UserAgentTest::extends_crawlers_as_string() - Code Metrics - Inspection of "Feature: Extend Crawlers data array" - JayBizzle/Crawler-Detect - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Pull Request — master (#328)

unknown

created 2019-05-18 16:11 UTC

UserAgentTest::extends_crawlers_as_string() A

↳ Parent: UserAgentTest

Complexity

Conditions	1
Paths	1

Size

Total Lines

Duplication

Lines	8
Ratio	100 %

Importance

Changes

Metric	Value
cc	1
nc	1
nop	0
dl	8
loc	8
rs	10
c	0
b	0
f	0

<?php

/*
 * This file is part of Crawler Detect - the web crawler detection library.
 *
 * (c) Mark Beech <[email protected]>
 *
 * This source file is subject to the MIT license that is bundled
 * with this source code in the file LICENSE.
 */

use PHPUnit\Framework\TestCase;
use Jaybizzle\CrawlerDetect\CrawlerDetect;
use Jaybizzle\CrawlerDetect\Fixtures\Crawlers;

class UserAgentTest extends TestCase
{
    protected $CrawlerDetect;

    public function setUp()
    {
        $this->CrawlerDetect = new CrawlerDetect();
    }

    /** @test */
    public function user_agents_are_bots()
    {
        $lines = file(__DIR__.'/crawlers.txt', FILE_IGNORE_NEW_LINES | FILE_SKIP_EMPTY_LINES);

        foreach ($lines as $line) {
            $test = $this->CrawlerDetect->isCrawler($line);
            $this->assertTrue($test, $line);
        }
    }

    /** @test */
    public function user_agents_are_devices()
    {
        $lines = file(__DIR__.'/devices.txt', FILE_IGNORE_NEW_LINES | FILE_SKIP_EMPTY_LINES);

        foreach ($lines as $line) {
            $test = $this->CrawlerDetect->isCrawler($line);
            $this->assertFalse($test, $line);
        }
    }

    /** @test */
    public function it_returns_correct_matched_bot_name()
    {
        $test = $this->CrawlerDetect->isCrawler('Mozilla/5.0 (iPhone; CPU iPhone OS 7_1 like Mac OS X) AppleWebKit (KHTML, like Gecko) Mobile (compatible; Yahoo Ad monitoring; https://help.yahoo.com/kb/yahoo-ad-monitoring-SLN24857.html)');
$myVar = 'Value';
$higher = false;

if (rand(1, 6) > 3) {
    $higher = true;
} else {
    $higher = false;
}

        $matches = $this->CrawlerDetect->getMatches();

        $this->assertEquals($this->CrawlerDetect->getMatches(), 'monitoring', $matches);
    }

    /** @test */
    public function it_returns_null_when_no_bot_detected()
    {
        $test = $this->CrawlerDetect->isCrawler('nothing to see here');
$myVar = 'Value';
$higher = false;

if (rand(1, 6) > 3) {
    $higher = true;
} else {
    $higher = false;
}

        $matches = $this->CrawlerDetect->getMatches();

        $this->assertEquals($this->CrawlerDetect->getMatches(), null, $matches);
    }

    /** @test */
    public function empty_user_agent()
    {
        $test = $this->CrawlerDetect->isCrawler('      ');

