UserAgentClient::isUrlApplicable() - Code Metrics - Inspection of "Refactoring" - VIPnytt/RobotsTxtParser - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Branch — 2.0-dev (d250b8)

by Jan-Petter

created 2016-05-29 19:53 UTC

UserAgentClient::isUrlApplicable() B

↳ Parent: UserAgentClient

Complexity

Conditions	5
Paths	7

Size

Total Lines	14
Code Lines	10

Duplication

Lines	0
Ratio	0 %

Importance

Changes	1
Bugs	0	Features	0

Metric	Value
c	1
b	0
f	0
dl	0
loc	14
rs	8.8571
cc	5
eloc	10
nc	7
nop	1

<?php
namespace vipnytt\RobotsTxtParser\Client\Directives;

use vipnytt\RobotsTxtParser\Exceptions\ClientException;
use vipnytt\RobotsTxtParser\Parser\Directives\UserAgentParser;
use vipnytt\RobotsTxtParser\Parser\StatusCodeParser;
use vipnytt\RobotsTxtParser\Parser\UrlParser;
use vipnytt\RobotsTxtParser\RobotsTxtInterface;
use vipnytt\UserAgentParser as UAStringParser;

/**
 * Class UserAgentClient
 *
 * @package vipnytt\RobotsTxtParser\Client\Directives
 */
class UserAgentClient implements RobotsTxtInterface
{
    use UrlParser;

    /**
     * Rules
     * @var array
     */
    private $rules;

    /**
     * User-agent
     * @var string
     */
    private $userAgent;

    /**
     * Origin user-agent
     * @var string
     */
    private $userAgentOrigin;

    /**
     * Robots.txt base URL
     * @var string
     */
    private $base;

    /**
     * Status code parser
     * @var StatusCodeParser
     */
    private $statusCodeParser;

    /**
     * Comment export status
     * @var bool
     */
    private $commentsExported = false;

    /**
     * UserAgentClient constructor.
     *
     * @param string $userAgent
     * @param UserAgentParser $rules
     * @param string $baseUri
     * @param int|null $statusCode
     */
    public function __construct($userAgent, UserAgentParser $rules, $baseUri, $statusCode)
    {
        $this->statusCodeParser = new StatusCodeParser($statusCode, parse_url($baseUri, PHP_URL_SCHEME));
        $this->rules = $rules;

        $this->base = $baseUri;
        $this->userAgentOrigin = mb_strtolower($userAgent);
        $userAgentParser = new UAStringParser($this->userAgentOrigin);
        if (($this->userAgent = $userAgentParser->match($rules->userAgents)) === false) {
class Id
{
    public $id;

    public function __construct($id)
    {
        $this->id = $id;
    }

}

class Account
{
    /** @var  Id $id */
    public $id;
}

$account_id = false;

if (starsAreRight()) {
    $account_id = new Id(42);
}

$account = new Account();
if ($account instanceof Id)
{
    $account->id = $account_id;
}
            $this->userAgent = self::USER_AGENT;
        }
    }

    /**
     * Check if URL is allowed to crawl
     *
     * @param string $url
     * @return bool
     */
    public function isAllowed($url)
    {
        return $this->check(self::DIRECTIVE_ALLOW, $url);
    }

    /**
     * Check
     *
     * @param string $directive
     * @param string $url - URL to check
     * @return bool
     * @throws ClientException
     */
    private function check($directive, $url)
    {
        $url = $this->urlConvertToFull($url, $this->base);
        if (!$this->isUrlApplicable([$url, $this->base])) {
            throw new ClientException('URL belongs to a different robots.txt');
        }
        $this->statusCodeParser->replaceUnofficial();
        if (($result = $this->statusCodeParser->check()) !== null) {
            return $directive === $result;
        }
        $result = self::DIRECTIVE_ALLOW;
        foreach (
            [
                self::DIRECTIVE_DISALLOW => $this->rules->disallow[$this->userAgent],
                self::DIRECTIVE_ALLOW => $this->rules->allow[$this->userAgent]
            ] as $currentDirective => $currentRules
        ) {
            if ($currentRules->check($url)) {
                $result = $currentDirective;
            }
        }
        return $directive === $result;
    }

    /**
     * Check if the URL belongs to current robots.txt
     *
     * @param $urls
     * @return bool
     */
    private function isUrlApplicable($urls)
    {
        foreach ($urls as $url) {
            $parsed = parse_url($url);
            $parsed['port'] = is_int($port = parse_url($url, PHP_URL_PORT)) ? $port : getservbyname($parsed['scheme'], 'tcp');
            $assembled = $parsed['scheme'] . '://' . $parsed['host'] . ':' . $parsed['port'];
            if (!isset($result)) {
                $result = $assembled;
            } elseif ($result !== $assembled) {
                return false;
            }
        }
        return true;
    }

