StaticPagesQueue::remove_old_cache() - Code Metrics - Inspection of "Converted to PSR-2" - silverstripe/silverstripe-staticpublishqueue - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Pull Request — master (#47)

by Helpful

created 2015-12-17 18:18 UTC

StaticPagesQueue::remove_old_cache() A

↳ Parent: StaticPagesQueue

Complexity

Conditions	4
Paths	4

Size

Total Lines	14
Code Lines	8

Duplication

Lines	0
Ratio	0 %

Metric	Value
dl	0
loc	14
rs	9.2
cc	4
eloc	8
nc	4
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
{

    /**
     *
     * @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 $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 (DB::getConn() instanceof MySQLDatabase) {   //locking currently only works on MySQL


                    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 $ID - ID of the object whose duplicates we want to remove
     * @return void
     */
    public static function remove_duplicates($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());
    }
}


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
13			{
14
15			/**
16			*
17			* @var array
18			*/
19			public static $create_table_options = array(
20			'MySQLDatabase' => 'ENGINE=InnoDB'
21			);
22
23			/**
24			*
25			* @var array
26			*/
27			public static $db = array(
28			'Priority' => 'Int',
29			'URLSegment' => 'Varchar(255)',
30			'Freshness' => "Enum('stale, regenerating, error', 'stale')"
31			);
32
33			/**
34			*
35			* @var array
36			*/
37			public static $defaults = array(
38			"Priority" => 3
39			);
40
41			/**
42			*
43			* @var array
44			*/
45			public static $default_sort = "\"Priority\"";
46
47			/**
48			* Sets database indexes
49			*
50			* @var array
51			*/
52			public static $indexes = array(
53			'freshness_priority_created' => '(Freshness, Priority, Created)',
54			);
55
56			/**
57			*
58			* @var boolean
59			*/
60			private static $realtime = false;
61
62			/**
63			*
64			* @var int
65			*/
66			protected static $minutes_until_force_regeneration = 1;
67
68			/**
69			*
70			* @var array
71			*/
72			protected static $insert_statements = array();
73
74			/**
75			*
76			* @var array
77			*/
78			protected static $urls = array();
79
80			/**
81			*
82			* @return bool
83			*/
84			public static function is_realtime()
85			{
86			return Config::inst()->get('StaticPagesQueue', 'realtime');
87			}
88
89			/**
90			*
91			* @param type $priority
92			* @param type $URLSegment
93			* @return type
94			*/
95			public static function add_to_queue($priority, $URLSegment)
96			{
97			$now = date("Y-m-d H:i:s");
98			self::$insert_statements[$URLSegment] = '(\''.$now.'\',\''.$now.'\', \''.Convert::raw2sql($priority).'\',\''.Convert::raw2sql($URLSegment).'\')';
99			self::$urls[md5($URLSegment)] = $URLSegment;
100			}
101
102			/**
103			* This will push all the currently cached insert statements to be pushed
104			* into the database
105			*
106			* @return void
107			*/
108			public static function push_urls_to_db()
109			{
110			foreach (self::$insert_statements as $stmt) {
111			$insertSQL = 'INSERT INTO "StaticPagesQueue" ("Created", "LastEdited", "Priority", "URLSegment") VALUES ' . $stmt;
112			DB::query($insertSQL);
113			}
114			self::remove_old_cache(self::$urls);
115			// Flush the cache so DataObject::get works correctly
116			if (!empty(self::$insert_statements) && DB::affectedRows()) {
			0 ignored issues – show Deprecated Code introduced 2015-12-02 22:54 UTC by Report Bug Copy Issue Report The method `DB::affectedRows()` has been deprecated with message: since version 4.0 Use DB::affected_rows instead This method has been deprecated. The supplier of the class has supplied an explanatory message. The explanatory message should give you some clue as to whether and when the method will be removed from the class and what other method or class to use instead. Loading history...
117			singleton(__CLASS__)->flushCache();
118			}
119			self::$insert_statements = array();
120			}
121
122			/**
123			* Remove an object by the url
124			*
125			* @param string $URLSegment
126			* @return bool - if there was an queue item removed
127			*
128			*/
129			public static function delete_by_link($URLSegment)
130			{
131			$object = self::get_by_link($URLSegment);
132			if (!$object) {
133			return false;
134			}
135
136			$object->delete();
137			unset($object);
138			return true;
139			}
140
141			/**
142			* Update the queue with the information that this url renders an error somehow
143			*
144			* @param string $url
145			*/
146			public static function has_error($url)
147			{
148			if (!$url) {
149			return;
150			}
151
152			$existingObject = self::get_by_link($url);
153			$existingObject->Freshness = 'error';
154			$existingObject->write();
155			}
156
157			/**
158			* Returns a single queue object according to a particular priority and freshness measure.
159			* This method removes any duplicates and makes the object as "regenerating", so other calls to this method
160			* don't grab the same object.
161			* If we are using MySQLDatabase with InnoDB, we do row-level locking when updating the dataobject to allow for
162			* distributed cache rebuilds
163			* @static
164			* @param $freshness
165			* @param $sortOrder
166			*/
167			protected static function get_queue_object($freshness, $interval = null, $sortOrder = array('Priority'=>'DESC', 'ID'=>'ASC'))
168			{
