|
1
|
|
|
<?php |
|
2
|
|
|
/** |
|
3
|
|
|
* This file is part of PHP-Typography. |
|
4
|
|
|
* |
|
5
|
|
|
* Copyright 2014-2017 Peter Putzer. |
|
6
|
|
|
* Copyright 2009-2011 KINGdesk, LLC. |
|
7
|
|
|
* |
|
8
|
|
|
* This program is free software; you can redistribute it and/or modify modify |
|
9
|
|
|
* it under the terms of the GNU General Public License as published by |
|
10
|
|
|
* the Free Software Foundation; either version 2 of the License, or |
|
11
|
|
|
* (at your option) any later version. |
|
12
|
|
|
* |
|
13
|
|
|
* This program is distributed in the hope that it will be useful, |
|
14
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
|
15
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
|
16
|
|
|
* GNU General Public License for more details. |
|
17
|
|
|
* |
|
18
|
|
|
* You should have received a copy of the GNU General Public License along |
|
19
|
|
|
* with this program; if not, write to the Free Software Foundation, Inc., |
|
20
|
|
|
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. |
|
21
|
|
|
* |
|
22
|
|
|
* *** |
|
23
|
|
|
* |
|
24
|
|
|
* @package mundschenk-at/php-typography |
|
25
|
|
|
* @license http://www.gnu.org/licenses/gpl-2.0.html |
|
26
|
|
|
*/ |
|
27
|
|
|
|
|
28
|
|
|
namespace PHP_Typography\Fixes\Node_Fixes; |
|
29
|
|
|
|
|
30
|
|
|
use PHP_Typography\DOM; |
|
31
|
|
|
use PHP_Typography\RE; |
|
32
|
|
|
use PHP_Typography\Settings; |
|
33
|
|
|
use PHP_Typography\Strings; |
|
34
|
|
|
use PHP_Typography\U; |
|
35
|
|
|
|
|
36
|
|
|
/** |
|
37
|
|
|
* Prevents widows (if enabled). |
|
38
|
|
|
* |
|
39
|
|
|
* @author Peter Putzer <[email protected]> |
|
40
|
|
|
* |
|
41
|
|
|
* @since 5.0.0 |
|
42
|
|
|
*/ |
|
43
|
|
|
class Dewidow_Fix extends Abstract_Node_Fix { |
|
44
|
|
|
const SPACE_BETWEEN = '[\s]+'; // \s includes all special spaces (but not ZWSP) with the u flag. |
|
45
|
|
|
const WIDOW = '[\w\p{M}\-' . U::ZERO_WIDTH_SPACE . U::SOFT_HYPHEN . ']+?'; // \w includes all alphanumeric Unicode characters but not composed characters. |
|
46
|
|
|
|
|
47
|
|
|
// Mandatory UTF-8 modifer. |
|
48
|
|
|
const REGEX_START = '/ |
|
49
|
|
|
(?: |
|
50
|
|
|
\A |
|
51
|
|
|
| |
|
52
|
|
|
(?: |
|
53
|
|
|
(?<space_before> # subpattern 1: space before (note: ZWSP is not a space) |
|
54
|
|
|
[\s' . U::ZERO_WIDTH_SPACE . U::SOFT_HYPHEN . ']+ |
|
55
|
|
|
) |
|
56
|
|
|
(?<neighbor> # subpattern 2: neighbors widow (short as possible) |
|
57
|
|
|
[^\s' . U::ZERO_WIDTH_SPACE . U::SOFT_HYPHEN . ']+? |
|
58
|
|
|
) |
|
59
|
|
|
) |
|
60
|
|
|
) |
|
61
|
|
|
(?<space_between> # subpattern 3: space between |
|
62
|
|
|
' . self::SPACE_BETWEEN . ' |
|
63
|
|
|
) |
|
64
|
|
|
(?<widow> # subpattern 4: widow |
|
65
|
|
|
' . self::WIDOW . ' |
|
66
|
|
|
(?: |
|
67
|
|
|
' . self::SPACE_BETWEEN . self::WIDOW . ' |
|
68
|
|
|
){0,'; // The maximum number of repetitions is missing. |
|
69
|
|
|
|
|
70
|
|
|
const REGEX_END = |
|
71
|
|
|
'}) |
|
72
|
|
|
(?<trailing> # subpattern 5: any trailing punctuation or spaces |
