changes	5
Bugs	0	Features	1

Metric	Value
c	5
b	0
f	1
dl	0
loc	7
rs	9.4285
cc	2
eloc	5
nc	2
nop	1

<?php
/**
 * This class responsibility is twofold:
 * 1) Holding the data for a prioritized queue of URLs that needs to be static cached
 * 2) Interaction with that queue
 *
 * @TODO: would be good to refactor this queue to hold not only URLSegment, but also ClassName and ID of the
 * associated object (or any other metadata). This would allow FilesystemPublisher::publishPages and others
 * to stop having to smuggle the metadata within the URL (see URLArrayData::get_object).
 *
 */
class StaticPagesQueue extends DataObject {
namespace YourVendor;

class YourClass { }

	/**
	 *
	 * @var array
	 */
	public static $create_table_options = array(
		'MySQLDatabase' => 'ENGINE=InnoDB'
	);

	/**
	 *
	 * @var array
	 */
	public static $db = array(
		'Priority' => 'Int',
		'URLSegment' => 'Varchar(255)',
		'Freshness' => "Enum('stale, regenerating, error', 'stale')"
	);

	/**
	 *
	 * @var array
	 */
	public static $defaults = array(
		"Priority" => 3
	);

	/**
	 *
	 * @var array
	 */
	public static $default_sort = "\"Priority\"";

	/**
	 * Sets database indexes
	 *
	 * @var array
	 */
	public static $indexes = array(
		'freshness_priority_created' => '(Freshness, Priority, Created)',
	);

	/**
	 *
	 * @var boolean
	 */
	private static $disable_mysql_locks = false;


	/**
	 *
	 * @var boolean
	 */
	private static $realtime = false;


	/**
	 *
	 * @var int
	 */
	protected static $minutes_until_force_regeneration = 1;

	/**
	 *
	 * @var array
	 */
	protected static $insert_statements = array();

	/**
	 *
	 * @var array
	 */
	protected static $urls = array();
	
	/**
	 *
	 * @return bool

	 */
	public static function is_realtime() {
		return Config::inst()->get('StaticPagesQueue','realtime');
	}

	/**
	 *
	 * @param type $priority
	 * @param type $URLSegment
	 * @return type

	 */
	public static function add_to_queue($priority, $URLSegment) {
		$now = date("Y-m-d H:i:s");
		self::$insert_statements[$URLSegment] = '(\''.$now.'\',\''.$now.'\', \''.Convert::raw2sql($priority).'\',\''.Convert::raw2sql($URLSegment).'\')';
		self::$urls[md5($URLSegment)] = $URLSegment;
	}

		/**
	 * This will push all the currently cached insert statements to be pushed 
	 * into the database
	 *
	 * @return void
	 */
	public static function push_urls_to_db() {
		foreach(self::$insert_statements as $stmt) {
			$insertSQL = 'INSERT INTO "StaticPagesQueue" ("Created", "LastEdited", "Priority", "URLSegment") VALUES ' . $stmt;
			DB::query($insertSQL);
		}
		self::remove_old_cache(self::$urls);
		// Flush the cache so DataObject::get works correctly
		if(!empty(self::$insert_statements) && DB::affectedRows()) {
			singleton(__CLASS__)->flushCache();
		}
		self::$insert_statements = array();
	}
	
	/**
	 * Remove an object by the url
	 *
	 * @param string $URLSegment
	 * @return bool - if there was an queue item removed
	 *
	 */
	public static function delete_by_link($URLSegment) {
		$object = self::get_by_link($URLSegment);
		if(!$object) return false;

		$object->delete();
		unset($object);
		return true;
	}
	
	/**
	 * Update the queue with the information that this url renders an error somehow
	 *
	 * @param string $url
	 */
	public static function has_error( $url ) {
		if(!$url) return;
		
		$existingObject = self::get_by_link($url);
		$existingObject->Freshness = 'error';
		$existingObject->write();
	}

