29783978122

Committed 20 Jul 2026 10:27PM UTC coverage: 98.684% (+1.9%) from 96.738%

Build # 29783978122

Build Type

Pull #2567

github

Committed by

web-flow

Commit Message

Merge f64cfee48 into 5579bd257

Pull Request Pull Request #2567: WIP Do Not Install

Coverage Stats

15594 of 15802 relevant lines covered (98.68%)

55.42 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

98.57

/src/PhpWord/Shared/Text.php

<?php

/**
 * This file is part of PHPWord - A pure PHP library for reading and writing
 * word processing documents.
 *
 * PHPWord 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/PHPWord/contributors.
 *
 * @see         https://github.com/PHPOffice/PHPWord
 *
 * @license     http://www.gnu.org/licenses/lgpl.txt LGPL version 3
 */

namespace PhpOffice\PhpWord\Shared;

use PhpOffice\PhpWord\Exception\Exception as WordException;

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

    /**
     * Build control characters array.
     */
    private static function buildControlCharacters(): void
    {
        if (!empty(self::$controlCharacters)) {
            return;
        }
        for ($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 = '')
    {
        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 int $decimals
     *
     * @return string
     */
    public static function numberFormat($number, $decimals)
    {
        return number_format($number, $decimals, '.', '');
    }

    /**
     * @param non-negative-int $dec
     *
     * @see http://stackoverflow.com/a/7153133/2235790
     *
     * @author velcrow
     *
     * @return string
     */
    public static function chr($dec)
    {
        if ($dec < 0) {
            return '';
        }
        if ($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 = '')
    {
        self::buildControlCharacters();

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

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

    /**
     * Return UTF8 encoded value.
     *
     * @param null|string $value
     *
     * @return ?string
     */
    public static function toUTF8($value = '')
    {
        if (null !== $value && !self::isUTF8($value)) {
            $value = static::mbConvertEncoding($value);
            if ($value === false) {
                throw new WordException('Unable to convert text to UTF-8');
            }
        }

        return $value;
    }

    /**
     * @param string $value
     *
     * @return false|string
     */
    protected static function mbConvertEncoding($value)
    {
        return mb_convert_encoding($value, 'UTF-8', 'ISO-8859-1');
    }

    /**
     * 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 $text UTF8 text
     *
     * @return array
     *
     * @since 0.11.0
     * @see http://www.randomchaos.com/documents/?source=php_and_unicode
     */
    public static function utf8ToUnicode($text)
    {
        $unicode = [];
        $values = [];
        $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 = [];
                    $lookingFor = 1;
                }
            }
        }

        return $unicode;
    }

    /**
     * Returns entites from unicode array.
     *
     * @param array $unicode
     *
     * @return string
     *
     * @since 0.11.0
     * @see http://www.randomchaos.com/documents/?source=php_and_unicode
     */
    private static function unicodeToEntities($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 (null !== $value) {
            if (substr($value, 0, 1) == '_') {
                $value = substr($value, 1);
            }
        }

