DerDeserializer::paginate() - Code Metrics - Inspection of "Update DER deserializer:" - webeweb/core-library - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( fedb3d...ad9766 )

by WEBEWEB

created 2020-03-24 13:44 UTC

DerDeserializer::paginate() A

↳ Parent: DerDeserializer

Complexity

Conditions	3
Paths	3

Size

Total Lines

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
dl	0
loc	19
rs	9.6333
c	0
b	0
f	0
cc	3
nc	3
nop	1

<?php

/*
 * This file is part of the core-library package.
 *
 * (c) 2020 WEBEWEB
 *
 * For the full copyright and license information, please view the LICENSE
 * file that was distributed with this source code.
 */

namespace WBW\Library\Core\ThirdParty\OcrLad\Serializer;

use WBW\Library\Core\ThirdParty\OcrLad\Model\Document;
use WBW\Library\Core\ThirdParty\OcrLad\Model\Page;
use WBW\Library\Core\ThirdParty\OcrLad\Model\Word;

/**
 * DER deserializer.
 *
 * @author webeweb <https://github.com/webeweb>
 * @package WBW\Library\Core\ThirdParty\OcrLad\Serializer
 */
class DerDeserializer {

    /**
     * DER delimiter.
     *
     * @var string
     */
    const DER_DELIMITER = ";";

    /**
     * Deserialize a document.
     *
     * @param string $filename The filename.
     * @return Document|null Returns the document in case of success, null otherwise.
     */
    public static function deserializeDocument($filename) {

        $model = new Document();
        $model->setFilename($filename);

        $stream = fopen($filename, "r");

        $headers = DerDeserializer::splitHeader(fgets($stream));
        foreach ($headers as $current) {

            $page = DerDeserializer::deserializePage($current);
            if (null !== $page) {
                $model->addPage($page);
            }
        }

        while (true !== feof($stream)) {

            $word = DerDeserializer::deserializeWord(fgets($stream));
            if (null !== $word) {
                $model->addWord($word);
            }
        }

        fclose($stream);

        return DerDeserializer::paginate($model);
    }

    /**
     * Deserialize a page.
     *
     * @param string $rawData The raw data.
     * @return Page|null Returns the page in case of success, null otherwise.
     */
    protected static function deserializePage($rawData) {

        $data = explode(DerDeserializer::DER_DELIMITER, $rawData);
        if (6 !== count($data)) {
            return null;
        }

        $model = new Page();
        $model->setResolution(intval(preg_replace("/[^0-9]/", "", $data[0])));
        $model->setWidth(intval($data[1]));
        $model->setHeight(intval($data[2]));
        $model->setRotation(intval($data[3]));
        $model->setTag(intval($data[4]));

        return $model;
    }

    /**
     * Deserialize a word.
     *
     * @param string $rawData The raw data.
     * @return Word|null Returns the word in case of success, null otherwise.
     */
    protected static function deserializeWord($rawData) {

        $data = explode(DerDeserializer::DER_DELIMITER, $rawData);
        if (7 !== count($data)) {
            return null;
        }

        $model = new Word();
        $model->setContent(trim($data[0]));
        $model->setType(trim($data[1]));
        $model->setOcrConfidence(floatval(str_replace(",", ".", $data[2])));
        $model->setX1(floatval($data[3]));
        $model->setY1(floatval($data[4]));
        $model->setX2(floatval($data[5]));
        $model->setY2(floatval($data[6]));

        return $model;
    }

    /**
     * Paginate.
     *
     * @param Document $document The document.
     * @return Document Returns the document.
     */
    protected static function paginate(Document $document) {

        $last = 0;
        $page = 0;

        foreach ($document->getWords() as $current) {

            if ($current->getY1() < $last) {
                $last = 0;
$myVar = 'Value';
$higher = false;

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

            $last = $current->getY1();

            $document->getPages()[$page]->addWord($current);
        }

        return $document;
    }

    /**
     * Split an header.
     *
     * @param string $rawData The raw data.
     * @return string[] Returns the headers.
     */
    protected static function splitHeader($rawData) {

        $data = explode(DerDeserializer::DER_DELIMITER, $rawData);
        if (6 === count($data)) {
            return [$rawData];
        }