        $this->assertFalse($test);
    }

    /** @test */
    public function current_visitor()
    {
        $headers = (array) json_decode('{"DOCUMENT_ROOT":"\/home\/test\/public_html","GATEWAY_INTERFACE":"CGI\/1.1","HTTP_ACCEPT":"*\/*","HTTP_ACCEPT_ENCODING":"gzip, deflate","HTTP_CACHE_CONTROL":"no-cache","HTTP_CONNECTION":"Keep-Alive","HTTP_FROM":"bingbot(at)microsoft.com","HTTP_HOST":"www.test.com","HTTP_PRAGMA":"no-cache","HTTP_USER_AGENT":"Mozilla\/5.0 (compatible; bingbot\/2.0; +http:\/\/www.bing.com\/bingbot.htm)","PATH":"\/bin:\/usr\/bin","QUERY_STRING":"order=closingDate","REDIRECT_STATUS":"200","REMOTE_ADDR":"127.0.0.1","REMOTE_PORT":"3360","REQUEST_METHOD":"GET","REQUEST_URI":"\/?test=testing","SCRIPT_FILENAME":"\/home\/test\/public_html\/index.php","SCRIPT_NAME":"\/index.php","SERVER_ADDR":"127.0.0.1","SERVER_ADMIN":"[email protected]","SERVER_NAME":"www.test.com","SERVER_PORT":"80","SERVER_PROTOCOL":"HTTP\/1.1","SERVER_SIGNATURE":"","SERVER_SOFTWARE":"Apache","UNIQUE_ID":"Vx6MENRxerBUSDEQgFLAAAAAS","PHP_SELF":"\/index.php","REQUEST_TIME_FLOAT":1461619728.0705,"REQUEST_TIME":1461619728}');

        $cd = new CrawlerDetect($headers);

        $this->assertTrue($cd->isCrawler());
    }

    /** @test */
    public function user_agent_passed_via_contructor()
    {
        $cd = new CrawlerDetect(null, 'Mozilla/5.0 (iPhone; CPU iPhone OS 7_1 like Mac OS X) AppleWebKit (KHTML, like Gecko) Mobile (compatible; Yahoo Ad monitoring; https://help.yahoo.com/kb/yahoo-ad-monitoring-SLN24857.html)');

        $this->assertTrue($cd->isCrawler());
    }

    /** @test */
    public function http_from_header()
    {
        $headers = (array) json_decode('{"DOCUMENT_ROOT":"\/home\/test\/public_html","GATEWAY_INTERFACE":"CGI\/1.1","HTTP_ACCEPT":"*\/*","HTTP_ACCEPT_ENCODING":"gzip, deflate","HTTP_CACHE_CONTROL":"no-cache","HTTP_CONNECTION":"Keep-Alive","HTTP_FROM":"googlebot(at)googlebot.com","HTTP_HOST":"www.test.com","HTTP_PRAGMA":"no-cache","HTTP_USER_AGENT":"Mozilla\/5.0 (Macintosh; Intel Mac OS X 10_8_4) AppleWebKit\/537.36 (KHTML, like Gecko) Chrome\/28.0.1500.71 Safari\/537.36","PATH":"\/bin:\/usr\/bin","QUERY_STRING":"order=closingDate","REDIRECT_STATUS":"200","REMOTE_ADDR":"127.0.0.1","REMOTE_PORT":"3360","REQUEST_METHOD":"GET","REQUEST_URI":"\/?test=testing","SCRIPT_FILENAME":"\/home\/test\/public_html\/index.php","SCRIPT_NAME":"\/index.php","SERVER_ADDR":"127.0.0.1","SERVER_ADMIN":"[email protected]","SERVER_NAME":"www.test.com","SERVER_PORT":"80","SERVER_PROTOCOL":"HTTP\/1.1","SERVER_SIGNATURE":"","SERVER_SOFTWARE":"Apache","UNIQUE_ID":"Vx6MENRxerBUSDEQgFLAAAAAS","PHP_SELF":"\/index.php","REQUEST_TIME_FLOAT":1461619728.0705,"REQUEST_TIME":1461619728}');

        $cd = new CrawlerDetect($headers);

        $this->assertTrue($cd->isCrawler());
    }

    /** @test */
    public function the_regex_patterns_are_unique()
    {
        $crawlers = new Crawlers();

        $this->assertEquals(count($crawlers->getAll()), count(array_unique($crawlers->getAll())));
    }

    /** @test */
    public function there_are_no_regex_collisions()
    {
        $crawlers = new Crawlers();

        foreach ($crawlers->getAll() as $key1 => $regex) {
            foreach ($crawlers->getAll() as $key2 => $compare) {
                // Dont check this regex against itself
                if ($key1 != $key2) {
                    preg_match('/'.$regex.'/i', stripslashes($compare), $matches);