	/**
	 * Returns a single queue object according to a particular priority and freshness measure.
	 * This method removes any duplicates and makes the object as "regenerating", so other calls to this method
	 * don't grab the same object.
	 * If we are using MySQLDatabase with InnoDB, we do row-level locking when updating the dataobject to allow for
	 * distributed cache rebuilds
	 * @static
	 * @param $freshness
	 * @param $sortOrder
	 */
	protected static function get_queue_object($freshness, $interval = null, $sortOrder = array('Priority'=>'DESC', 'ID'=>'ASC')) {

		$className = __CLASS__;
		$queueObject = null;
		$filterQuery = array("Freshness" => $freshness);
		if ($interval) $filterQuery["LastEdited:LessThan"] = $interval;

		$query = self::get();
		if ($query->Count() > 0) {
			$offset = 0;
			$filteredQuery = $query->filter($filterQuery)->sort($sortOrder);

			if ($filteredQuery->Count() > 0) {
				if (!self::config()->disable_mysql_locks && DB::getConn() instanceof MySQLDatabase) {   //locking currently only works on MySQL
if ($x instanceof DoesNotExist) {
    // Do something.
}

					do {
						$queueObject = $filteredQuery->limit(1, $offset)->first();   //get first item

						if ($queueObject) $lockName = md5($queueObject->URLSegment . $className);
						//try to locking the item's URL, keep trying new URLs until we find one that is free to lock
						$offset++;
					} while($queueObject && !LockMySQL::isFreeToLock($lockName));
function myFunction($a) {
    switch ($a) {
        case 'foo':
            $x = 1;
            break;

        case 'bar':
            $x = 2;
            break;
    }

    // $x is potentially undefined here.
    echo $x;
}

					if ($queueObject) {
						$lockSuccess = LockMySQL::getLock($lockName);  //acquire a lock with the URL of the queue item we have just fetched
						if ($lockSuccess) {
							self::remove_duplicates($queueObject->ID);  //remove any duplicates
							self::mark_as_regenerating($queueObject);   //mark as regenerating so nothing else grabs it
							LockMySQL::releaseLock($lockName);			//return the object and release the lock
						}
					}
				} else {
					$queueObject = $filteredQuery->first();
					self::remove_duplicates($queueObject->ID);
					self::mark_as_regenerating($queueObject);
				}
			}
		}

		return $queueObject;    //return the object or null
	}

	/**
	 * Finds the next most prioritized url that needs recaching
	 *
	 * @return string
	 */
	public static function get_next_url() {
		$object = self::get_queue_object('stale');
		if($object) return $object->URLSegment;

		$interval = date('Y-m-d H:i:s', strtotime('-'.self::$minutes_until_force_regeneration.' minutes'));

		// Find URLs that has been stuck in regeneration
		$object = self::get_queue_object('regenerating', $interval);
		if($object) return $object->URLSegment;

		// Find URLs that is erronous and might work now (flush issues etc)
		$object = self::get_queue_object('error', $interval);
		if($object) return $object->URLSegment;

		return '';
	}

	/**
	 * Removes the .html fresh copy of the cache.
	 * Keeps the *.stale.html copy in place,
	 * in order to notify the user of the stale content.
	 *
	 * @param array $URLSegments
	 */
	protected static function remove_old_cache( array $URLSegments ) {
		$publisher = singleton('SiteTree')->getExtensionInstance('FilesystemPublisher');
		if ($publisher) {
			$paths = $publisher->urlsToPaths($URLSegments);
			foreach($paths as $absolutePath) {

				if(!file_exists($publisher->getDestDir().'/'.$absolutePath)) {
					continue;
				}

				unlink($publisher->getDestDir().'/'.$absolutePath);
			}
		}
	}

	/**
	 * Mark this current StaticPagesQueue as a work in progress
	 *
	 * @param StaticPagesQueue $object 
	 */
	protected static function mark_as_regenerating(StaticPagesQueue $object) {
		$now = date('Y-m-d H:i:s');
		DB::query('UPDATE "StaticPagesQueue" SET "LastEdited" = \''.$now.'\', "Freshness"=\'regenerating\' WHERE "ID" = '.$object->ID);
		singleton(__CLASS__)->flushCache();
	}