    /**
     * Check if URL is disallowed to crawl
     *
     * @param string $url
     * @return bool
     */
    public function isDisallowed($url)
    {
        return $this->check(self::DIRECTIVE_DISALLOW, $url);
    }

    /**
     * Cache-delay
     *
     * @return DelayClient
     */
    public function cacheDelay()
    {
        $export = $this->rules->cacheDelay[$this->userAgent]->export();
        $delay = isset($export[self::DIRECTIVE_CACHE_DELAY]) ? $export[self::DIRECTIVE_CACHE_DELAY] : 0;
        return new DelayClient($this->base, $this->userAgent, $delay, $this->crawlDelay()->get());
    }

    /**
     * Crawl-delay
     *
     * @return DelayClient
     */
    public function crawlDelay()
    {
        $export = $this->rules->crawlDelay[$this->userAgent]->export();
        $delay = isset($export[self::DIRECTIVE_CRAWL_DELAY]) ? $export[self::DIRECTIVE_CRAWL_DELAY] : 0;
        return new DelayClient($this->base, $this->userAgent, $delay, $this->requestRate()->get());
    }

    /**
     * RequestClient-rate
     *
     * @return RequestRateClient
     */
    public function requestRate()
    {
        $array = $this->rules->requestRate[$this->userAgent]->export();
        $rates = isset($array[self::DIRECTIVE_REQUEST_RATE]) ? $array[self::DIRECTIVE_REQUEST_RATE] : [];
        return new RequestRateClient($this->base, $this->userAgent, $rates);
    }

    /**
     * Robot-version
     *
     * @return RobotVersionClient
     */
    public function robotVersion()
    {
        $export = $this->rules->robotVersion[$this->userAgent]->export();
        return new RobotVersionClient(isset($export[self::DIRECTIVE_ROBOT_VERSION]) ? $export[self::DIRECTIVE_ROBOT_VERSION] : null);
    }

    /**
     * Rule export
     *
     * @return array
     */
    public function export()
    {
        return array_merge(
            $this->rules->allow[$this->userAgent]->export(),
            $this->rules->comment[$this->userAgent]->export(),
            $this->rules->cacheDelay[$this->userAgent]->export(),
            $this->rules->crawlDelay[$this->userAgent]->export(),
            $this->rules->disallow[$this->userAgent]->export(),
            $this->rules->requestRate[$this->userAgent]->export(),
            $this->rules->robotVersion[$this->userAgent]->export(),
            $this->rules->visitTime[$this->userAgent]->export()
        );
    }

    /**
     * Visit-time
     *
     * @return VisitTimeClient
     */
    public function visitTime()
    {
        $export = $this->rules->visitTime[$this->userAgent]->export();
        $times = isset($export[self::DIRECTIVE_VISIT_TIME]) ? $export[self::DIRECTIVE_VISIT_TIME] : [];
        return new VisitTimeClient($times);
    }

    /**
     * UserAgentClient destructor.
     */
    public function __destruct()
    {
        if (!$this->commentsExported && $this->userAgent != self::USER_AGENT) {
            // Comment from the `Comments` directive exists, but has not been read.
            foreach ($this->comment()->export() as $message) {
                trigger_error($this->userAgent . ' @ ' . $this->base . self::PATH . ': ' . $message, E_USER_NOTICE);
            }
        }
    }

    /**
     * Comment
     *
     * @return CommentClient
     */
    public function comment()
    {
        $this->commentsExported = true;
        $export = $this->rules->comment[$this->userAgent]->export();
        $comments = isset($export[self::DIRECTIVE_COMMENT]) ? $export[self::DIRECTIVE_COMMENT] : [];
        return new CommentClient($comments);
    }
}


