Text::toUTF8() - Code Metrics - Inspection of "Merge pull request #21 from Trainmaster/fix-write-..." - PHPOffice/Common - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — develop ( 3a4c83...2cf724 )

by Franck

created 2018-05-23 08:10 UTC

Text::toUTF8() A

↳ Parent: Text

Complexity

Conditions	3
Paths	2

Size

Total Lines	8
Code Lines	4

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	0
CRAP Score	12

Importance

Changes

Metric	Value
dl	0
loc	8
ccs	0
cts	5
cp	0
rs	9.4285
c	0
b	0
f	0
cc	3
eloc	4
nc	2
nop	1
crap	12

<?php
/**
 * This file is part of PHPOffice Common
 *
 * PHPOffice Common is free software distributed under the terms of the GNU Lesser
 * General Public License version 3 as published by the Free Software Foundation.
 *
 * For the full copyright and license information, please read the LICENSE
 * file that was distributed with this source code. For the full list of
 * contributors, visit https://github.com/PHPOffice/Common/contributors.
 *
 * @link        https://github.com/PHPOffice/Common
 * @copyright   2009-2016 PHPOffice Common contributors
 * @license     http://www.gnu.org/licenses/lgpl.txt LGPL version 3
 */

namespace PhpOffice\Common;

/**
 * Text
 */
class Text
{
    /**
     * Control characters array
     *
     * @var string[]
     */
    private static $controlCharacters = array();

    /**
     * Build control characters array
     */
    private static function buildControlCharacters()
    {
        for ($i = 0; $i <= 19; ++$i) {
            if ($i != 9 && $i != 10 && $i != 13) {
                $find                            = '_x' . sprintf('%04s', strtoupper(dechex($i))) . '_';
                $replace                         = chr($i);
                self::$controlCharacters[$find] = $replace;
            }
        }
    }

    /**
     * Convert from PHP control character to OpenXML escaped control character
     *
     * Excel 2007 team:
     * ----------------
     * That's correct, control characters are stored directly in the shared-strings table.
     * We do encode characters that cannot be represented in XML using the following escape sequence:
     * _xHHHH_ where H represents a hexadecimal character in the character's value...
     * So you could end up with something like _x0008_ in a string (either in a cell value (<v>)
     * element or in the shared string <t> element.
     *
     * @param  string $value Value to escape
     * @return string
     */
    public static function controlCharacterPHP2OOXML($value = '')
    {
        if (empty(self::$controlCharacters)) {
            self::buildControlCharacters();
        }

        return str_replace(array_values(self::$controlCharacters), array_keys(self::$controlCharacters), $value);
    }

    /**
     * Return a number formatted for being integrated in xml files
     * @param float $number
     * @param integer $decimals
     */
    public static function numberFormat($number, $decimals)
    {
        return number_format($number, $decimals, '.', '');
    }

    /**
     * @param int $dec
     * @link http://stackoverflow.com/a/7153133/2235790
     * @author velcrow
     */
    public static function chr($dec)
    {
        if ($dec<=0x7F) {
            return chr($dec);
        }
        if ($dec<=0x7FF) {
            return chr(($dec>>6)+192).chr(($dec&63)+128);
        }
        if ($dec<=0xFFFF) {
            return chr(($dec>>12)+224).chr((($dec>>6)&63)+128).chr(($dec&63)+128);
        }
        if ($dec<=0x1FFFFF) {
            return chr(($dec>>18)+240).chr((($dec>>12)&63)+128).chr((($dec>>6)&63)+128).chr(($dec&63)+128);
        }
        return '';
    }

    /**
     * Convert from OpenXML escaped control character to PHP control character
     *
     * @param string $value Value to unescape
     * @return string
     */
    public static function controlCharacterOOXML2PHP($value = '')
    {
        if (empty(self::$controlCharacters)) {
            self::buildControlCharacters();
        }

        return str_replace(array_keys(self::$controlCharacters), array_values(self::$controlCharacters), $value);
    }

    /**
     * Check if a string contains UTF-8 data
     *
     * @param string $value
     * @return boolean
     */
    public static function isUTF8($value = '')
    {
        return is_string($value) && ($value === '' || preg_match('/^./su', $value) == 1);
    }

    /**
     * Return UTF8 encoded value
     *
     * @param string $value
     * @return string
     */
    public static function toUTF8($value = '')
    {
        if (!is_null($value) && !self::isUTF8($value)) {
            $value = utf8_encode($value);
        }

        return $value;
    }

    /**
     * Returns unicode from UTF8 text
     *
     * The function is splitted to reduce cyclomatic complexity
     *
     * @param string $text UTF8 text
     * @return string Unicode text
     * @since 0.11.0
     */
    public static function toUnicode($text)
    {
        return self::unicodeToEntities(self::utf8ToUnicode($text));
    }