                    $this->assertEmpty($matches, $regex.' collided with '.$compare);
                }
            }
        }
    }

    /** @test */
    public function extends_crawlers()
    {
        $crawlers = new Crawlers();

        $crawlersCount = count($crawlers->getAll());
        $newUserAgents = [
            'some_user_agent',
            'some_other_user_agent'
        ];

        $crawlers->extend($newUserAgents);

        $this->assertNotEquals($crawlersCount, count($crawlers->getAll()));
        $this->assertEquals($crawlersCount + count($newUserAgents), count($crawlers->getAll()));
    }

    /** @test */
    public function extends_crawlers_as_string()
    {
        $cd = new CrawlerDetect(null, 'Mozilla/5.0 (iPhone; CPU iPhone OS 7_1 like Mac OS X) AppleWebKit (KHTML, like Gecko) Mobile (compatible; some_user_agent)');
        $this->assertFalse($cd->isCrawler());

        $cd->extendCrawlers('some_user_agent');
        $this->assertTrue($cd->isCrawler());
    }

    /** @test */
    public function extends_crawlers_as_array()
    {
        $cd = new CrawlerDetect(null, 'Mozilla/5.0 (iPhone; CPU iPhone OS 7_1 like Mac OS X) AppleWebKit (KHTML, like Gecko) Mobile (compatible; some_other_user_agent)');
        $this->assertFalse($cd->isCrawler());

        $cd->extendCrawlers(['some_user_agent', 'some_other_user_agent']);
        $this->assertTrue($cd->isCrawler());
    }
}


