class-tiny-source-base.php — TinyPNG – JPEG, PNG & WebP image compression trunk

tiny-compress-images / src / class-tiny-source-base.php

tiny-compress-images / src Last commit date

class-tiny-source-base.php

312 lines

1	<?php
2
3	/*
4	* Tiny Compress Images - WordPress plugin.
5	* Copyright (C) 2015-2018 Tinify B.V.
6	*
7	* This program is free software; you can redistribute it and/or modify it
8	* under the terms of the GNU General Public License as published by the Free
9	* Software Foundation; either version 2 of the License, or (at your option)
10	* any later version.
11	*
12	* This program is distributed in the hope that it will be useful, but WITHOUT
13	* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
14	* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
15	* more details.
16	*
17	* You should have received a copy of the GNU General Public License along
18	* with this program; if not, write to the Free Software Foundation, Inc., 51
19	* Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
20	*/
21
22	abstract class Tiny_Source_Base {
23
24	public $raw_html;
25	protected $base_dir;
26	protected $allowed_domains;
27	protected $valid_mimetypes;
28
29	public function __construct( $html, $base_dir, $domains ) {
30	$this->raw_html = $html;
31	$this->base_dir = $base_dir;
32	$this->allowed_domains = $domains;
33	$this->valid_mimetypes = array( 'image/avif', 'image/webp' );
34	}
35
36	protected static function get_attribute_value( $element, $name ) {
37	// Match the exact attribute name (not part of data-media, mediaType, etc.)
38	// and capture a single- or double-quoted value.
39	$delim = '~';
40	$attr = preg_quote( $name, $delim );
41	$regex = $delim . '(?<![\w:-])' . $attr . '\s=\s(["\'])(.*?)\1' . $delim . 'is';
42
43	if ( preg_match( $regex, $element, $m ) ) {
44	return $m[2];
45	}
46	return null;
47	}
48
49	/**
50	* Extract elements by tag name from an HTML string (regex-based).
51	*
52	* @param string $html The HTML string to search in.
53	* @param string $tagname The tag name (e.g., 'div', 'source', 'img').
54	* @return array Array of matched elements as strings.
55	*/
56	protected function get_element_by_tag( $html, $tagname ) {
57	$results = array();
58
59	// Self-closing / void tag (e.g. <source />, <img />, <br />)
60	if ( preg_match_all(
61	'~<' . preg_quote( $tagname, '~' ) . '\b(?:[^>"\']+\|"[^"]"\|\'[^\']\')*/?>~i',
62	$html,
63	$matches
64	) ) {
65	$results = array_merge( $results, $matches[0] );
66	}
67
68	// Normal paired tags (e.g. <div>…</div>)
69	$regex_tag = preg_quote( $tagname, '~' );
70	if ( preg_match_all(
71	'~<' . $regex_tag .
72	'\b(?:[^>"\']+\|"[^"]"\|\'[^\']\')>.?</' .
73	$regex_tag .
74	'>~is',
75	$html,
76	$matches
77	) ) {
78	$results = array_merge( $results, $matches[0] );
79	}
80
81	return $results;
82	}
83
84	protected function get_local_path( $url ) {
85	if ( strpos( $url, 'http' ) === 0 ) {
86	$matched_domain = null;
87
88	foreach ( $this->allowed_domains as $domain ) {
89	if ( strpos( $url, $domain ) === 0 ) {
90	$matched_domain = $domain;
91	break;
92	}
93	}
94
95	if ( null === $matched_domain ) {
96	return '';
97	}
98
99	$url = substr( $url, strlen( $matched_domain ) );
100	}
101	$url = $this->base_dir . $url;
102
103	return $url;
104	}
105
106	protected function get_formatted_source( $image_source_data, $mimetype ) {
107	$format_url = Tiny_Helpers::replace_file_extension( $mimetype, $image_source_data['path'] );
108	$local_path = $this->get_local_path( $format_url );
109	if ( empty( $local_path ) ) {
110	return null;
111	}
112
113	$exists_local = file_exists( $local_path );
114	if ( $exists_local ) {
115	return array(
116	'src' => $format_url,
117	'size' => $image_source_data['size'],
118	'type' => $mimetype,
119	);
120	}
121	return null;
122	}
123
124	/**
125	* Retrieves the sources from the <img> or <source> element
126	*
127	* @return array{path: string, size: string}[] The image sources
128	*/
129	protected function get_image_srcsets( $html ) {
130	$result = array();
131	$srcset = $this::get_attribute_value( $html, 'srcset' );
132
133	if ( $srcset ) {
134	// Split the srcset to get individual entries
135	$srcset_entries = explode( ',', $srcset );
136
137	foreach ( $srcset_entries as $entry ) {
138	// Trim whitespace
139	$entry = trim( $entry );
140
141	// Split by whitespace to separate path and size/density descriptor
142	$parts = preg_split( '/\s+/', $entry, 2 );
143