169			$className = __CLASS__;
170			$queueObject = null;
171			$filterQuery = array("Freshness" => $freshness);
172			if ($interval) {
173			$filterQuery["LastEdited:LessThan"] = $interval;
174			}
175
176			$query = self::get();
177			if ($query->Count() > 0) {
178			$offset = 0;
179			$filteredQuery = $query->filter($filterQuery)->sort($sortOrder);
180
181			if ($filteredQuery->Count() > 0) {
182			if (DB::getConn() instanceof MySQLDatabase) { //locking currently only works on MySQL
			0 ignored issues – show Deprecated Code introduced 2015-12-02 22:54 UTC by Report Bug Copy Issue Report The method `DB::getConn()` has been deprecated with message: since version 4.0 Use DB::get_conn instead This method has been deprecated. The supplier of the class has supplied an explanatory message. The explanatory message should give you some clue as to whether and when the method will be removed from the class and what other method or class to use instead. Loading history...
183
184			do {
185			$queueObject = $filteredQuery->limit(1, $offset)->first(); //get first item
186
187			if ($queueObject) {
188			$lockName = md5($queueObject->URLSegment . $className);
189			}
190			//try to locking the item's URL, keep trying new URLs until we find one that is free to lock
191			$offset++;
192			} while ($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 $x; } Loading history...
193
194			if ($queueObject) {
195			$lockSuccess = LockMySQL::getLock($lockName); //acquire a lock with the URL of the queue item we have just fetched
196			if ($lockSuccess) {
197			self::remove_duplicates($queueObject->ID); //remove any duplicates
198			self::mark_as_regenerating($queueObject); //mark as regenerating so nothing else grabs it
199			LockMySQL::releaseLock($lockName); //return the object and release the lock
200			}
201			}
202			} else {
203			$queueObject = $filteredQuery->first();
204			self::remove_duplicates($queueObject->ID);
205			self::mark_as_regenerating($queueObject);
206			}
207			}
208			}
209
210			return $queueObject; //return the object or null
211			}
212
213			/**
214			* Finds the next most prioritized url that needs recaching
215			*
216			* @return string
217			*/
218			public static function get_next_url()
219			{
220			$object = self::get_queue_object('stale');
221			if ($object) {
222			return $object->URLSegment;
223			}
224
225			$interval = date('Y-m-d H:i:s', strtotime('-'.self::$minutes_until_force_regeneration.' minutes'));
226
227			// Find URLs that has been stuck in regeneration
228			$object = self::get_queue_object('regenerating', $interval);
229			if ($object) {
230			return $object->URLSegment;
231			}
232
233			// Find URLs that is erronous and might work now (flush issues etc)
234			$object = self::get_queue_object('error', $interval);
235			if ($object) {
236			return $object->URLSegment;
237			}
238
239			return '';
240			}
241
242			/**
243			* Removes the .html fresh copy of the cache.
244			* Keeps the *.stale.html copy in place,
245			* in order to notify the user of the stale content.
246			*
247			* @param array $URLSegments
248			*/
249			protected static function remove_old_cache(array $URLSegments)
250			{
251			$publisher = singleton('SiteTree')->getExtensionInstance('FilesystemPublisher');
252			if ($publisher) {
253			$paths = $publisher->urlsToPaths($URLSegments);
254			foreach ($paths as $absolutePath) {
255			if (!file_exists($publisher->getDestDir().'/'.$absolutePath)) {
256			continue;
257			}
258
259			unlink($publisher->getDestDir().'/'.$absolutePath);
260			}
261			}
262			}
263
264			/**
265			* Mark this current StaticPagesQueue as a work in progress
266			*
267			* @param StaticPagesQueue $object
268			*/
269			protected static function mark_as_regenerating(StaticPagesQueue $object)
270			{
271			$now = date('Y-m-d H:i:s');
272			DB::query('UPDATE "StaticPagesQueue" SET "LastEdited" = \''.$now.'\', "Freshness"=\'regenerating\' WHERE "ID" = '.$object->ID);
273			singleton(__CLASS__)->flushCache();
274			}
275
276			/**
277			* Removes all duplicates that has the same URLSegment as $ID
278			*
279			* @param int $ID - ID of the object whose duplicates we want to remove
280			* @return void
281			*/
282			public static function remove_duplicates($ID)
283			{
284			$obj = DataObject::get_by_id('StaticPagesQueue', $ID);
285			if (!$obj) {
286			return 0;
287			}
288			DB::query(
289			sprintf('DELETE FROM "StaticPagesQueue" WHERE "URLSegment" = \'%s\' AND "ID" != %d', $obj->URLSegment, (int)$ID)
290			);
291			}
292
293			/**
294			*
295			* @param string $url
296			* @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, $island) { return "2:1"; } The most likely cause is that the parameter was removed, but the annotation was not. Loading history...
297			* @return DataObject \|\| false - The first item matching the query
298			*/
299			protected static function get_by_link($url)
300			{
301			$filter = '"URLSegment" = \''.Convert::raw2sql($url).'\'';
302			$res = DB::query('SELECT * FROM "StaticPagesQueue" WHERE '.$filter.' LIMIT 1;');
303			if (!$res->numRecords()) {
304			return false;
305			}
306			return new StaticPagesQueue($res->first());
307			}
308			}
309

silverstripe / silverstripe-staticpublishqueue

Pull Request — master (#47)

StaticPagesQueue::remove_old_cache() A

Complexity

Size

Duplication

Available Fixes

Duplication Side-by-Side

Filter issues like