1		<?php
2
3		/*
4		* This file is part of Crawler Detect - the web crawler detection library.
5		*
6		* (c) Mark Beech <[email protected]>
7		*
8		* This source file is subject to the MIT license that is bundled
9		* with this source code in the file LICENSE.
10		*/
11
12		use PHPUnit\Framework\TestCase;
13		use Jaybizzle\CrawlerDetect\CrawlerDetect;
14		use Jaybizzle\CrawlerDetect\Fixtures\Crawlers;
15
16		class UserAgentTest extends TestCase
17		{
18		protected $CrawlerDetect;
19
20		public function setUp()
21		{
22		$this->CrawlerDetect = new CrawlerDetect();
23		}
24
25		/** @test */
26	View Code Duplication	public function user_agents_are_bots()
27		{
28		$lines = file(__DIR__.'/crawlers.txt', FILE_IGNORE_NEW_LINES \| FILE_SKIP_EMPTY_LINES);
29
30		foreach ($lines as $line) {
31		$test = $this->CrawlerDetect->isCrawler($line);
32		$this->assertTrue($test, $line);
33		}
34		}
35
36		/** @test */
37	View Code Duplication	public function user_agents_are_devices()
38		{
39		$lines = file(__DIR__.'/devices.txt', FILE_IGNORE_NEW_LINES \| FILE_SKIP_EMPTY_LINES);
40
41		foreach ($lines as $line) {
42		$test = $this->CrawlerDetect->isCrawler($line);
43		$this->assertFalse($test, $line);
44		}
45		}
46
47		/** @test */
48	View Code Duplication	public function it_returns_correct_matched_bot_name()
49		{
50		$test = $this->CrawlerDetect->isCrawler('Mozilla/5.0 (iPhone; CPU iPhone OS 7_1 like Mac OS X) AppleWebKit (KHTML, like Gecko) Mobile (compatible; Yahoo Ad monitoring; https://help.yahoo.com/kb/yahoo-ad-monitoring-SLN24857.html)');
		0 ignored issues – show Unused Code introduced 2016-01-14 12:35 UTC by Report Bug Copy Issue Report `$test` is not used, you could remove the assignment. This check looks for variable assignements that are either overwritten by other assignments or where the variable is not used subsequently. $myVar = 'Value'; $higher = false; if (rand(1, 6) > 3) { $higher = true; } else { $higher = false; } Both the `$myVar` assignment in line 1 and the `$higher` assignment in line 2 are dead. The first because `$myVar` is never used and the second because `$higher` is always overwritten for every possible time line. Loading history...
51
52		$matches = $this->CrawlerDetect->getMatches();
53
54		$this->assertEquals($this->CrawlerDetect->getMatches(), 'monitoring', $matches);
55		}
56
57		/** @test */
58	View Code Duplication	public function it_returns_null_when_no_bot_detected()
59		{
60		$test = $this->CrawlerDetect->isCrawler('nothing to see here');
		0 ignored issues – show Unused Code introduced 2016-07-18 20:42 UTC by Report Bug Copy Issue Report `$test` is not used, you could remove the assignment. This check looks for variable assignements that are either overwritten by other assignments or where the variable is not used subsequently. $myVar = 'Value'; $higher = false; if (rand(1, 6) > 3) { $higher = true; } else { $higher = false; } Both the `$myVar` assignment in line 1 and the `$higher` assignment in line 2 are dead. The first because `$myVar` is never used and the second because `$higher` is always overwritten for every possible time line. Loading history...
61
62		$matches = $this->CrawlerDetect->getMatches();
63
64		$this->assertEquals($this->CrawlerDetect->getMatches(), null, $matches);
65		}
66
67		/** @test */
68		public function empty_user_agent()
69		{
70		$test = $this->CrawlerDetect->isCrawler(' ');
71
72		$this->assertFalse($test);
73		}
74
75		/** @test */
76		public function current_visitor()
77		{
78		$headers = (array) json_decode('{"DOCUMENT_ROOT":"\/home\/test\/public_html","GATEWAY_INTERFACE":"CGI\/1.1","HTTP_ACCEPT":"\/","HTTP_ACCEPT_ENCODING":"gzip, deflate","HTTP_CACHE_CONTROL":"no-cache","HTTP_CONNECTION":"Keep-Alive","HTTP_FROM":"bingbot(at)microsoft.com","HTTP_HOST":"www.test.com","HTTP_PRAGMA":"no-cache","HTTP_USER_AGENT":"Mozilla\/5.0 (compatible; bingbot\/2.0; +http:\/\/www.bing.com\/bingbot.htm)","PATH":"\/bin:\/usr\/bin","QUERY_STRING":"order=closingDate","REDIRECT_STATUS":"200","REMOTE_ADDR":"127.0.0.1","REMOTE_PORT":"3360","REQUEST_METHOD":"GET","REQUEST_URI":"\/?test=testing","SCRIPT_FILENAME":"\/home\/test\/public_html\/index.php","SCRIPT_NAME":"\/index.php","SERVER_ADDR":"127.0.0.1","SERVER_ADMIN":"[email protected]","SERVER_NAME":"www.test.com","SERVER_PORT":"80","SERVER_PROTOCOL":"HTTP\/1.1","SERVER_SIGNATURE":"","SERVER_SOFTWARE":"Apache","UNIQUE_ID":"Vx6MENRxerBUSDEQgFLAAAAAS","PHP_SELF":"\/index.php","REQUEST_TIME_FLOAT":1461619728.0705,"REQUEST_TIME":1461619728}');