144	if ( count( $parts ) === 2 ) {
145	// We have both path and size
146	$result[] = array(
147	'path' => $parts[0],
148	'size' => $parts[1],
149	);
150	} elseif ( count( $parts ) === 1 ) {
151	// We only have a path, will be interpreted as pixel
152	// density 1x (unusual in srcset)
153	$result[] = array(
154	'path' => $parts[0],
155	'size' => '',
156	);
157	}
158	}
159	}
160	return $result;
161	}
162
163	/**
164	* Retrieves the sources from the <img> or <source> element
165	*
166	* @return array{path: string, size: string}[] The image sources
167	*/
168	private function get_image_src( $html ) {
169	$source = $this::get_attribute_value( $html, 'src' );
170	if ( ! empty( $source ) ) {
171	// No srcset, but we have a src attribute
172	return array(
173	'path' => $source,
174	'size' => '',
175	);
176	}
177	return array();
178	}
179
180
181	/**
182	* Creates one or more <source> elements if alternative formats
183	* are available.
184	*
185	* @param string $original_source_html, either <source> or <img>
186	* @return array{string} array of <source> html
187	*/
188	protected function create_alternative_sources( $original_source_html ) {
189	$srcsets = $this->get_image_srcsets( $original_source_html );
190	if ( empty( $srcsets ) ) {
191	// no srcset, try src attribute
192	$src = $this->get_image_src( $original_source_html );
193	if ( ! empty( $src ) ) {
194	$srcsets[] = $src;
195	}
196	}
197
198	if ( empty( $srcsets ) ) {
199	return array();
200	}
201
202	$is_source_tag = (bool) preg_match( '#<source\b#i', $original_source_html );
203
204	$sources = array();
205	$width_descriptor = $this->get_largest_width_descriptor( $srcsets );
206
207	foreach ( $this->valid_mimetypes as $mimetype ) {
208	$srcset_parts = array();
209
210	foreach ( $srcsets as $srcset ) {
211	$alt_source = $this->get_formatted_source( $srcset, $mimetype );
212	if ( $alt_source ) {
213	$srcset_parts[] = trim( $alt_source['src'] . ' ' . $alt_source['size'] );
214	}
215	}
216
217	if (
218	$width_descriptor &&
219	! self::srcset_contains_width_descriptor(
220	$srcset_parts,
221	$width_descriptor
222	)
223	) {
224	continue;
225	}
226
227	if ( empty( $srcset_parts ) ) {
228	continue;
229	}
230
231	$source_attr_parts = array();
232
233	$srcset_attr = implode( ', ', $srcset_parts );
234	$source_attr_parts['srcset'] = $srcset_attr;
235
236	if ( $is_source_tag ) {
237	foreach ( array( 'sizes', 'media', 'width', 'height' ) as $attr ) {
238	$attr_value = $this->get_attribute_value( $original_source_html, $attr );
239	if ( $attr_value ) {
240	$source_attr_parts[ $attr ] = $attr_value;
241	}
242	}
243	} else {
244	$sizes_value = $this->get_attribute_value( $original_source_html, 'sizes' );
245	if ( $sizes_value ) {
246	$source_attr_parts['sizes'] = $sizes_value;
247	}
248	}
249
250	$source_attr_parts['type'] = $mimetype;
251	$source_parts = array( '<source' );
252	foreach ( $source_attr_parts as $source_attr_name => $source_attr_val ) {
253	$source_parts[] = $source_attr_name . '="' . esc_attr( $source_attr_val ) . '"';
254	}
255	$source_parts[] = '/>';
256	$sources[] = implode( ' ', $source_parts );
257	} // End foreach().
258
259	return $sources;
260	}
261
262	/**
263	* Returns the largest numeric width descriptor
264	* (e.g. 2000 from "2000w") found in the srcset data.
265	*
266	* @param array<array{path: string, size: string}> $srcsets
267	* @return int
268	*/
269	public static function get_largest_width_descriptor( $srcsets ) {
270	$largest = 0;
271
272	foreach ( $srcsets as $srcset ) {
273	if ( empty( $srcset['size'] ) ) {
274	continue;
275	}
276
277	if ( preg_match( '/(\d+)w/', $srcset['size'], $matches ) ) {
278	$width = (int) $matches[1];
279	if ( $width > $largest ) {
280	$largest = $width;
281	}
282	}
283	}
284
285	return $largest;
286	}
287
288	/**
289	* Determines whether a srcset list contains the provided width descriptor.
290	*
291	* @param string[] $srcset_parts
292	* @param int $width_descriptor
293	* @return bool true if width is in srcset
294	*/
295	public static function srcset_contains_width_descriptor( $srcset_parts, $width_descriptor ) {
296	if ( empty( $srcset_parts ) \|\| $width_descriptor <= 0 ) {
297	return false;
298	}
299
300	$suffix = ' ' . $width_descriptor . 'w';
301	$suffix_length = strlen( $suffix );
302
303	foreach ( $srcset_parts as $srcset_part ) {
304	if ( substr( $srcset_part, -$suffix_length ) === $suffix ) {
305	return true;
306	}
307	}
308
309	return false;
310	}
311	}
312