class-tiny-source-base.php — TinyPNG – JPEG, PNG & WebP image compression 3.8.0

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

tiny-compress-images / src Last commit date

class-tiny-source-base.php

313 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	* @param string $html string containg html
128	* @return array{path: string, size: string}[] The image sources
129	*/
130	protected function get_image_srcsets( $html ) {
131	$result = array();
132	$srcset = $this::get_attribute_value( $html, 'srcset' );
133
134	if ( $srcset ) {
135	// Split the srcset to get individual entries
136	$srcset_entries = explode( ',', $srcset );
137
138	foreach ( $srcset_entries as $entry ) {
139	// Trim whitespace
140	$entry = trim( $entry );
141
142	// Split by whitespace to separate path and size/density descriptor
143	$parts = preg_split( '/\s+/', $entry, 2 );
144
145	if ( count( $parts ) === 2 ) {
146	// We have both path and size
147	$result[] = array(
148	'path' => $parts[0],
149	'size' => $parts[1],
150	);
151	} elseif ( count( $parts ) === 1 ) {
152	// We only have a path, will be interpreted as pixel
153	// density 1x (unusual in srcset)
154	$result[] = array(
155	'path' => $parts[0],
156	'size' => '',
157	);
158	}
159	}
160	}
161	return $result;
162	}
163
164	/**
165	* Retrieves the sources from the <img> or <source> element
166	*
167	* @return array{path: string, size: string}[] The image sources
168	*/
169	private function get_image_src( $html ) {
170	$source = $this::get_attribute_value( $html, 'src' );
171	if ( ! empty( $source ) ) {
172	// No srcset, but we have a src attribute
173	return array(
174	'path' => $source,
175	'size' => '',
176	);
177	}
178	return array();
179	}
180
181
182	/**
183	* Creates one or more <source> elements if alternative formats
184	* are available.
185	*
186	* @param string $original_source_html, either <source> or <img>
187	* @return array{string} array of <source> html
188	*/
189	protected function create_alternative_sources( $original_source_html ) {
190	$srcsets = $this->get_image_srcsets( $original_source_html );
191	if ( empty( $srcsets ) ) {
192	// no srcset, try src attribute
193	$src = $this->get_image_src( $original_source_html );
194	if ( ! empty( $src ) ) {
195	$srcsets[] = $src;
196	}
197	}
198
199	if ( empty( $srcsets ) ) {
200	return array();
201	}
202
203	$is_source_tag = (bool) preg_match( '#<source\b#i', $original_source_html );
204
205	$sources = array();
206	$width_descriptor = $this->get_largest_width_descriptor( $srcsets );
207
208	foreach ( $this->valid_mimetypes as $mimetype ) {
209	$srcset_parts = array();
210
211	foreach ( $srcsets as $srcset ) {
212	$alt_source = $this->get_formatted_source( $srcset, $mimetype );
213	if ( $alt_source ) {
214	$srcset_parts[] = trim( $alt_source['src'] . ' ' . $alt_source['size'] );
215	}
216	}
217
218	if (
219	$width_descriptor &&
220	! self::srcset_contains_width_descriptor(
221	$srcset_parts,
222	$width_descriptor
223	)
224	) {
225	continue;
226	}
227
228	if ( empty( $srcset_parts ) ) {
229	continue;
230	}
231
232	$source_attr_parts = array();
233
234	$srcset_attr = implode( ', ', $srcset_parts );
235	$source_attr_parts['srcset'] = $srcset_attr;
236
237	if ( $is_source_tag ) {
238	foreach ( array( 'sizes', 'media', 'width', 'height' ) as $attr ) {
239	$attr_value = $this->get_attribute_value( $original_source_html, $attr );
240	if ( $attr_value ) {
241	$source_attr_parts[ $attr ] = $attr_value;
242	}
243	}
244	} else {
245	$sizes_value = $this->get_attribute_value( $original_source_html, 'sizes' );
246	if ( $sizes_value ) {
247	$source_attr_parts['sizes'] = $sizes_value;
248	}
249	}
250
251	$source_attr_parts['type'] = $mimetype;
252	$source_parts = array( '<source' );
253	foreach ( $source_attr_parts as $source_attr_name => $source_attr_val ) {
254	$source_parts[] = $source_attr_name . '="' . esc_attr( $source_attr_val ) . '"';
255	}
256	$source_parts[] = '/>';
257	$sources[] = implode( ' ', $source_parts );
258	} // End foreach().
259
260	return $sources;
261	}
262
263	/**
264	* Returns the largest numeric width descriptor
265	* (e.g. 2000 from "2000w") found in the srcset data.
266	*
267	* @param array<array{path: string, size: string}> $srcsets
268	* @return int
269	*/
270	public static function get_largest_width_descriptor( $srcsets ) {
271	$largest = 0;
272
273	foreach ( $srcsets as $srcset ) {
274	if ( empty( $srcset['size'] ) ) {
275	continue;
276	}
277
278	if ( preg_match( '/(\d+)w/', $srcset['size'], $matches ) ) {
279	$width = (int) $matches[1];
280	if ( $width > $largest ) {
281	$largest = $width;
282	}
283	}
284	}
285
286	return $largest;
287	}
288
289	/**
290	* Determines whether a srcset list contains the provided width descriptor.
291	*
292	* @param string[] $srcset_parts
293	* @param int $width_descriptor
294	* @return bool true if width is in srcset
295	*/
296	public static function srcset_contains_width_descriptor( $srcset_parts, $width_descriptor ) {
297	if ( empty( $srcset_parts ) \|\| $width_descriptor <= 0 ) {
298	return false;
299	}
300
301	$suffix = ' ' . $width_descriptor . 'w';
302	$suffix_length = strlen( $suffix );
303
304	foreach ( $srcset_parts as $srcset_part ) {
305	if ( substr( $srcset_part, -$suffix_length ) === $suffix ) {
306	return true;
307	}
308	}
309
310	return false;
311	}
312	}
313