	/**
	 * Removes all duplicates that has the same URLSegment as $ID
	 *
	 * @param int $ID - ID of the object whose duplicates we want to remove
	 * @return void

	 */
	static function remove_duplicates( $ID ) {

		$obj = DataObject::get_by_id('StaticPagesQueue', $ID);
		if(!$obj) return 0;
		DB::query(
			sprintf('DELETE FROM "StaticPagesQueue" WHERE "URLSegment" = \'%s\' AND "ID" != %d', $obj->URLSegment, (int)$ID)
		);
	}

	/**
	 *
	 * @param string $url
	 * @param bool $onlyStale - Get only stale entries
/**
 * @param array $germany
 * @param array $island
 * @param array $italy
 */
function finale($germany, $island) {
    return "2:1";
}
	 * @return DataObject || false - The first item matching the query

	 */
	protected static function get_by_link($url) {
		$filter = '"URLSegment" = \''.Convert::raw2sql($url).'\'';
		$res = DB::query('SELECT * FROM "StaticPagesQueue" WHERE '.$filter.' LIMIT 1;');
		if(!$res->numRecords()){
			return false;
		}
		return new StaticPagesQueue($res->first());
	}
}


Push — master ( bb3e3e...12ae4c )

StaticPagesQueue::remove_duplicates() A

Complexity

Size

Duplication

Importance

1. Missing dependencies

2. Missing use statement

Available Fixes

1			<?php
2			/**
3			* This class responsibility is twofold:
4			* 1) Holding the data for a prioritized queue of URLs that needs to be static cached
5			* 2) Interaction with that queue
6			*
7			* @TODO: would be good to refactor this queue to hold not only URLSegment, but also ClassName and ID of the
8			* associated object (or any other metadata). This would allow FilesystemPublisher::publishPages and others
9			* to stop having to smuggle the metadata within the URL (see URLArrayData::get_object).
10			*
11			*/
12			class StaticPagesQueue extends DataObject {
			0 ignored issues – show Coding Style Compatibility introduced 2016-03-22 06:43 UTC by Report Bug Copy Issue Report PSR1 recommends that each class must be in a namespace of at least one level to avoid collisions. You can fix this by adding a namespace to your class: namespace YourVendor; class YourClass { } When choosing a vendor namespace, try to pick something that is not too generic to avoid conflicts with other libraries. Loading history...
13
14			/**
15			*
16			* @var array
17			*/
18			public static $create_table_options = array(
19			'MySQLDatabase' => 'ENGINE=InnoDB'
20			);
21
22			/**
23			*
24			* @var array
25			*/
26			public static $db = array(
27			'Priority' => 'Int',
28			'URLSegment' => 'Varchar(255)',
29			'Freshness' => "Enum('stale, regenerating, error', 'stale')"
30			);
31
32			/**
33			*
34			* @var array
35			*/
36			public static $defaults = array(
37			"Priority" => 3
38			);
39
40			/**
41			*
42			* @var array
43			*/
44			public static $default_sort = "\"Priority\"";
45
46			/**
47			* Sets database indexes
48			*
49			* @var array
50			*/
51			public static $indexes = array(
52			'freshness_priority_created' => '(Freshness, Priority, Created)',
53			);
54
55			/**
56			*
57			* @var boolean
58			*/
59			private static $disable_mysql_locks = false;
			0 ignored issues – show Unused Code introduced 2016-07-11 13:02 UTC by Report Bug Copy Issue Report The property `$disable_mysql_locks` is not used and could be removed. This check marks private properties in classes that are never used. Those properties can be removed. Loading history...
60
61			/**
62			*
63			* @var boolean
64			*/
65			private static $realtime = false;
			0 ignored issues – show Unused Code introduced 2016-03-22 06:43 UTC by Report Bug Copy Issue Report The property `$realtime` is not used and could be removed. This check marks private properties in classes that are never used. Those properties can be removed. Loading history...
66
67			/**
68			*
69			* @var int
70			*/
71			protected static $minutes_until_force_regeneration = 1;
72
73			/**
74			*
75			* @var array
76			*/
77			protected static $insert_statements = array();
78
79			/**
80			*
81			* @var array
82			*/
83			protected static $urls = array();
84
85			/**
86			*
87			* @return bool
			0 ignored issues – show Documentation introduced 2016-03-22 06:43 UTC by Report Bug Copy Issue Report Should the return type not be `array\|integer\|double\|string\|boolean`? This check compares the return type specified in the `@return` annotation of a function or method doc comment with the types returned by the function and raises an issue if they mismatch. Loading history...
88			*/
89			public static function is_realtime() {
90			return Config::inst()->get('StaticPagesQueue','realtime');
91			}
92
93			/**
94			*
95			* @param type $priority
96			* @param type $URLSegment
97			* @return type
			0 ignored issues – show Documentation introduced 2016-03-22 06:43 UTC by Report Bug Copy Issue Report Should the return type not be `type\|null`? This check compares the return type specified in the `@return` annotation of a function or method doc comment with the types returned by the function and raises an issue if they mismatch. Loading history...
98			*/
99			public static function add_to_queue($priority, $URLSegment) {
100			$now = date("Y-m-d H:i:s");
101			self::$insert_statements[$URLSegment] = '(\''.$now.'\',\''.$now.'\', \''.Convert::raw2sql($priority).'\',\''.Convert::raw2sql($URLSegment).'\')';
102			self::$urls[md5($URLSegment)] = $URLSegment;
103			}
104
105			/**
106			* This will push all the currently cached insert statements to be pushed
107			* into the database
108			*
109			* @return void
110			*/
111			public static function push_urls_to_db() {
112			foreach(self::$insert_statements as $stmt) {
113			$insertSQL = 'INSERT INTO "StaticPagesQueue" ("Created", "LastEdited", "Priority", "URLSegment") VALUES ' . $stmt;
114			DB::query($insertSQL);
115			}
116			self::remove_old_cache(self::$urls);
117			// Flush the cache so DataObject::get works correctly
118			if(!empty(self::$insert_statements) && DB::affectedRows()) {
119			singleton(__CLASS__)->flushCache();
120			}
121			self::$insert_statements = array();
122			}
123
124			/**
125			* Remove an object by the url
126			*
127			* @param string $URLSegment
128			* @return bool - if there was an queue item removed
129			*
130			*/
131			public static function delete_by_link($URLSegment) {
132			$object = self::get_by_link($URLSegment);
133			if(!$object) return false;
134
135			$object->delete();
136			unset($object);
137			return true;
138			}
139
140			/**
141			* Update the queue with the information that this url renders an error somehow
142			*
143			* @param string $url
144			*/
145			public static function has_error( $url ) {
146			if(!$url) return;
147
148			$existingObject = self::get_by_link($url);
149			$existingObject->Freshness = 'error';
150			$existingObject->write();
151			}
152
153			/**
154			* Returns a single queue object according to a particular priority and freshness measure.
155			* This method removes any duplicates and makes the object as "regenerating", so other calls to this method
156			* don't grab the same object.
157			* If we are using MySQLDatabase with InnoDB, we do row-level locking when updating the dataobject to allow for
158			* distributed cache rebuilds
159			* @static
160			* @param $freshness
161			* @param $sortOrder
162			*/
163			protected static function get_queue_object($freshness, $interval = null, $sortOrder = array('Priority'=>'DESC', 'ID'=>'ASC')) {
			0 ignored issues – show Documentation introduced 2016-03-22 06:43 UTC by Report Bug Copy Issue Report The return type could not be reliably inferred; please add a `@return` annotation. Our type inference engine in quite powerful, but sometimes the code does not provide enough clues to go by. In these cases we request you to add a `@return` annotation as described here. Loading history...
164			$className = __CLASS__;
165			$queueObject = null;
166			$filterQuery = array("Freshness" => $freshness);
167			if ($interval) $filterQuery["LastEdited:LessThan"] = $interval;
168
169			$query = self::get();
170			if ($query->Count() > 0) {
171			$offset = 0;
172			$filteredQuery = $query->filter($filterQuery)->sort($sortOrder);
173
174			if ($filteredQuery->Count() > 0) {
175			if (!self::config()->disable_mysql_locks && DB::getConn() instanceof MySQLDatabase) { //locking currently only works on MySQL
