GlSpellChecker::checkSentences() - Code Metrics - Inspection of "change docker port" - emmanuelroecker/php-spellchecker - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( 6aeed9...ac4824 )

by Emmanuel

created 2016-01-11 20:28 UTC

GlSpellChecker::checkSentences() C

↳ Parent: GlSpellChecker

Complexity

Conditions	7
Paths	3

Size

Total Lines	47
Code Lines	34

Duplication

Lines	0
Ratio	0 %

Importance

Changes	4
Bugs	0	Features	2

Metric	Value
c	4
b	0
f	2
dl	0
loc	47
rs	6.7273
cc	7
eloc	34
nc	3
nop	2

<?php
/**
 * Spell check html files
 *
 * PHP version 5.4
 *
 * @category  GLICER
 * @package   GlSpellChecker
 * @author    Emmanuel ROECKER
 * @author    Rym BOUCHAGOUR
 * @copyright 2015 GLICER
 * @license   MIT
 * @link      http://dev.glicer.com/
 *
 * Created : 04/05/15
 * File : GlSpellChecker.php
 *
 */

namespace GlSpellChecker;

use GlHtml\GlHtml;
use Symfony\Component\Process\Process;
use Symfony\Component\Finder\SplFileInfo;
use Symfony\Component\Finder\Finder;
use Symfony\Component\Yaml\Exception\ParseException;
use Symfony\Component\Yaml\Yaml;
use GuzzleHttp\Client;


/**
 * Class GlSpellChecker
 * @package GlSpellChecker
 */
class GlSpellChecker
{
    /**
     * @var int
     */
    private $languageToolServerPort = 8081;

    /**
     * @var string
     */
    private $languageToolLanguage = 'fr';

    /**
     * @var Process $languagetoolServer ;
     */
    private $languagetoolServer = null;

    /**
     * @var Client
     */
    private $languagetoolClientHttp;


    /**
     * @var string
     */
    private $enchantLanguage = "fr_FR";
    private $enchantDictionnary = null;
    private $enchantBroker = null;

    /**
     * @param string $languageToolDirectory
     * @param string $languageToolLanguage
     * @param string $enchantLanguage
     * @param string $languageToolServerIP
     * @param int    $languageToolServerPort
     *
     * @throws \Exception
     */
    public function __construct(
        $languageToolLanguage,
        $enchantLanguage,
        $languageToolDirectory = null,
        $languageToolServerIP = 'localhost',
        $languageToolServerPort = 8081
    ) {
        $this->languageToolLanguage   = $languageToolLanguage;
        $this->enchantLanguage        = $enchantLanguage;
        $this->languageToolServerPort = $languageToolServerPort;
        $this->languageToolServerIP   = $languageToolServerIP;


        if ($languageToolDirectory) {
''   == false // true
''   == null  // true
'ab' == false // false
'ab' == null  // false

// It is often better to use strict comparison
'' === false // false
'' === null  // false
            $this->startLanguageToolServer($languageToolDirectory);
        }

        $this->languagetoolClientHttp = new Client();

        if (extension_loaded('enchant')) {
            $this->enchantBroker = enchant_broker_init();
            if (!enchant_broker_dict_exists($this->enchantBroker, $this->enchantLanguage)) {
                throw new \Exception("Cannot find dictionnaries for enchant");
            } else {
                $this->enchantDictionnary = enchant_broker_request_dict($this->enchantBroker, $this->enchantLanguage);
            }
        }
    }

    public function __destruct()
    {
        $this->stopLanguageToolServer();
        if ($this->enchantBroker) {
            enchant_broker_free_dict($this->enchantDictionnary);
            enchant_broker_free($this->enchantBroker);
        }
    }

    /**
     * @param string                   $title
     * @param GlSpellCheckerSentence[] $sentences
     *
     * @return string
     */
    public function convertToHtml($title, $sentences)
    {
        $html = '<!DOCTYPE HTML>';
        $html .= '<html><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8">';
        $html .= '<title>' . $title . '</title>';
        $html .= '<style>';
        $html .= '.error {  color: red  }';

        $html .= '.tooltip
                    {
                        display: inline;
                        position: relative;
                        text-decoration: none;
                        top: 0px;
                        left: 0px;
                    }';

        $html .= '.tooltip:hover:after
                    {
                        background: #333;
                        background: rgba(0,0,0,.8);
                        border-radius: 5px;
                        top: -5px;
                        color: #fff;
                        content: attr(data-tooltip);
                        left: 160px;
                        padding: 5px 15px;
                        position: absolute;
                        z-index: 98;
                        width: 150px;
                    }';
        $html .= '</style>';
        $html .= '</head><body>';

        foreach ($sentences as $sentence) {
            $html .= '<div class="sentence">';
            $text   = $sentence->getText();
            $errors = $sentence->mergeErrors();

            if (count($errors) <= 0) {
                $html .= $text;
                $html .= '</div>';
                continue;
            }

            $cons  = "";
            $start = 0;
            foreach ($errors as $error) {
                $offset = $error->getOffset();
                $length = $error->getLength();
                $cons .= mb_substr($text, $start, $offset - $start, 'UTF-8');

                $tooltip = $error->getMessage();
                $suggs   = $error->getSuggestions();
                if (count($suggs) > 0) {
                    $tooltip .= " : " . $suggs[0];
                }
                $zone = mb_substr($text, $offset, $length, 'UTF-8');
                $cons .= '<span class="error tooltip" data-tooltip="' . $tooltip . '">' . $zone . '</span>';

                $start = $offset + $length;
            }
            $cons .= mb_substr($text, $start, mb_strlen($text) - $start, 'UTF-8');

            $html .= $cons;
            $html .= '</div>';
        }
        $html .= '<br><br><br></body></html>';

        return $html;
    }

    public function checkYamlFiles(
        Finder $files,
        array    $fields,
        callable $checkfilestart,
        callable $checksentence,
        callable $checkfileend
    ) {
        $results = [];
        /**
         * @var SplFileInfo $file
         */
        foreach ($files as $file) {
            try {
                $data = Yaml::parse(
                            file_get_contents(
                                $file->getRealPath()
                            )
                );
            } catch (ParseException $e) {
                throw new \Exception("Unable to parse YAML string: {$e->getMessage()}");
            }
            $sentences = [];
            foreach ($data as $item) {
                foreach ($item as $key => $valueitem) {
                    foreach ($fields as $field) {
                        if ($key == $field) {
                            $sentences[] = $valueitem;
                        }
                    }
                }
            }
            $checkfilestart($file, count($sentences));
            $sentences = $this->checkSentences(
                              $sentences,
                                  $checksentence
            );
            $htmlcode  = $this->convertToHtml($file->getFilename(), $sentences);

            $checkerfile = sys_get_temp_dir() . "/" . uniqid("spellcheck") . ".html";
            file_put_contents($checkerfile, $htmlcode);
            $results[] = $checkerfile;

            $checkfileend();
        }

        return $results;
    }

    /**
     * @param Finder   $files
     * @param callable $checkfilestart
     * @param callable $checksentence
     * @param callable $checkfileend
     *
     * @return array
     */
    public function checkHtmlFiles(
        Finder $files,
        callable $checkfilestart,
        callable $checksentence,
        callable $checkfileend
    ) {
        $results = [];
        /**
         * @var SplFileInfo $file
         */
        foreach ($files as $file) {
            $html = file_get_contents($file->getRealPath());
            $html = new GlHtml($html);

            $title = $html->get("head title");

            if ($title && sizeof($title) > 0) {

                $title = $title[0]->getText();
            } else {
                $title = $file->getFilename();
            }

            $sentences = $html->getSentences();
            $checkfilestart($file, count($sentences));
            $sentences = $this->checkSentences(
                              $sentences,
                                  $checksentence
            );
            $htmlcode  = $this->convertToHtml($title, $sentences);

            $checkerfile = sys_get_temp_dir() . "/" . uniqid("spellcheck") . ".html";
            file_put_contents($checkerfile, $htmlcode);
            $results[] = $checkerfile;

            $checkfileend();
        }

        return $results;
    }

    /**
     * @param array    $sentences
     *
     * @param callable $closure
     *
     * @return GlSpellCheckerSentence[]
     */
    public
    function checkSentences(
        array $sentences,
        callable $closure
    ) {
        $url              = "http://{$this->languageToolServerIP}:{$this->languageToolServerPort}";

        $sentencesChecked = [];
        foreach ($sentences as $sentence) {
            $response        = $this->languagetoolClientHttp->get(
                                                            $url,
                                                                [
                                                                    'query' => [
                                                                        'language' => $this->languageToolLanguage,
                                                                        'text'     => $sentence
                                                                    ]
                                                                ]
            );
            $xml             = $response->getBody()->getContents();
            $glxml           = new GlHtml($xml);
            $errors          = $glxml->get('error');
            $sentenceChecked = new GlSpellCheckerSentence($sentence);
            if (count($errors) > 0) {
                foreach ($errors as $error) {
                    $msg    = $error->getAttribute('msg');
                    $offset = (int)$error->getAttribute('offset');
                    $length = (int)$error->getAttribute('errorlength');
                    $suggs  = [];
                    $word   = null;
                    if ($error->getAttribute('locqualityissuetype') == 'misspelling') {
                        $word = mb_substr($sentence, $offset, $length, 'UTF-8');
                        if ($this->enchantDictionnary) {
                            $wordcorrect = enchant_dict_check($this->enchantDictionnary, $word);
                            if (!$wordcorrect) {
                                $suggs = enchant_dict_suggest($this->enchantDictionnary, $word);
                            }
                        }
                    }
                    $glerror = new GlSpellCheckerError($msg, $offset, $length, $word, $suggs);
                    $sentenceChecked->addError($glerror);
                }
            }
            $sentencesChecked[] = $sentenceChecked;
            $closure($sentence);
        }

        return $sentencesChecked;
    }

    /**
     * @param string $directory
     */
    private function startLanguageToolServer($directory)
    {
        $jar                      = $directory . "languagetool-server.jar";
        $command                  = "java -cp $jar org.languagetool.server.HTTPServer --port {$this->languageToolServerPort}";
        $this->languagetoolServer = new Process($command);
        $this->languagetoolServer->start();
    }

    private function stopLanguageToolServer()
    {
        if ($this->languagetoolServer) {
            $this->languagetoolServer->stop();
            $this->languagetoolServer = null;
        }
    }
} 


1			<?php
2			/**
3			* Spell check html files
4			*
5			* PHP version 5.4
6			*
7			* @category GLICER
8			* @package GlSpellChecker
9			* @author Emmanuel ROECKER
10			* @author Rym BOUCHAGOUR
11			* @copyright 2015 GLICER
12			* @license MIT
13			* @link http://dev.glicer.com/
14			*
15			* Created : 04/05/15
16			* File : GlSpellChecker.php
17			*
18			*/
19
20			namespace GlSpellChecker;
21
22			use GlHtml\GlHtml;
23			use Symfony\Component\Process\Process;
24			use Symfony\Component\Finder\SplFileInfo;
25			use Symfony\Component\Finder\Finder;
26			use Symfony\Component\Yaml\Exception\ParseException;
27			use Symfony\Component\Yaml\Yaml;
28			use GuzzleHttp\Client;
29
30
31			/**
32			* Class GlSpellChecker
33			* @package GlSpellChecker
34			*/
35			class GlSpellChecker
36			{
37			/**
38			* @var int
39			*/
40			private $languageToolServerPort = 8081;
41
42			/**
43			* @var string
44			*/
45			private $languageToolLanguage = 'fr';
46
47			/**
48			* @var Process $languagetoolServer ;
49			*/
50			private $languagetoolServer = null;
51
52			/**
53			* @var Client
54			*/
55			private $languagetoolClientHttp;
56
57
58			/**
59			* @var string
60			*/
61			private $enchantLanguage = "fr_FR";
62			private $enchantDictionnary = null;
63			private $enchantBroker = null;
64
65			/**
66			* @param string $languageToolDirectory
67			* @param string $languageToolLanguage
68			* @param string $enchantLanguage
69			* @param string $languageToolServerIP
70			* @param int $languageToolServerPort
71			*
72			* @throws \Exception
73			*/
74			public function __construct(
75			$languageToolLanguage,
76			$enchantLanguage,
77			$languageToolDirectory = null,
78			$languageToolServerIP = 'localhost',
79			$languageToolServerPort = 8081
80			) {
81			$this->languageToolLanguage = $languageToolLanguage;
82			$this->enchantLanguage = $enchantLanguage;
83			$this->languageToolServerPort = $languageToolServerPort;
84			$this->languageToolServerIP = $languageToolServerIP;
			0 ignored issues – show Bug introduced 2016-01-09 20:02 UTC by Report Bug Copy Issue Report The property `languageToolServerIP` does not seem to exist. Did you mean `languagetoolServer`? An attempt at access to an undefined property has been detected. This may either be a typographical error or the property has been renamed but there are still references to its old name. If you really want to allow access to undefined properties, you can define magic methods to allow access. See the php core documentation on Overloading. Loading history...
85
86			if ($languageToolDirectory) {
			0 ignored issues – show Bug Best Practice introduced 2016-01-09 20:02 UTC by Report Bug Copy Issue Report The expression `$languageToolDirectory` of type `string\|null` is loosely compared to `true`; this is ambiguous if the string can be empty. You might want to explicitly use `!== null` instead. In PHP, under loose comparison (like `==`, or `!=`, or `switch` conditions), values of different types might be equal. For `string` values, the empty string `''` is a special case, in particular the following results might be unexpected: '' == false // true '' == null // true 'ab' == false // false 'ab' == null // false // It is often better to use strict comparison '' === false // false '' === null // false Loading history...
87			$this->startLanguageToolServer($languageToolDirectory);
88			}
89
90			$this->languagetoolClientHttp = new Client();
91
92			if (extension_loaded('enchant')) {
93			$this->enchantBroker = enchant_broker_init();
94			if (!enchant_broker_dict_exists($this->enchantBroker, $this->enchantLanguage)) {
95			throw new \Exception("Cannot find dictionnaries for enchant");
96			} else {
97			$this->enchantDictionnary = enchant_broker_request_dict($this->enchantBroker, $this->enchantLanguage);
98			}
99			}
100			}
101
102			public function __destruct()
103			{
104			$this->stopLanguageToolServer();
105			if ($this->enchantBroker) {
106			enchant_broker_free_dict($this->enchantDictionnary);
107			enchant_broker_free($this->enchantBroker);
108			}