1			<?php
2			namespace vipnytt\RobotsTxtParser\Client\Directives;
3
4			use vipnytt\RobotsTxtParser\Exceptions\ClientException;
5			use vipnytt\RobotsTxtParser\Parser\Directives\UserAgentParser;
6			use vipnytt\RobotsTxtParser\Parser\StatusCodeParser;
7			use vipnytt\RobotsTxtParser\Parser\UrlParser;
8			use vipnytt\RobotsTxtParser\RobotsTxtInterface;
9			use vipnytt\UserAgentParser as UAStringParser;
10
11			/**
12			* Class UserAgentClient
13			*
14			* @package vipnytt\RobotsTxtParser\Client\Directives
15			*/
16			class UserAgentClient implements RobotsTxtInterface
17			{
18			use UrlParser;
19
20			/**
21			* Rules
22			* @var array
23			*/
24			private $rules;
25
26			/**
27			* User-agent
28			* @var string
29			*/
30			private $userAgent;
31
32			/**
33			* Origin user-agent
34			* @var string
35			*/
36			private $userAgentOrigin;
37
38			/**
39			* Robots.txt base URL
40			* @var string
41			*/
42			private $base;
43
44			/**
45			* Status code parser
46			* @var StatusCodeParser
47			*/
48			private $statusCodeParser;
49
50			/**
51			* Comment export status
52			* @var bool
53			*/
54			private $commentsExported = false;
55
56			/**
57			* UserAgentClient constructor.
58			*
59			* @param string $userAgent
60			* @param UserAgentParser $rules
61			* @param string $baseUri
62			* @param int\|null $statusCode
63			*/
64			public function __construct($userAgent, UserAgentParser $rules, $baseUri, $statusCode)
65			{
66			$this->statusCodeParser = new StatusCodeParser($statusCode, parse_url($baseUri, PHP_URL_SCHEME));
67			$this->rules = $rules;
			0 ignored issues – show Documentation Bug introduced 2016-05-29 19:56 UTC by Report Bug Copy Issue Report It seems like `$rules` of type `object<vipnytt\RobotsTxt...ctives\UserAgentParser>` is incompatible with the declared type `array` of property `$rules`. Our type inference engine has found an assignment to a property that is incompatible with the declared type of that property. Either this assignment is in error or the assigned type should be added to the documentation/type hint for that property.. Loading history...
68			$this->base = $baseUri;
69			$this->userAgentOrigin = mb_strtolower($userAgent);
70			$userAgentParser = new UAStringParser($this->userAgentOrigin);
71			if (($this->userAgent = $userAgentParser->match($rules->userAgents)) === false) {
			0 ignored issues – show Documentation Bug introduced 2016-05-23 18:04 UTC by Report Bug Copy Issue Report It seems like `$userAgentParser->match($rules->userAgents)` can also be of type `false`. However, the property `$userAgent` is declared as type `string`. Maybe add an additional type check? Our type inference engine has found a suspicous assignment of a value to a property. This check raises an issue when a value that can be of a mixed type is assigned to a property that is type hinted more strictly. For example, imagine you have a variable `$accountId` that can either hold an Id object or false (if there is no account id yet). Your code now assigns that value to the `id` property of an instance of the `Account` class. This class holds a proper account, so the id value must no longer be false. Either this assignment is in error or a type check should be added for that assignment. class Id { public $id; public function __construct($id) { $this->id = $id; } } class Account { /** @var Id $id */ public $id; } $account_id = false; if (starsAreRight()) { $account_id = new Id(42); } $account = new Account(); if ($account instanceof Id) { $account->id = $account_id; } Loading history...
72			$this->userAgent = self::USER_AGENT;
73			}
74			}
75
76			/**
77			* Check if URL is allowed to crawl
78			*
79			* @param string $url
80			* @return bool
81			*/
82			public function isAllowed($url)
83			{
84			return $this->check(self::DIRECTIVE_ALLOW, $url);
85			}
86
87			/**
88			* Check
89			*
90			* @param string $directive
91			* @param string $url - URL to check
92			* @return bool
93			* @throws ClientException
94			*/
95			private function check($directive, $url)
96			{
97			$url = $this->urlConvertToFull($url, $this->base);
98			if (!$this->isUrlApplicable([$url, $this->base])) {
99			throw new ClientException('URL belongs to a different robots.txt');
100			}
101			$this->statusCodeParser->replaceUnofficial();
102			if (($result = $this->statusCodeParser->check()) !== null) {
103			return $directive === $result;
104			}
105			$result = self::DIRECTIVE_ALLOW;
106			foreach (
107			[
108			self::DIRECTIVE_DISALLOW => $this->rules->disallow[$this->userAgent],