    /**
     * Returns unicode array from UTF8 text
     *
     * @param string $text UTF8 text
     * @return array
     * @since 0.11.0
     * @link http://www.randomchaos.com/documents/?source=php_and_unicode
     */
    public static function utf8ToUnicode($text)
    {
        $unicode = array();
        $values = array();
        $lookingFor = 1;

        // Gets unicode for each character
        for ($i = 0; $i < strlen($text); $i++) {
            $thisValue = ord($text[$i]);
            if ($thisValue < 128) {
                $unicode[] = $thisValue;
            } else {
                if (count($values) == 0) {
                    $lookingFor = $thisValue < 224 ? 2 : 3;
                }
                $values[] = $thisValue;
                if (count($values) == $lookingFor) {
                    if ($lookingFor == 3) {
                        $number = (($values[0] % 16) * 4096) + (($values[1] % 64) * 64) + ($values[2] % 64);
                    } else {
                        $number = (($values[0] % 32) * 64) + ($values[1] % 64);
                    }
                    $unicode[] = $number;
                    $values = array();
                    $lookingFor = 1;
                }
            }
        }

        return $unicode;
    }

    /**
     * Returns entites from unicode array
     *
     * @param array $unicode
     * @return string
     * @since 0.11.0
     * @link http://www.randomchaos.com/documents/?source=php_and_unicode
     */
    private static function unicodeToEntities($unicode)
    {
        $entities = '';

        foreach ($unicode as $value) {
            if ($value != 65279) {
                $entities .= $value > 127 ? '\uc0{\u' . $value . '}' : chr($value);
            }
        }

        return $entities;
    }

    /**
     * Return name without underscore for < 0.10.0 variable name compatibility
     *
     * @param string $value
     * @return string
     */
    public static function removeUnderscorePrefix($value)
    {
        if (!is_null($value)) {
            if (substr($value, 0, 1) == '_') {
                $value = substr($value, 1);
            }
        }