			0 ignored issues – show Bug introduced 2016-07-08 16:33 UTC by Report Bug Copy Issue Report The class `MySQLDatabase` does not exist. Did you forget a USE statement, or did you not list all dependencies? This error could be the result of: 1. Missing dependencies PHP Analyzer uses your `composer.json` file (if available) to determine the dependencies of your project and to determine all the available classes and functions. It expects the `composer.json` to be in the root folder of your repository. Are you sure this class is defined by one of your dependencies, or did you maybe not list a dependency in either the `require` or `require-dev` section? 2. Missing use statement PHP does not complain about undefined classes in `ìnstanceof` checks. For example, the following PHP code will work perfectly fine: if ($x instanceof DoesNotExist) { // Do something. } If you have not tested against this specific condition, such errors might go unnoticed. Loading history...
176
177			do {
178			$queueObject = $filteredQuery->limit(1, $offset)->first(); //get first item
179
180			if ($queueObject) $lockName = md5($queueObject->URLSegment . $className);
181			//try to locking the item's URL, keep trying new URLs until we find one that is free to lock
182			$offset++;
183			} while($queueObject && !LockMySQL::isFreeToLock($lockName));
			0 ignored issues – show Bug introduced 2015-12-02 22:54 UTC by Report Bug Copy Issue Report The variable `$lockName` does not seem to be defined for all execution paths leading up to this point. If you define a variable conditionally, it can happen that it is not defined for all execution paths. Let’s take a look at an example: function myFunction($a) { switch ($a) { case 'foo': $x = 1; break; case 'bar': $x = 2; break; } // $x is potentially undefined here. echo $x; } In the above example, the variable `$x` is defined if you pass “foo” or “bar” as argument for `$a`. However, since the `switch` statement has no default case statement, if you pass any other value, the variable `$x` would be undefined. Available Fixes Check for existence of the variable explicitly: function myFunction($a) { switch ($a) { case 'foo': $x = 1; break; case 'bar': $x = 2; break; } if (isset($x)) { // Make sure it's always set. echo $x; } } Define a default value for the variable: function myFunction($a) { $x = ''; // Set a default which gets overridden for certain paths. switch ($a) { case 'foo': $x = 1; break; case 'bar': $x = 2; break; } echo $x; } Add a value for the missing path: function myFunction($a) { switch ($a) { case 'foo': $x = 1; break; case 'bar': $x = 2; break; // We add support for the missing case. default: $x = ''; break; } echo $x; } Loading history...
184
185			if ($queueObject) {
186			$lockSuccess = LockMySQL::getLock($lockName); //acquire a lock with the URL of the queue item we have just fetched
187			if ($lockSuccess) {
188			self::remove_duplicates($queueObject->ID); //remove any duplicates
189			self::mark_as_regenerating($queueObject); //mark as regenerating so nothing else grabs it
190			LockMySQL::releaseLock($lockName); //return the object and release the lock
191			}
192			}
193			} else {
194			$queueObject = $filteredQuery->first();
195			self::remove_duplicates($queueObject->ID);
196			self::mark_as_regenerating($queueObject);
197			}
198			}
199			}
200
201			return $queueObject; //return the object or null
202			}
203
204			/**
205			* Finds the next most prioritized url that needs recaching
206			*
207			* @return string
208			*/
209			public static function get_next_url() {
210			$object = self::get_queue_object('stale');
211			if($object) return $object->URLSegment;
212
213			$interval = date('Y-m-d H:i:s', strtotime('-'.self::$minutes_until_force_regeneration.' minutes'));
214
215			// Find URLs that has been stuck in regeneration
216			$object = self::get_queue_object('regenerating', $interval);
217			if($object) return $object->URLSegment;
218
219			// Find URLs that is erronous and might work now (flush issues etc)
220			$object = self::get_queue_object('error', $interval);
221			if($object) return $object->URLSegment;
222
223			return '';
224			}
225
226			/**
227			* Removes the .html fresh copy of the cache.
228			* Keeps the *.stale.html copy in place,
229			* in order to notify the user of the stale content.
230			*
231			* @param array $URLSegments
232			*/
233			protected static function remove_old_cache( array $URLSegments ) {
234			$publisher = singleton('SiteTree')->getExtensionInstance('FilesystemPublisher');
235			if ($publisher) {
236			$paths = $publisher->urlsToPaths($URLSegments);
237			foreach($paths as $absolutePath) {
238
239			if(!file_exists($publisher->getDestDir().'/'.$absolutePath)) {
240			continue;
241			}
242
243			unlink($publisher->getDestDir().'/'.$absolutePath);
244			}
245			}
246			}
247
248			/**
249			* Mark this current StaticPagesQueue as a work in progress
250			*
251			* @param StaticPagesQueue $object
252			*/
253			protected static function mark_as_regenerating(StaticPagesQueue $object) {
254			$now = date('Y-m-d H:i:s');
255			DB::query('UPDATE "StaticPagesQueue" SET "LastEdited" = \''.$now.'\', "Freshness"=\'regenerating\' WHERE "ID" = '.$object->ID);
256			singleton(__CLASS__)->flushCache();
257			}
258
259			/**
260			* Removes all duplicates that has the same URLSegment as $ID
261			*
262			* @param int $ID - ID of the object whose duplicates we want to remove
263			* @return void
			0 ignored issues – show Documentation introduced 2016-03-22 06:43 UTC by Report Bug Copy Issue Report Should the return type not be `integer\|null`? This check compares the return type specified in the `@return` annotation of a function or method doc comment with the types returned by the function and raises an issue if they mismatch. Loading history...
264			*/
265			static function remove_duplicates( $ID ) {
			0 ignored issues – show Best Practice introduced 2015-12-02 22:54 UTC by Report Bug Copy Issue Report It is generally recommended to explicitly declare the visibility for methods. Adding explicit visibility (`private`, `protected`, or `public`) is generally recommend to communicate to other developers how, and from where this method is intended to be used. Loading history...
266			$obj = DataObject::get_by_id('StaticPagesQueue', $ID);
267			if(!$obj) return 0;
268			DB::query(
269			sprintf('DELETE FROM "StaticPagesQueue" WHERE "URLSegment" = \'%s\' AND "ID" != %d', $obj->URLSegment, (int)$ID)
270			);
271			}
272
273			/**
274			*
275			* @param string $url
276			* @param bool $onlyStale - Get only stale entries
			0 ignored issues – show Bug introduced 2015-12-02 22:54 UTC by Report Bug Copy Issue Report There is no parameter named `$onlyStale`. Was it maybe removed? This check looks for PHPDoc comments describing methods or function parameters that do not exist on the corresponding method or function. Consider the following example. The parameter `$italy` is not defined by the method `finale(...)`. /** * @param array $germany * @param array $island * @param array $italy */ function finale($germany, $island) { return "2:1"; } The most likely cause is that the parameter was removed, but the annotation was not. Loading history...
277			* @return DataObject \|\| false - The first item matching the query
			0 ignored issues – show Documentation introduced 2016-03-22 06:43 UTC by Report Bug Copy Issue Report Should the return type not be `false\|StaticPagesQueue`? This check compares the return type specified in the `@return` annotation of a function or method doc comment with the types returned by the function and raises an issue if they mismatch. Loading history...
278			*/
279			protected static function get_by_link($url) {
280			$filter = '"URLSegment" = \''.Convert::raw2sql($url).'\'';
281			$res = DB::query('SELECT * FROM "StaticPagesQueue" WHERE '.$filter.' LIMIT 1;');
282			if(!$res->numRecords()){
283			return false;
284			}
285			return new StaticPagesQueue($res->first());
286			}
287			}
288

silverstripe / silverstripe-staticpublishqueue

Push — master ( bb3e3e...12ae4c )

StaticPagesQueue::remove_duplicates() A

Complexity

Size

Duplication

Importance

1. Missing dependencies

2. Missing use statement

Available Fixes

Duplication Side-by-Side

Filter issues like