109			}
110
111			/**
112			* @param string $title
113			* @param GlSpellCheckerSentence[] $sentences
114			*
115			* @return string
116			*/
117			public function convertToHtml($title, $sentences)
118			{
119			$html = '<!DOCTYPE HTML>';
120			$html .= '<html><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8">';
121			$html .= '<title>' . $title . '</title>';
122			$html .= '<style>';
123			$html .= '.error { color: red }';
124
125			$html .= '.tooltip
126			{
127			display: inline;
128			position: relative;
129			text-decoration: none;
130			top: 0px;
131			left: 0px;
132			}';
133
134			$html .= '.tooltip:hover:after
135			{
136			background: #333;
137			background: rgba(0,0,0,.8);
138			border-radius: 5px;
139			top: -5px;
140			color: #fff;
141			content: attr(data-tooltip);
142			left: 160px;
143			padding: 5px 15px;
144			position: absolute;
145			z-index: 98;
146			width: 150px;
147			}';
148			$html .= '</style>';
149			$html .= '</head><body>';
150
151			foreach ($sentences as $sentence) {
152			$html .= '<div class="sentence">';
153			$text = $sentence->getText();
154			$errors = $sentence->mergeErrors();
155
156			if (count($errors) <= 0) {
157			$html .= $text;
158			$html .= '</div>';
159			continue;
160			}
161
162			$cons = "";
163			$start = 0;
164			foreach ($errors as $error) {
165			$offset = $error->getOffset();
166			$length = $error->getLength();
167			$cons .= mb_substr($text, $start, $offset - $start, 'UTF-8');
168
169			$tooltip = $error->getMessage();
170			$suggs = $error->getSuggestions();
171			if (count($suggs) > 0) {
172			$tooltip .= " : " . $suggs[0];
173			}
174			$zone = mb_substr($text, $offset, $length, 'UTF-8');
175			$cons .= '<span class="error tooltip" data-tooltip="' . $tooltip . '">' . $zone . '</span>';
176
177			$start = $offset + $length;
178			}
179			$cons .= mb_substr($text, $start, mb_strlen($text) - $start, 'UTF-8');
180
181			$html .= $cons;
182			$html .= '</div>';
183			}
184			$html .= '<br><br><br></body></html>';
185
186			return $html;
187			}
188
189			public function checkYamlFiles(
190			Finder $files,
191			array $fields,
192			callable $checkfilestart,
193			callable $checksentence,
194			callable $checkfileend
195			) {
196			$results = [];
197			/**
198			* @var SplFileInfo $file
199			*/
200			foreach ($files as $file) {
201			try {
202			$data = Yaml::parse(
203			file_get_contents(
204			$file->getRealPath()
205			)
206			);
207			} catch (ParseException $e) {
208			throw new \Exception("Unable to parse YAML string: {$e->getMessage()}");
209			}
210			$sentences = [];
211			foreach ($data as $item) {
212			foreach ($item as $key => $valueitem) {
213			foreach ($fields as $field) {
214			if ($key == $field) {
215			$sentences[] = $valueitem;
216			}
217			}
218			}
219			}
220			$checkfilestart($file, count($sentences));
221			$sentences = $this->checkSentences(
222			$sentences,
223			$checksentence
224			);
225			$htmlcode = $this->convertToHtml($file->getFilename(), $sentences);
226
227			$checkerfile = sys_get_temp_dir() . "/" . uniqid("spellcheck") . ".html";
228			file_put_contents($checkerfile, $htmlcode);
229			$results[] = $checkerfile;
230
231			$checkfileend();
232			}
233
234			return $results;
235			}
236
237			/**
238			* @param Finder $files
239			* @param callable $checkfilestart
240			* @param callable $checksentence
241			* @param callable $checkfileend
242			*
243			* @return array
244			*/
245			public function checkHtmlFiles(
246			Finder $files,
247			callable $checkfilestart,
248			callable $checksentence,
249			callable $checkfileend
250			) {
251			$results = [];
252			/**
253			* @var SplFileInfo $file
254			*/
255			foreach ($files as $file) {
256			$html = file_get_contents($file->getRealPath());
257			$html = new GlHtml($html);
258
259			$title = $html->get("head title");
260
261			if ($title && sizeof($title) > 0) {
			0 ignored issues – show Bug Best Practice introduced 2016-01-09 19:58 UTC by Report Bug Copy Issue Report The expression `$title` of type `GlHtml\GlHtmlNode[]` is implicitly converted to a boolean; are you sure this is intended? If so, consider using `! empty($expr)` instead to make it clear that you intend to check for an array without elements. This check marks implicit conversions of arrays to boolean values in a comparison. While in PHP an empty array is considered to be equal (but not identical) to false, this is not always apparent. Consider making the comparison explicit by using `empty(..)` or `! empty(...)` instead. Loading history...
262			$title = $title[0]->getText();
263			} else {
264			$title = $file->getFilename();
265			}
266
267			$sentences = $html->getSentences();
268			$checkfilestart($file, count($sentences));
269			$sentences = $this->checkSentences(
270			$sentences,
271			$checksentence
272			);
273			$htmlcode = $this->convertToHtml($title, $sentences);
274
275			$checkerfile = sys_get_temp_dir() . "/" . uniqid("spellcheck") . ".html";
276			file_put_contents($checkerfile, $htmlcode);
277			$results[] = $checkerfile;
278
279			$checkfileend();
280			}
281
282			return $results;
283			}
284
285			/**
286			* @param array $sentences
287			*
288			* @param callable $closure
289			*
290			* @return GlSpellCheckerSentence[]
291			*/
292			public
293			function checkSentences(
294			array $sentences,
295			callable $closure
296			) {
297			$url = "http://{$this->languageToolServerIP}:{$this->languageToolServerPort}";
			0 ignored issues – show Bug introduced 2016-01-09 20:02 UTC by Report Bug Copy Issue Report The property `languageToolServerIP` does not seem to exist. Did you mean `languagetoolServer`? An attempt at access to an undefined property has been detected. This may either be a typographical error or the property has been renamed but there are still references to its old name. If you really want to allow access to undefined properties, you can define magic methods to allow access. See the php core documentation on Overloading. Loading history...
298			$sentencesChecked = [];
299			foreach ($sentences as $sentence) {
300			$response = $this->languagetoolClientHttp->get(
301			$url,
302			[
303			'query' => [
304			'language' => $this->languageToolLanguage,
305			'text' => $sentence
306			]
307			]
308			);
309			$xml = $response->getBody()->getContents();
310			$glxml = new GlHtml($xml);
311			$errors = $glxml->get('error');
312			$sentenceChecked = new GlSpellCheckerSentence($sentence);
313			if (count($errors) > 0) {
314			foreach ($errors as $error) {
315			$msg = $error->getAttribute('msg');
316			$offset = (int)$error->getAttribute('offset');
317			$length = (int)$error->getAttribute('errorlength');
318			$suggs = [];
319			$word = null;
320			if ($error->getAttribute('locqualityissuetype') == 'misspelling') {
321			$word = mb_substr($sentence, $offset, $length, 'UTF-8');
322			if ($this->enchantDictionnary) {
323			$wordcorrect = enchant_dict_check($this->enchantDictionnary, $word);
324			if (!$wordcorrect) {
325			$suggs = enchant_dict_suggest($this->enchantDictionnary, $word);
326			}
327			}
328			}
329			$glerror = new GlSpellCheckerError($msg, $offset, $length, $word, $suggs);
330			$sentenceChecked->addError($glerror);
331			}
332			}
333			$sentencesChecked[] = $sentenceChecked;
334			$closure($sentence);
335			}
336
337			return $sentencesChecked;
338			}
339
340			/**
341			* @param string $directory
342			*/
343			private function startLanguageToolServer($directory)
344			{
345			$jar = $directory . "languagetool-server.jar";
346			$command = "java -cp $jar org.languagetool.server.HTTPServer --port {$this->languageToolServerPort}";
347			$this->languagetoolServer = new Process($command);
348			$this->languagetoolServer->start();
349			}
350
351			private function stopLanguageToolServer()
352			{
353			if ($this->languagetoolServer) {
354			$this->languagetoolServer->stop();
355			$this->languagetoolServer = null;
356			}
357			}
358			}
359

emmanuelroecker / php-spellchecker

Push — master ( 6aeed9...ac4824 )

GlSpellChecker::checkSentences() C

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like