|
73
|
|
|
[^\w\p{M}]* |
|
74
|
|
|
) |
|
75
|
|
|
\Z |
|
76
|
|
|
/Sxu'; |
|
77
|
|
|
|
|
78
|
|
|
const MASKED_NARROW_SPACE = '__NO_BREAK_NARROW_SPACE__'; |
|
79
|
|
|
|
|
80
|
|
|
/** |
|
81
|
|
|
* Apply the fix to a given textnode. |
|
82
|
|
|
* |
|
83
|
|
|
* @param \DOMText $textnode Required. |
|
84
|
|
|
* @param Settings $settings Required. |
|
85
|
|
|
* @param bool $is_title Optional. Default false. |
|
86
|
|
|
*/ |
|
87
|
|
|
public function apply( \DOMText $textnode, Settings $settings, $is_title = false ) { |
|
88
|
|
|
// Intervening inline tags may interfere with widow identification, but that is a sacrifice of using the parser. |
|
89
|
|
|
// Intervening tags will only interfere if they separate the widow from previous or preceding whitespace. |
|
90
|
|
|
if ( empty( $settings['dewidow'] ) || empty( $settings['dewidowMaxPull'] ) || empty( $settings['dewidowMaxLength'] ) ) { |
|
91
|
|
|
return; |
|
92
|
|
|
} |
|
93
|
|
|
|
|
94
|
|
|
if ( '' === DOM::get_next_chr( $textnode ) ) { |
|
95
|
|
|
// We have the last type "text" child of a block level element. |
|
96
|
|
|
$textnode->data = $this->dewidow( $textnode->data, Strings::functions( $textnode->data ), $settings['dewidowMaxPull'], $settings['dewidowMaxLength'], $settings['dewidowWordNumber'], $settings->no_break_narrow_space() ); |
|
97
|
|
|
} |
|
98
|
|
|
} |
|
99
|
|
|
|
|
100
|
|
|
/** |
|
101
|
|
|
* Dewidow a given text fragment. |
|
102
|
|
|
* |
|
103
|
|
|
* @param string $text The text fragment to dewidow. |
|
104
|
|
|
* @param array $func An array of string functions. |
|
105
|
|
|
* @param int $max_pull Maximum number of characters pulled from previous line. |
|
106
|
|
|
* @param int $max_length Maximum widow length. |
|
107
|
|
|
* @param int $word_number Maximum number of words allowed in widow. |
|
108
|
|
|
* @param string $narrow_space The narrow no-break space character. |
|
109
|
|
|
* |
|
110
|
|
|
* @return string |
|
111
|
|
|
*/ |
|
112
|
|
|
protected function dewidow( $text, array $func, $max_pull, $max_length, $word_number, $narrow_space ) { |
|
113
|
|
|
if ( $word_number < 1 ) { |
|
114
|
|
|
return $text; // We are done. |
|
115
|
|
|
} |
|
116
|
|
|
|
|
117
|
|
|
// Do what we have to do. |
|
118
|
|
|
return preg_replace_callback( self::REGEX_START . ( $word_number - 1 ) . self::REGEX_END, function( array $widow ) use ( $func, $max_pull, $max_length, $word_number, $narrow_space ) { |
|
119
|
|
|
|
|
120
|
|
|
// If we are here, we know that widows are being protected in some fashion |
|
121
|
|
|
// with that, we will assert that widows should never be hyphenated or wrapped |
|
122
|
|
|
// as such, we will strip soft hyphens and zero-width-spaces. |
|
123
|
|
|
$widow['widow'] = self::strip_breaking_characters( $widow['widow'] ); |
|
124
|
|
|
$widow['trailing'] = self::strip_breaking_characters( self::make_space_nonbreaking( $widow['trailing'], $narrow_space, $func['u'] ) ); |
|
125
|
|
|
|
|
126
|
|
|
if ( |
|
127
|
|
|