        return $value;
    }
}


1		<?php
2		/**
3		* This file is part of PHPOffice Common
4		*
5		* PHPOffice Common is free software distributed under the terms of the GNU Lesser
6		* General Public License version 3 as published by the Free Software Foundation.
7		*
8		* For the full copyright and license information, please read the LICENSE
9		* file that was distributed with this source code. For the full list of
10		* contributors, visit https://github.com/PHPOffice/Common/contributors.
11		*
12		* @link https://github.com/PHPOffice/Common
13		* @copyright 2009-2016 PHPOffice Common contributors
14		* @license http://www.gnu.org/licenses/lgpl.txt LGPL version 3
15		*/
16
17		namespace PhpOffice\Common;
18
19		/**
20		* Text
21		*/
22		class Text
23		{
24		/**
25		* Control characters array
26		*
27		* @var string[]
28		*/
29		private static $controlCharacters = array();
30
31		/**
32		* Build control characters array
33		*/
34	1	private static function buildControlCharacters()
35		{
36	1	for ($i = 0; $i <= 19; ++$i) {
37	1	if ($i != 9 && $i != 10 && $i != 13) {
38	1	$find = '_x' . sprintf('%04s', strtoupper(dechex($i))) . '_';
39	1	$replace = chr($i);
40	1	self::$controlCharacters[$find] = $replace;
41	1	}
42	1	}
43	1	}
44
45		/**
46		* Convert from PHP control character to OpenXML escaped control character
47		*
48		* Excel 2007 team:
49		* ----------------
50		* That's correct, control characters are stored directly in the shared-strings table.
51		* We do encode characters that cannot be represented in XML using the following escape sequence:
52		* _xHHHH_ where H represents a hexadecimal character in the character's value...
53		* So you could end up with something like _x0008_ in a string (either in a cell value (<v>)
54		* element or in the shared string <t> element.
55		*
56		* @param string $value Value to escape
57		* @return string
58		*/
59	1	public static function controlCharacterPHP2OOXML($value = '')
60		{
61	1	if (empty(self::$controlCharacters)) {
62	1	self::buildControlCharacters();
63	1	}
64
65	1	return str_replace(array_values(self::$controlCharacters), array_keys(self::$controlCharacters), $value);
66		}
67
68		/**
69		* Return a number formatted for being integrated in xml files
70		* @param float $number
71		* @param integer $decimals
72		*/
73	1	public static function numberFormat($number, $decimals)
74		{
75	1	return number_format($number, $decimals, '.', '');
76		}
77
78		/**
79		* @param int $dec
80		* @link http://stackoverflow.com/a/7153133/2235790
81		* @author velcrow
82		*/
83	1	public static function chr($dec)
84		{
85	1	if ($dec<=0x7F) {
86	1	return chr($dec);
87		}
88	1	if ($dec<=0x7FF) {
89	1	return chr(($dec>>6)+192).chr(($dec&63)+128);
90		}
91	1	if ($dec<=0xFFFF) {
92	1	return chr(($dec>>12)+224).chr((($dec>>6)&63)+128).chr(($dec&63)+128);
93		}
94	1	if ($dec<=0x1FFFFF) {
95	1	return chr(($dec>>18)+240).chr((($dec>>12)&63)+128).chr((($dec>>6)&63)+128).chr(($dec&63)+128);
96		}
97	1	return '';
98		}
99
100		/**
101		* Convert from OpenXML escaped control character to PHP control character
102		*
103		* @param string $value Value to unescape
104		* @return string
105		*/
106	1	public static function controlCharacterOOXML2PHP($value = '')
107		{
108	1	if (empty(self::$controlCharacters)) {
109		self::buildControlCharacters();
110		}
111
112	1	return str_replace(array_keys(self::$controlCharacters), array_values(self::$controlCharacters), $value);
113		}
114
115		/**
116		* Check if a string contains UTF-8 data
117		*
118		* @param string $value
119		* @return boolean
120		*/
121	1	public static function isUTF8($value = '')
122		{
123	1	return is_string($value) && ($value === '' \|\| preg_match('/^./su', $value) == 1);
124		}
125
126		/**
127		* Return UTF8 encoded value
128		*
129		* @param string $value
130		* @return string
131		*/
132		public static function toUTF8($value = '')
133		{
134		if (!is_null($value) && !self::isUTF8($value)) {
135		$value = utf8_encode($value);
136		}
137
138		return $value;
139		}
140
141		/**
142		* Returns unicode from UTF8 text
143		*
144		* The function is splitted to reduce cyclomatic complexity
145		*
146		* @param string $text UTF8 text
147		* @return string Unicode text
148		* @since 0.11.0
149		*/
150	1	public static function toUnicode($text)
151		{
152	1	return self::unicodeToEntities(self::utf8ToUnicode($text));
153		}
154
155		/**
156		* Returns unicode array from UTF8 text
157		*
158		* @param string $text UTF8 text
159		* @return array
160		* @since 0.11.0
161		* @link http://www.randomchaos.com/documents/?source=php_and_unicode
162		*/
163	1	public static function utf8ToUnicode($text)
164		{
165	1	$unicode = array();
166	1	$values = array();
167	1	$lookingFor = 1;
168
169		// Gets unicode for each character
170	1	for ($i = 0; $i < strlen($text); $i++) {
171	1	$thisValue = ord($text[$i]);
172	1	if ($thisValue < 128) {
173	1	$unicode[] = $thisValue;
174	1	} else {
175	1	if (count($values) == 0) {
176	1	$lookingFor = $thisValue < 224 ? 2 : 3;
177	1	}
178	1	$values[] = $thisValue;
179	1	if (count($values) == $lookingFor) {
180	1	if ($lookingFor == 3) {
181	1	$number = (($values[0] % 16) * 4096) + (($values[1] % 64) * 64) + ($values[2] % 64);
182	1	} else {
183	1	$number = (($values[0] % 32) * 64) + ($values[1] % 64);
184		}
185	1	$unicode[] = $number;
186	1	$values = array();
187	1	$lookingFor = 1;
188	1	}
189		}
190	1	}
191
192	1	return $unicode;
193		}
194
195		/**
196		* Returns entites from unicode array
197		*
198		* @param array $unicode
199		* @return string
200		* @since 0.11.0
201		* @link http://www.randomchaos.com/documents/?source=php_and_unicode
202		*/
203	1	private static function unicodeToEntities($unicode)
204		{
205	1	$entities = '';
206
207	1	foreach ($unicode as $value) {
208	1	if ($value != 65279) {
209	1	$entities .= $value > 127 ? '\uc0{\u' . $value . '}' : chr($value);
210	1	}
211	1	}
212
213	1	return $entities;
214		}
215
216		/**
217		* Return name without underscore for < 0.10.0 variable name compatibility
218		*
219		* @param string $value
220		* @return string
221		*/
222	1	public static function removeUnderscorePrefix($value)
223		{
224	1	if (!is_null($value)) {
225	1	if (substr($value, 0, 1) == '_') {
226	1	$value = substr($value, 1);
227	1	}
228	1	}
229
230	1	return $value;
231		}
232		}
233

PHPOffice / Common

Push — develop ( 3a4c83...2cf724 )

Text::toUTF8() A

Complexity

Size

Duplication

Code Coverage

Importance

Duplication Side-by-Side

Filter issues like