        $rows   = [];
        $buffer = [];

        foreach ($data as $current) {

            $buffer[] = $current;
            if (6 === count($buffer)) {

                $rows[] = implode(DerDeserializer::DER_DELIMITER, $buffer);
                $buffer = [];
            }
        }

        return $rows;
    }
}

1			<?php
2
3			/*
4			* This file is part of the core-library package.
5			*
6			* (c) 2020 WEBEWEB
7			*
8			* For the full copyright and license information, please view the LICENSE
9			* file that was distributed with this source code.
10			*/
11
12			namespace WBW\Library\Core\ThirdParty\OcrLad\Serializer;
13
14			use WBW\Library\Core\ThirdParty\OcrLad\Model\Document;
15			use WBW\Library\Core\ThirdParty\OcrLad\Model\Page;
16			use WBW\Library\Core\ThirdParty\OcrLad\Model\Word;
17
18			/**
19			* DER deserializer.
20			*
21			* @author webeweb <https://github.com/webeweb>
22			* @package WBW\Library\Core\ThirdParty\OcrLad\Serializer
23			*/
24			class DerDeserializer {
25
26			/**
27			* DER delimiter.
28			*
29			* @var string
30			*/
31			const DER_DELIMITER = ";";
32
33			/**
34			* Deserialize a document.
35			*
36			* @param string $filename The filename.
37			* @return Document\|null Returns the document in case of success, null otherwise.
38			*/
39			public static function deserializeDocument($filename) {
40
41			$model = new Document();
42			$model->setFilename($filename);
43
44			$stream = fopen($filename, "r");
45
46			$headers = DerDeserializer::splitHeader(fgets($stream));
47			foreach ($headers as $current) {
48
49			$page = DerDeserializer::deserializePage($current);
50			if (null !== $page) {
51			$model->addPage($page);
52			}
53			}
54
55			while (true !== feof($stream)) {
56
57			$word = DerDeserializer::deserializeWord(fgets($stream));
58			if (null !== $word) {
59			$model->addWord($word);
60			}
61			}
62
63			fclose($stream);
64
65			return DerDeserializer::paginate($model);
66			}
67
68			/**
69			* Deserialize a page.
70			*
71			* @param string $rawData The raw data.
72			* @return Page\|null Returns the page in case of success, null otherwise.
73			*/
74			protected static function deserializePage($rawData) {
75
76			$data = explode(DerDeserializer::DER_DELIMITER, $rawData);
77			if (6 !== count($data)) {
78			return null;
79			}
80
81			$model = new Page();
82			$model->setResolution(intval(preg_replace("/[^0-9]/", "", $data[0])));
83			$model->setWidth(intval($data[1]));
84			$model->setHeight(intval($data[2]));
85			$model->setRotation(intval($data[3]));
86			$model->setTag(intval($data[4]));
87
88			return $model;
89			}
90
91			/**
92			* Deserialize a word.
93			*
94			* @param string $rawData The raw data.
95			* @return Word\|null Returns the word in case of success, null otherwise.
96			*/
97			protected static function deserializeWord($rawData) {
98
99			$data = explode(DerDeserializer::DER_DELIMITER, $rawData);
100			if (7 !== count($data)) {
101			return null;
102			}
103
104			$model = new Word();
105			$model->setContent(trim($data[0]));
106			$model->setType(trim($data[1]));
107			$model->setOcrConfidence(floatval(str_replace(",", ".", $data[2])));
108			$model->setX1(floatval($data[3]));
109			$model->setY1(floatval($data[4]));
110			$model->setX2(floatval($data[5]));
111			$model->setY2(floatval($data[6]));
112
113			return $model;
114			}
115
116			/**
117			* Paginate.
118			*
119			* @param Document $document The document.
120			* @return Document Returns the document.
121			*/
122			protected static function paginate(Document $document) {
123
124			$last = 0;
125			$page = 0;
126
127			foreach ($document->getWords() as $current) {
128
129			if ($current->getY1() < $last) {
130			$last = 0;
			0 ignored issues – show Unused Code introduced 2020-03-24 13:45 UTC by Report Bug Copy Issue Report `$last` 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 `$higher` is always overwritten for every possible time line. Loading history...
131			++$page;
132			}
133
134			$last = $current->getY1();
135
136			$document->getPages()[$page]->addWord($current);
137			}
138
139			return $document;
140			}
141
142			/**
143			* Split an header.
144			*
145			* @param string $rawData The raw data.
146			* @return string[] Returns the headers.
147			*/
148			protected static function splitHeader($rawData) {
149
150			$data = explode(DerDeserializer::DER_DELIMITER, $rawData);
151			if (6 === count($data)) {
152			return [$rawData];
153			}
154
155			$rows = [];
156			$buffer = [];
157
158			foreach ($data as $current) {
159
160			$buffer[] = $current;
161			if (6 === count($buffer)) {
162
163			$rows[] = implode(DerDeserializer::DER_DELIMITER, $buffer);
164			$buffer = [];
165			}
166			}
167
168			return $rows;
169			}
170			}

webeweb / core-library

Push — master ( fedb3d...ad9766 )

DerDeserializer::paginate() A

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like