        return $value;
    }
}

1	<?php
2
3	/**
4	* This file is part of PHPWord - A pure PHP library for reading and writing
5	* word processing documents.
6	*
7	* PHPWord is free software distributed under the terms of the GNU Lesser
8	* General Public License version 3 as published by the Free Software Foundation.
9	*
10	* For the full copyright and license information, please read the LICENSE
11	* file that was distributed with this source code. For the full list of
12	* contributors, visit https://github.com/PHPOffice/PHPWord/contributors.
13	*
14	* @see https://github.com/PHPOffice/PHPWord
15	*
16	* @license http://www.gnu.org/licenses/lgpl.txt LGPL version 3
17	*/
18
19	namespace PhpOffice\PhpWord\Shared;
20
21	use PhpOffice\PhpWord\Exception\Exception as WordException;
22
23	class Text
24	{
25	/**
26	* Control characters array.
27	*
28	* @var string[]
29	*/
30	private static $controlCharacters = [];
31
32	/**
33	* Build control characters array.
34	*/
35	private static function buildControlCharacters(): void	153✔
36	{
37	if (!empty(self::$controlCharacters)) {	153✔
38	return;	148✔
39	}
40	for ($i = 0; $i <= 19; ++$i) {	43✔
41	if ($i != 9 && $i != 10 && $i != 13) {	43✔
42	$find = '_x' . sprintf('%04s', strtoupper(dechex($i))) . '_';	43✔
43	$replace = chr($i);	43✔
44	self::$controlCharacters[$find] = $replace;	43✔
45	}
46	}
47	}
48
49	/**
50	* Convert from PHP control character to OpenXML escaped control character.
51	*
52	* Excel 2007 team:
53	* ----------------
54	* That's correct, control characters are stored directly in the shared-strings table.
55	* We do encode characters that cannot be represented in XML using the following escape sequence:
56	* _xHHHH_ where H represents a hexadecimal character in the character's value...
57	* So you could end up with something like _x0008_ in a string (either in a cell value (<v>)
58	* element or in the shared string <t> element.
59	*
60	* @param string $value Value to escape
61	*
62	* @return string
63	*/
64	public static function controlCharacterPHP2OOXML($value = '')	153✔
65	{
66	self::buildControlCharacters();	153✔
67
68	return str_replace(array_values(self::$controlCharacters), array_keys(self::$controlCharacters), $value);	153✔
69	}
70
71	/**
72	* Return a number formatted for being integrated in xml files.
73	*
74	* @param float $number
75	* @param int $decimals
76	*
77	* @return string
78	*/
79	public static function numberFormat($number, $decimals)	1✔
80	{
81	return number_format($number, $decimals, '.', '');	1✔
82	}
83
84	/**
85	* @param non-negative-int $dec
86	*
87	* @see http://stackoverflow.com/a/7153133/2235790
88	*
89	* @author velcrow
90	*
91	* @return string
92	*/
93	public static function chr($dec)	1✔
94	{
95	if ($dec < 0) {	1✔
96	return '';	×
97	}
98	if ($dec <= 0x7F) {	1✔
99	return chr($dec);	1✔
100	}
101	if ($dec <= 0x7FF) {	1✔
102	return chr(($dec >> 6) + 192) . chr(($dec & 63) + 128);	1✔
103	}
104	if ($dec <= 0xFFFF) {	1✔
105	return chr(($dec >> 12) + 224) . chr((($dec >> 6) & 63) + 128) . chr(($dec & 63) + 128);	1✔
106	}
107	if ($dec <= 0x1FFFFF) {	1✔
108	return chr(($dec >> 18) + 240) . chr((($dec >> 12) & 63) + 128) . chr((($dec >> 6) & 63) + 128) . chr(($dec & 63) + 128);	1✔
109	}
110
111	return '';	1✔
112	}
113
114	/**
115	* Convert from OpenXML escaped control character to PHP control character.
116	*
117	* @param string $value Value to unescape
118	*
119	* @return string
120	*/
121	public static function controlCharacterOOXML2PHP($value = '')	1✔
122	{
123	self::buildControlCharacters();	1✔
124
125	return str_replace(array_keys(self::$controlCharacters), array_values(self::$controlCharacters), $value);	1✔
126	}
127
128	/**
129	* Check if a string contains UTF-8 data.
130	*
131	* @param string $value
132	*
133	* @return bool
134	*/
135	public static function isUTF8($value = '')	456✔
136	{
137	return is_string($value) && ($value === '' \|\| preg_match('/^./su', $value) == 1);	456✔
138	}
139
140	/**
141	* Return UTF8 encoded value.
142	*
143	* @param null\|string $value
144	*
145	* @return ?string
146	*/
147	public static function toUTF8($value = '')	465✔
148	{
149	if (null !== $value && !self::isUTF8($value)) {	465✔
150	$value = static::mbConvertEncoding($value);	12✔
151	if ($value === false) {	12✔
152	throw new WordException('Unable to convert text to UTF-8');	1✔
153	}
154	}
155
156	return $value;	464✔
157	}
158
159	/**
160	* @param string $value
161	*
162	* @return false\|string
163	*/
164	protected static function mbConvertEncoding($value)	11✔
165	{
166	return mb_convert_encoding($value, 'UTF-8', 'ISO-8859-1');	11✔
167	}
168
169	/**
170	* Returns unicode from UTF8 text.
171	*
172	* The function is splitted to reduce cyclomatic complexity
173	*
174	* @param string $text UTF8 text
175	*
176	* @return string Unicode text
177	*
178	* @since 0.11.0
179	*/
180	public static function toUnicode($text)	1✔
181	{
182	return self::unicodeToEntities(self::utf8ToUnicode($text));	1✔
183	}
184
185	/**
186	* Returns unicode array from UTF8 text.
187	*
188	* @param string $text UTF8 text
189	*
190	* @return array
191	*
192	* @since 0.11.0
193	* @see http://www.randomchaos.com/documents/?source=php_and_unicode
194	*/
195	public static function utf8ToUnicode($text)	1✔
196	{
197	$unicode = [];	1✔
198	$values = [];	1✔
199	$lookingFor = 1;	1✔
200
201	// Gets unicode for each character
202	for ($i = 0; $i < strlen($text); ++$i) {	1✔
203	$thisValue = ord($text[$i]);	1✔
204	if ($thisValue < 128) {	1✔
205	$unicode[] = $thisValue;	1✔
206	} else {
207	if (count($values) == 0) {	1✔
208	$lookingFor = $thisValue < 224 ? 2 : 3;	1✔
209	}
210	$values[] = $thisValue;	1✔
211	if (count($values) == $lookingFor) {	1✔
212	if ($lookingFor == 3) {	1✔
213	$number = (($values[0] % 16) * 4096) + (($values[1] % 64) * 64) + ($values[2] % 64);	1✔
214	} else {
215	$number = (($values[0] % 32) * 64) + ($values[1] % 64);	1✔
216	}
217	$unicode[] = $number;	1✔
218	$values = [];	1✔
219	$lookingFor = 1;	1✔
220	}
221	}
222	}
223
224	return $unicode;	1✔
225	}
226
227	/**
228	* Returns entites from unicode array.
229	*
230	* @param array $unicode
231	*
232	* @return string
233	*
234	* @since 0.11.0
235	* @see http://www.randomchaos.com/documents/?source=php_and_unicode
236	*/
237	private static function unicodeToEntities($unicode)	1✔
238	{
239	$entities = '';	1✔
240
241	foreach ($unicode as $value) {	1✔
242	if ($value != 65279) {	1✔
243	$entities .= $value > 127 ? '\uc0{\u' . $value . '}' : chr($value);	1✔
244	}
245	}
246
247	return $entities;	1✔
248	}
249
250	/**
251	* Return name without underscore for < 0.10.0 variable name compatibility.
252	*
253	* @param string $value
254	*
255	* @return string
256	*/
257	public static function removeUnderscorePrefix($value)	436✔
258	{
259	if (null !== $value) {	436✔
260	if (substr($value, 0, 1) == '_') {	436✔
261	$value = substr($value, 1);	1✔
262	}
263	}
264
265	return $value;	436✔
266	}
267	}

PHPOffice / PHPWord / 29783978122

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous