Reader::setDocumentContent() - Code Metrics - Inspection of "Allow to query new content" - spiderling-php/crawler - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( 0e62dc...fa05db )

by Ivan

created 2015-12-17 19:14 UTC

Reader::setDocumentContent() A

↳ Parent: Reader

Complexity

Conditions	1
Paths	1

Size

Total Lines	7
Code Lines	4

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	4
CRAP Score	1

Importance

Changes	1
Bugs	0	Features	0

Metric	Value
c	1
b	0
f	0
dl	0
loc	7
ccs	4
cts	4
cp	1
rs	9.4286
cc	1
eloc	4
nc	1
nop	1
crap	1

<?php

namespace SP\Crawler;

use SP\Spiderling\CrawlerInterface;
use SP\Crawler\Element\ClickRequestInterface;
use SP\Crawler\Element\ClickableInterface;
use SP\Crawler\Element\SelectableInterface;
use Psr\Http\Message\RequestInterface;
use SP\Spiderling\Query\AbstractQuery;
use GuzzleHttp\Psr7\Uri;
use Psr\Http\Message\UriInterface;
use DOMDocument;
use DOMElement;
use DOMXPath;
use InvalidArgumentException;
use BadMethodCallException;

/**
 * @author    Ivan Kerin <[email protected]>
 * @copyright 2015, Clippings Ltd.
 * @license   http://spdx.org/licenses/BSD-3-Clause
 */
class Reader implements CrawlerInterface
{
    /**
     * @var DOMDocument
     */
    private $document;

    /**
     * @var DOMXPath
     */
    private $xpath;

    /**
     * @var ElementMap
     */
    private $inputMap;

    /**
     * @param DOMDocument $document
     */
    public function __construct(DOMDocument $document)
    {
        $this->document = $document;

        $this->xpath = new DOMXPath($document);

        $this->inputMap = new InputMap($this);

    }

    /**
     * @param  string $content
     * @return self
     */
    public function setDocumentContent($content)
    {
        $this->document->loadHtml((string) $content);
        $this->xpath = new DOMXPath($this->document);

        return $this;
    }

    /**
     * @return DOMDocument
     */
    public function getDocument()
    {
        return $this->document;
    }

    /**
     * @return DOMXPath
     */
    public function getXPath()
    {
        return $this->xpath;
    }

    /**
     * @return ElementMap
     */
    public function getInputMap()
    {
        return $this->inputMap;
    }

    /**
     * @param  string $id
     * @throws BadMethodCallException
     */
    public function click($id)
    {
        $input = $this->getInput($this->getElement($id));

        if ($input instanceof ClickableInterface) {
            $input->click();
        } elseif ($input instanceof ClickRequestInterface) {
            $request = $input->clickRequest();
            $this->sendRequest($request);
        } else {
            throw new BadMethodCallException(
                sprintf('Cannot click on %s, %s', get_class($input), $id)
            );
        }
    }

    /**
     * @param  string $id
     * @throws BadMethodCallException
     */
    public function select($id)
    {
        $input = $this->getInput($this->getElement($id));

        if ($input instanceof SelectableInterface) {
            $input->select();
        } else {
            throw new BadMethodCallException(
                sprintf('Cannot select on %s, %s', get_class($input), $id)
            );
        }
    }

    /**
     * @param  RequestInterface $input
/**
 * @param array $germany
 * @param array $island
 * @param array $italy
 */
function finale($germany, $island) {
    return "2:1";
}
     * @throws BadMethodCallException
     */
    public function sendRequest(RequestInterface $request)
    {
        throw new BadMethodCallException(
            sprintf('Cannot send request to %s', $request->getUri())
        );
    }

    /**
     * @param  string $url
     * @throws BadMethodCallException
     */
    public function open(UriInterface $url)
    {
        throw new BadMethodCallException(
            sprintf('Method %s not supported by %s', __METHOD__, __CLASS__)
        );
    }

    /**
     * @return Psr\Http\Message\UriInterface
     */
    public function getUri()
    {
        return new Uri('');
    }

    /**
     * @param  string          $xpath
     * @param  DOMElement|null $scope
     * @return DOMNodeList
     */
    public function query($xpath, DOMElement $scope = null)
    {
        return $this->getXpath()->query($xpath, $scope);
    }

    /**
     * @param  string $xpath
     * @throws InvalidArgumentException when id not found
     * @return DOMElement
     */
    public function getElement($xpath)
    {
        $items = $this->query($xpath);

        if (0 === $items->length) {
            throw new InvalidArgumentException(
                sprintf('Node with id %s does not exist', $xpath)
            );
        }

        return $items->item(0);
    }

    /**
     * @param  string $id
     * @throws InvalidArgumentException when id not found
     * @return string
     */
    public function getText($id)
    {
        $element = $this->getElement($id);

        return trim(preg_replace('/[ \s\f\n\r\t\v ]+/u', ' ', $element->textContent));
    }