79
80		$cd = new CrawlerDetect($headers);
81
82		$this->assertTrue($cd->isCrawler());
83		}
84
85		/** @test */
86		public function user_agent_passed_via_contructor()
87		{
88		$cd = new CrawlerDetect(null, 'Mozilla/5.0 (iPhone; CPU iPhone OS 7_1 like Mac OS X) AppleWebKit (KHTML, like Gecko) Mobile (compatible; Yahoo Ad monitoring; https://help.yahoo.com/kb/yahoo-ad-monitoring-SLN24857.html)');
89
90		$this->assertTrue($cd->isCrawler());
91		}
92
93		/** @test */
94		public function http_from_header()
95		{
96		$headers = (array) json_decode('{"DOCUMENT_ROOT":"\/home\/test\/public_html","GATEWAY_INTERFACE":"CGI\/1.1","HTTP_ACCEPT":"\/","HTTP_ACCEPT_ENCODING":"gzip, deflate","HTTP_CACHE_CONTROL":"no-cache","HTTP_CONNECTION":"Keep-Alive","HTTP_FROM":"googlebot(at)googlebot.com","HTTP_HOST":"www.test.com","HTTP_PRAGMA":"no-cache","HTTP_USER_AGENT":"Mozilla\/5.0 (Macintosh; Intel Mac OS X 10_8_4) AppleWebKit\/537.36 (KHTML, like Gecko) Chrome\/28.0.1500.71 Safari\/537.36","PATH":"\/bin:\/usr\/bin","QUERY_STRING":"order=closingDate","REDIRECT_STATUS":"200","REMOTE_ADDR":"127.0.0.1","REMOTE_PORT":"3360","REQUEST_METHOD":"GET","REQUEST_URI":"\/?test=testing","SCRIPT_FILENAME":"\/home\/test\/public_html\/index.php","SCRIPT_NAME":"\/index.php","SERVER_ADDR":"127.0.0.1","SERVER_ADMIN":"[email protected]","SERVER_NAME":"www.test.com","SERVER_PORT":"80","SERVER_PROTOCOL":"HTTP\/1.1","SERVER_SIGNATURE":"","SERVER_SOFTWARE":"Apache","UNIQUE_ID":"Vx6MENRxerBUSDEQgFLAAAAAS","PHP_SELF":"\/index.php","REQUEST_TIME_FLOAT":1461619728.0705,"REQUEST_TIME":1461619728}');
97
98		$cd = new CrawlerDetect($headers);
99
100		$this->assertTrue($cd->isCrawler());
101		}
102
103		/** @test */
104		public function the_regex_patterns_are_unique()
105		{
106		$crawlers = new Crawlers();
107
108		$this->assertEquals(count($crawlers->getAll()), count(array_unique($crawlers->getAll())));
109		}
110
111		/** @test */
112		public function there_are_no_regex_collisions()
113		{
114		$crawlers = new Crawlers();
115
116		foreach ($crawlers->getAll() as $key1 => $regex) {
117		foreach ($crawlers->getAll() as $key2 => $compare) {
118		// Dont check this regex against itself
119		if ($key1 != $key2) {
120		preg_match('/'.$regex.'/i', stripslashes($compare), $matches);
121
122		$this->assertEmpty($matches, $regex.' collided with '.$compare);
123		}
124		}
125		}
126		}
127
128		/** @test */
129		public function extends_crawlers()
130		{
131		$crawlers = new Crawlers();
132
133		$crawlersCount = count($crawlers->getAll());
134		$newUserAgents = [
135		'some_user_agent',
136		'some_other_user_agent'
137		];
138
139		$crawlers->extend($newUserAgents);
140
141		$this->assertNotEquals($crawlersCount, count($crawlers->getAll()));
142		$this->assertEquals($crawlersCount + count($newUserAgents), count($crawlers->getAll()));
143		}
144
145		/** @test */
146	View Code Duplication	public function extends_crawlers_as_string()
147		{
148		$cd = new CrawlerDetect(null, 'Mozilla/5.0 (iPhone; CPU iPhone OS 7_1 like Mac OS X) AppleWebKit (KHTML, like Gecko) Mobile (compatible; some_user_agent)');
149		$this->assertFalse($cd->isCrawler());
150
151		$cd->extendCrawlers('some_user_agent');
152		$this->assertTrue($cd->isCrawler());
153		}
154
155		/** @test */
156	View Code Duplication	public function extends_crawlers_as_array()
157		{
158		$cd = new CrawlerDetect(null, 'Mozilla/5.0 (iPhone; CPU iPhone OS 7_1 like Mac OS X) AppleWebKit (KHTML, like Gecko) Mobile (compatible; some_other_user_agent)');
159		$this->assertFalse($cd->isCrawler());
160
161		$cd->extendCrawlers(['some_user_agent', 'some_other_user_agent']);
162		$this->assertTrue($cd->isCrawler());
163		}
164		}
165

JayBizzle / Crawler-Detect

Pull Request — master (#328)

UserAgentTest::extends_crawlers_as_string() A

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like