// Eject if widows neighbor is proceeded by a no break space (the pulled text would be too long). |
|
128
|
|
|
'' === $widow['space_before'] || false !== \strpos( $widow['space_before'], U::NO_BREAK_SPACE ) || |
|
129
|
|
|
|
|
130
|
|
|
// Eject if widows neighbor length exceeds the max allowed or widow length exceeds max allowed. |
|
131
|
|
|
$func['strlen']( $widow['neighbor'] ) > $max_pull || $func['strlen']( $widow['widow'] ) > $max_length || |
|
132
|
|
|
|
|
133
|
|
|
// Never replace thin and hair spaces with . |
|
134
|
|
|
self::is_narrow_space( $widow['space_between'] ) |
|
135
|
|
|
) { |
|
136
|
|
|
return $widow['space_before'] . $widow['neighbor'] . $this->dewidow( $widow['space_between'] . $widow['widow'] . $widow['trailing'], $func, $max_pull, $max_length, $word_number - 1, $narrow_space ); |
|
137
|
|
|
} |
|
138
|
|
|
|
|
139
|
|
|
// Let's protect some widows! |
|
140
|
|
|
return $widow['space_before'] . $widow['neighbor'] . U::NO_BREAK_SPACE . self::make_space_nonbreaking( $widow['widow'], $narrow_space, $func['u'] ) . $widow['trailing']; |
|
141
|
|
|
}, $text ); |
|
142
|
|
|
} |
|
143
|
|
|
|
|
144
|
|
|
/** |
|
145
|
|
|
* Strip zero-width space and soft hyphens from the given string. |
|
146
|
|
|
* |
|
147
|
|
|
* @param string $string Required. |
|
148
|
|
|
* |
|
149
|
|
|
* @return string |
|
150
|
|
|
*/ |
|
151
|
|
|
protected static function strip_breaking_characters( $string ) { |
|
152
|
|
|
return str_replace( [ U::ZERO_WIDTH_SPACE, U::SOFT_HYPHEN ], '', $string ); |
|
153
|
|
|
} |
|
154
|
|
|
|
|
155
|
|
|
/** |
|
156
|
|
|
* Is the given string one of the narrow space characters? |
|
157
|
|
|
* |
|
158
|
|
|
* @since 6.0.0 |
|
159
|
|
|
* |
|
160
|
|
|
* @param string $string The whitespace to test. |
|
161
|
|
|
* |
|
162
|
|
|
* @return bool |
|
163
|
|
|
*/ |
|
164
|
|
|
protected static function is_narrow_space( $string ) { |
|
165
|
|
|
return U::THIN_SPACE === $string || U::HAIR_SPACE === $string || U::NO_BREAK_NARROW_SPACE === $string; |
|
166
|
|
|
} |
|
167
|
|
|
|
|
168
|
|
|
/** |
|
169
|
|
|
* Strip zero-width space and soft hyphens from the given string. |
|
170
|
|
|
* |
|
171
|
|
|
* @param string $string Required. |
|
172
|
|
|
* @param string $narrow_space The narrow no-break space character. |
|
173
|
|
|
* @param string $u Either 'u' or the empty string. |
|
174
|
|
|
* |
|
175
|
|
|
* @return string |
|
176
|
|
|
*/ |
|
177
|
|
|
protected static function make_space_nonbreaking( $string, $narrow_space, $u ) { |
|
178
|
|
|
return preg_replace( |
|
179
|
|
|
[ |
|
180
|
|
|
'/\s*' . U::THIN_SPACE . '\s*/u', |
|
181
|
|
|
'/\s*' . U::NO_BREAK_NARROW_SPACE . '\s*/u', |
|
182
|
|
|
"/\\s+/$u", |
|
183
|
|
|
'/' . self::MASKED_NARROW_SPACE . "/$u", |
|
184
|
|
|
], [ |
|
185
|
|
|
self::MASKED_NARROW_SPACE, |
|
186
|
|
|
self::MASKED_NARROW_SPACE, |
|
187
|
|
|
U::NO_BREAK_SPACE, |
|
188
|
|
|
$narrow_space, |
|
189
|
|
|
], $string |
|
190
|
|
|
); |
|
191
|
|
|
} |
|
192
|
|
|
} |
|
193
|
|
|
|
mundschenk-at /
php-typography