    /**
     * @param  string $id
     * @throws InvalidArgumentException when id not found
     * @return string
     */
    public function getTagName($id)
    {
        return $this->getElement($id)->tagName;
    }

    /**
     * @param  string $id
     * @param  string $name
     * @throws InvalidArgumentException when id not found
     * @return string
     */
    public function getAttribute($id, $name)
    {
        return $this->getElement($id)->getAttribute($name);
    }

    /**
     * @param  string $id
     * @throws InvalidArgumentException when id not found
     * @return string
     */
    public function getHtml($id)
    {
        return $this->document->saveXml($this->getElement($id));
    }

    /**
     * @return string
     */
    public function getFullHtml()
    {
        return $this->document->saveHtml();
    }

    /**
     * @param  string $id
     * @throws InvalidArgumentException when id not found
     * @return boolean
     */
    public function isVisible($id)
    {
        $element = $this->getElement($id);

        $conditions = [
            "contains(@style, 'display:none')",
            "contains(@style, 'display: none')",
            "self::script",
            "self::head",
        ];

        $hidden = $this->xpath->query(
            './ancestor-or-self::*['.join(' or ', $conditions).']',
            $element
        );

        return $hidden->length == 0;
    }

    /**
     * @param  string $id
     * @throws InvalidArgumentException when id not found
     * @return boolean
     */
    public function isSelected($id)
    {
        return $this->getElement($id)->hasAttribute('selected');
    }

    /**
     * @param  string $id
     * @throws InvalidArgumentException when id not found
     * @return boolean
     */
    public function isChecked($id)
    {
        return $this->getElement($id)->hasAttribute('checked');
    }

    /**
     * @param  DOMElement $element
     * @throws InvalidArgumentException when id not found
     * @return Element\AbstractElement
     */
    public function getInput(DOMElement $element)
    {
        return $this->inputMap->get($element);
    }

    /**
     * @param  string $id
     * @throws InvalidArgumentException when id not found
     * @return mixed
     */
    public function getValue($id)
    {
        return $this->getInput($this->getElement($id))->getValue();
abstract class User
{
    /** @return string */
    abstract public function getPassword();
}

class MyUser extends User
{
    public function getPassword()
    {
        // return something
    }

    public function getDisplayName()
    {
        // return some name.
    }
}

class AuthSystem
{
    public function authenticate(User $user)
    {
        $this->logger->info(sprintf('Authenticating %s.', $user->getDisplayName()));
        // do something.
    }
}
    }

    /**
     * @param  string $id
     * @param  string $value
     * @throws InvalidArgumentException when id not found
     */
    public function setValue($id, $value)
    {
        $input = $this->getInput($this->getElement($id));

        if (false === $input->isDisabled()) {
            $input->setValue($value);
        }
    }

    /**
     * @param  AbstractQuery $query
     * @param  string        $parent
     * @return array
     */
    public function queryIds(AbstractQuery $query, $parent = null)
    {
        $xpath = $parent.$query->getXPath();

        $ids = [];

        foreach ($this->query($xpath) as $index => $element) {
            $ids []= "($xpath)[".($index+1)."]";
        }

        return $query->getFilters()->matchAll($this, $ids);
    }
}


1		<?php
2
3		namespace SP\Crawler;
4
5		use SP\Spiderling\CrawlerInterface;
6		use SP\Crawler\Element\ClickRequestInterface;
7		use SP\Crawler\Element\ClickableInterface;
8		use SP\Crawler\Element\SelectableInterface;
9		use Psr\Http\Message\RequestInterface;
10		use SP\Spiderling\Query\AbstractQuery;
11		use GuzzleHttp\Psr7\Uri;
12		use Psr\Http\Message\UriInterface;
13		use DOMDocument;
14		use DOMElement;
15		use DOMXPath;
16		use InvalidArgumentException;
17		use BadMethodCallException;
18
19		/**
20		* @author Ivan Kerin <[email protected]>
21		* @copyright 2015, Clippings Ltd.
22		* @license http://spdx.org/licenses/BSD-3-Clause
23		*/
24		class Reader implements CrawlerInterface
25		{
26		/**
27		* @var DOMDocument
28		*/
29		private $document;
30
31		/**
32		* @var DOMXPath
33		*/
34		private $xpath;
35
36		/**
37		* @var ElementMap
38		*/
39		private $inputMap;
40
41		/**
42		* @param DOMDocument $document
43		*/
44	1	public function __construct(DOMDocument $document)
45		{
46	1	$this->document = $document;
47
48	1	$this->xpath = new DOMXPath($document);
49
50	1	$this->inputMap = new InputMap($this);
		0 ignored issues – show Documentation Bug introduced 2015-11-28 10:37 UTC by Report Bug Copy Issue Report It seems like `new \SP\Crawler\InputMap($this)` of type `object<SP\Crawler\InputMap>` is incompatible with the declared type `object<SP\Crawler\ElementMap>` of property `$inputMap`. 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...
51	1	}
52
53		/**
54		* @param string $content
55		* @return self
56		*/
57	1	public function setDocumentContent($content)
58		{
59	1	$this->document->loadHtml((string) $content);
60	1	$this->xpath = new DOMXPath($this->document);
61
62	1	return $this;
63		}
64
65		/**
66		* @return DOMDocument
67		*/
68	1	public function getDocument()
69		{
70	1	return $this->document;
71		}
72
73		/**
74		* @return DOMXPath
75		*/
76	1	public function getXPath()
77		{
78	1	return $this->xpath;
79		}
80
81		/**
82		* @return ElementMap
83		*/
84	1	public function getInputMap()
85		{
86	1	return $this->inputMap;
87		}
88
89		/**
90		* @param string $id
91		* @throws BadMethodCallException
92		*/
93	3	public function click($id)
94		{
95	3	$input = $this->getInput($this->getElement($id));
96
97	3	if ($input instanceof ClickableInterface) {
98	1	$input->click();
99	3	} elseif ($input instanceof ClickRequestInterface) {
100	1	$request = $input->clickRequest();
101	1	$this->sendRequest($request);
102	1	} else {
103	1	throw new BadMethodCallException(
104	1	sprintf('Cannot click on %s, %s', get_class($input), $id)
105	1	);
106		}
107	2	}
108
109		/**
110		* @param string $id
111		* @throws BadMethodCallException
112		*/
113	2	public function select($id)
114		{
115	2	$input = $this->getInput($this->getElement($id));
116
117	2	if ($input instanceof SelectableInterface) {
118	1	$input->select();
119	1	} else {
120	1	throw new BadMethodCallException(
121	1	sprintf('Cannot select on %s, %s', get_class($input), $id)
122	1	);
123		}
124	1	}
125
126		/**
127		* @param RequestInterface $input
		0 ignored issues – show Bug introduced 2015-11-28 10:37 UTC by Report Bug Copy Issue Report There is no parameter named `$input`. 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...
128		* @throws BadMethodCallException
129		*/
130	1	public function sendRequest(RequestInterface $request)
131		{
132	1	throw new BadMethodCallException(
133	1	sprintf('Cannot send request to %s', $request->getUri())
134	1	);
135		}
136
137		/**
138		* @param string $url
139		* @throws BadMethodCallException
140		*/
141	1	public function open(UriInterface $url)
142		{
143	1	throw new BadMethodCallException(
144	1	sprintf('Method %s not supported by %s', __METHOD__, __CLASS__)
145	1	);
146		}
147
148		/**
149		* @return Psr\Http\Message\UriInterface
150		*/
151	1	public function getUri()
152		{
153	1	return new Uri('');
154		}
155
156		/**
157		* @param string $xpath
158		* @param DOMElement\|null $scope
159		* @return DOMNodeList
160		*/
161	1	public function query($xpath, DOMElement $scope = null)
162		{
163	1	return $this->getXpath()->query($xpath, $scope);
164		}
165
166		/**
167		* @param string $xpath
168		* @throws InvalidArgumentException when id not found
169		* @return DOMElement
170		*/
171	1	public function getElement($xpath)
172		{
173	1	$items = $this->query($xpath);
174
175	1	if (0 === $items->length) {
176	1	throw new InvalidArgumentException(
177	1	sprintf('Node with id %s does not exist', $xpath)
178	1	);
179		}
180
181	1	return $items->item(0);
182		}
183
184		/**
185		* @param string $id
186		* @throws InvalidArgumentException when id not found
187		* @return string
188		*/
189	1	public function getText($id)
190		{
191	1	$element = $this->getElement($id);
192
193	1	return trim(preg_replace('/[ \s\f\n\r\t\v ]+/u', ' ', $element->textContent));
194		}
195
196		/**
197		* @param string $id
198		* @throws InvalidArgumentException when id not found
199		* @return string
200		*/
201	1	public function getTagName($id)
202		{
203	1	return $this->getElement($id)->tagName;
204		}
205
206		/**
207		* @param string $id