109			self::DIRECTIVE_ALLOW => $this->rules->allow[$this->userAgent]
110			] as $currentDirective => $currentRules
111			) {
112			if ($currentRules->check($url)) {
113			$result = $currentDirective;
114			}
115			}
116			return $directive === $result;
117			}
118
119			/**
120			* Check if the URL belongs to current robots.txt
121			*
122			* @param $urls
123			* @return bool
124			*/
125			private function isUrlApplicable($urls)
126			{
127			foreach ($urls as $url) {
128			$parsed = parse_url($url);
129			$parsed['port'] = is_int($port = parse_url($url, PHP_URL_PORT)) ? $port : getservbyname($parsed['scheme'], 'tcp');
130			$assembled = $parsed['scheme'] . '://' . $parsed['host'] . ':' . $parsed['port'];
131			if (!isset($result)) {
132			$result = $assembled;
133			} elseif ($result !== $assembled) {
134			return false;
135			}
136			}
137			return true;
138			}
139
140			/**
141			* Check if URL is disallowed to crawl
142			*
143			* @param string $url
144			* @return bool
145			*/
146			public function isDisallowed($url)
147			{
148			return $this->check(self::DIRECTIVE_DISALLOW, $url);
149			}
150
151			/**
152			* Cache-delay
153			*
154			* @return DelayClient
155			*/
156			public function cacheDelay()
157			{
158			$export = $this->rules->cacheDelay[$this->userAgent]->export();
159			$delay = isset($export[self::DIRECTIVE_CACHE_DELAY]) ? $export[self::DIRECTIVE_CACHE_DELAY] : 0;
160			return new DelayClient($this->base, $this->userAgent, $delay, $this->crawlDelay()->get());
161			}
162
163			/**
164			* Crawl-delay
165			*
166			* @return DelayClient
167			*/
168			public function crawlDelay()
169			{
170			$export = $this->rules->crawlDelay[$this->userAgent]->export();
171			$delay = isset($export[self::DIRECTIVE_CRAWL_DELAY]) ? $export[self::DIRECTIVE_CRAWL_DELAY] : 0;
172			return new DelayClient($this->base, $this->userAgent, $delay, $this->requestRate()->get());
173			}
174
175			/**
176			* RequestClient-rate
177			*
178			* @return RequestRateClient
179			*/
180			public function requestRate()
181			{
182			$array = $this->rules->requestRate[$this->userAgent]->export();
183			$rates = isset($array[self::DIRECTIVE_REQUEST_RATE]) ? $array[self::DIRECTIVE_REQUEST_RATE] : [];
184			return new RequestRateClient($this->base, $this->userAgent, $rates);
185			}
186
187			/**
188			* Robot-version
189			*
190			* @return RobotVersionClient
191			*/
192			public function robotVersion()
193			{
194			$export = $this->rules->robotVersion[$this->userAgent]->export();
195			return new RobotVersionClient(isset($export[self::DIRECTIVE_ROBOT_VERSION]) ? $export[self::DIRECTIVE_ROBOT_VERSION] : null);
196			}
197
198			/**
199			* Rule export
200			*
201			* @return array
202			*/
203			public function export()
204			{
205			return array_merge(
206			$this->rules->allow[$this->userAgent]->export(),
207			$this->rules->comment[$this->userAgent]->export(),
208			$this->rules->cacheDelay[$this->userAgent]->export(),
209			$this->rules->crawlDelay[$this->userAgent]->export(),
210			$this->rules->disallow[$this->userAgent]->export(),
211			$this->rules->requestRate[$this->userAgent]->export(),
212			$this->rules->robotVersion[$this->userAgent]->export(),
213			$this->rules->visitTime[$this->userAgent]->export()
214			);
215			}
216
217			/**
218			* Visit-time
219			*
220			* @return VisitTimeClient
221			*/
222			public function visitTime()
223			{
224			$export = $this->rules->visitTime[$this->userAgent]->export();
225			$times = isset($export[self::DIRECTIVE_VISIT_TIME]) ? $export[self::DIRECTIVE_VISIT_TIME] : [];
226			return new VisitTimeClient($times);
227			}
228
229			/**
230			* UserAgentClient destructor.
231			*/
232			public function __destruct()
233			{
234			if (!$this->commentsExported && $this->userAgent != self::USER_AGENT) {
235			// Comment from the `Comments` directive exists, but has not been read.
236			foreach ($this->comment()->export() as $message) {
237			trigger_error($this->userAgent . ' @ ' . $this->base . self::PATH . ': ' . $message, E_USER_NOTICE);
238			}
239			}
240			}
241
242			/**
243			* Comment
244			*
245			* @return CommentClient
246			*/
247			public function comment()
248			{
249			$this->commentsExported = true;
250			$export = $this->rules->comment[$this->userAgent]->export();
251			$comments = isset($export[self::DIRECTIVE_COMMENT]) ? $export[self::DIRECTIVE_COMMENT] : [];
252			return new CommentClient($comments);
253			}
254			}
255

VIPnytt / RobotsTxtParser

Branch — 2.0-dev (d250b8)

UserAgentClient::isUrlApplicable() B

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like