208		* @param string $name
209		* @throws InvalidArgumentException when id not found
210		* @return string
211		*/
212	1	public function getAttribute($id, $name)
213		{
214	1	return $this->getElement($id)->getAttribute($name);
215		}
216
217		/**
218		* @param string $id
219		* @throws InvalidArgumentException when id not found
220		* @return string
221		*/
222	1	public function getHtml($id)
223		{
224	1	return $this->document->saveXml($this->getElement($id));
225		}
226
227		/**
228		* @return string
229		*/
230	1	public function getFullHtml()
231		{
232	1	return $this->document->saveHtml();
233		}
234
235		/**
236		* @param string $id
237		* @throws InvalidArgumentException when id not found
238		* @return boolean
239		*/
240	3	public function isVisible($id)
241		{
242	3	$element = $this->getElement($id);
243
244		$conditions = [
245	3	"contains(@style, 'display:none')",
246	3	"contains(@style, 'display: none')",
247	3	"self::script",
248	3	"self::head",
249	3	];
250
251	3	$hidden = $this->xpath->query(
252	3	'./ancestor-or-self::*['.join(' or ', $conditions).']',
253		$element
254	3	);
255
256	3	return $hidden->length == 0;
257		}
258
259		/**
260		* @param string $id
261		* @throws InvalidArgumentException when id not found
262		* @return boolean
263		*/
264	4	public function isSelected($id)
265		{
266	4	return $this->getElement($id)->hasAttribute('selected');
267		}
268
269		/**
270		* @param string $id
271		* @throws InvalidArgumentException when id not found
272		* @return boolean
273		*/
274	6	public function isChecked($id)
275		{
276	6	return $this->getElement($id)->hasAttribute('checked');
277		}
278
279		/**
280		* @param DOMElement $element
281		* @throws InvalidArgumentException when id not found
282		* @return Element\AbstractElement
283		*/
284	9	public function getInput(DOMElement $element)
285		{
286	9	return $this->inputMap->get($element);
287		}
288
289		/**
290		* @param string $id
291		* @throws InvalidArgumentException when id not found
292		* @return mixed
293		*/
294	1	public function getValue($id)
295		{
296	1	return $this->getInput($this->getElement($id))->getValue();
		0 ignored issues – show Bug introduced 2015-11-28 10:37 UTC by Report Bug Copy Issue Report It seems like you code against a specific sub-type and not the parent class `SP\Crawler\Element\AbstractElement` as the method `getValue()` does only exist in the following sub-classes of `SP\Crawler\Element\AbstractElement`: `SP\Crawler\Element\Checkbox`, `SP\Crawler\Element\File`, `SP\Crawler\Element\Input`, `SP\Crawler\Element\Option`, `SP\Crawler\Element\Radio`, `SP\Crawler\Element\Select`, `SP\Crawler\Element\Textarea`. Maybe you want to `instanceof` check for one of these explicitly? Let’s take a look at an example: abstract class User { /** @return string / abstract public function getPassword(); } class MyUser extends User { public function getPassword() { // return something } public function getDisplayName() { // return some name. } } class AuthSystem { public function authenticate(User $user) { $this->logger->info(sprintf('Authenticating %s.', $user->getDisplayName())); // do something. } } In the above example, the authenticate() method works fine as long as you just pass instances of MyUser. However, if you now also want to pass a different sub-classes of User which does not have a getDisplayName() method, the code will break. Available Fixes Change the type-hint for the parameter: class AuthSystem { public function authenticate(MyUser $user) { / ... / } } Add an additional type-check: class AuthSystem { public function authenticate(User $user) { if ($user instanceof MyUser) { $this->logger->info(/* ... /); } // or alternatively if ( ! $user instanceof MyUser) { throw new \LogicException( '$user must be an instance of MyUser, ' .'other instances are not supported.' ); } } } Note:* PHP Analyzer uses reverse abstract interpretation to narrow down the types inside the if block in such a case. Add the method to the parent class: abstract class User { /** @return string / abstract public function getPassword(); /* @return string */ abstract public function getDisplayName(); } Loading history...
297		}
298
299		/**
300		* @param string $id
301		* @param string $value
302		* @throws InvalidArgumentException when id not found
303		*/
304	1	public function setValue($id, $value)
305		{
306	1	$input = $this->getInput($this->getElement($id));
307
308	1	if (false === $input->isDisabled()) {
309	1	$input->setValue($value);
310	1	}
311	1	}
312
313		/**
314		* @param AbstractQuery $query
315		* @param string $parent
316		* @return array
317		*/
318	1	public function queryIds(AbstractQuery $query, $parent = null)
319		{
320	1	$xpath = $parent.$query->getXPath();
321
322	1	$ids = [];
323
324	1	foreach ($this->query($xpath) as $index => $element) {
325	1	$ids []= "($xpath)[".($index+1)."]";
326	1	}
327
328	1	return $query->getFilters()->matchAll($this, $ids);
329		}
330		}
331

spiderling-php / crawler

Push — master ( 0e62dc...fa05db )

Reader::setDocumentContent() A

Complexity

Size

Duplication

Code Coverage

Importance

Available Fixes

Duplication Side-by-Side

Filter issues like