image-source-control-isc / includes / image-sources / analyze-html.php

analyze-html.php in Image Source Control Lite – Show Image Credits and Captions 3.12.0, at includes/image-sources/analyze-html.php

224 lines 6.8 KB

Raw Download Zip

1	<?php
2
3	namespace ISC\Image_Sources;
4
5	use ISC_Log;
6
7	/**
8	* Analyze chunks of HTML
9	*/
10	class Analyze_HTML {
11
12	/**
13	* Extract image-related markup from larger chunks of HTML.
14	*
15	* @param string $html HTML to extract images from.
16	* @return array Array of image markup.
17	*/
18	public function extract_images_from_html( string $html ): array {
19	/**
20	* Removed [caption], because this check runs after the hook that interprets shortcodes
21	* img tag is checked individually since there is a different order of attributes when images are used in gallery or individually
22	*
23	* 0 – full match
24	* 1 - <figure> if set and having a class attribute
25	* 2 – classes from figure tag
26	* 3 – inner code starting with <a>
27	* 4 – opening link tag
28	* 5 – "rel" attribute from link tag
29	* 6 – image id from link wp-att- value in "rel" attribute
30	* 7 – full img tag
31	* 8 – image URL
32	* 9 – closing link tag
33	*
34	* tested with:
35	* * with and without [caption]
36	* * with and without link attribute
37	*
38	* potential issues:
39	* * line breaks in the code – use \s* where potential line breaks could appear
40	*
41	* Use (\x20\|\x9\|\xD\|\xA)+ to match whitespace following HTML starting tag name according to W3C REC 3.1. See issue PR #136
42	*/
43	$pattern = apply_filters( 'isc_public_caption_regex', '#(?:<figure[^>]class=["\']([^"\'])["\'][^>]>\s)?((<a[\x20\|\x9\|\xD\|\xA]+[^>]>)?\s(<img[\x20\|\x9\|\xD\|\xA]+[^>][^>]src=["\'](.+)["\'].\/?>).(\s</a>)??[^<])#isU', $html );
44	preg_match_all( $pattern, $html, $matches, PREG_SET_ORDER );
45
46	/**
47	* Convert the matches into a multidimensional array with string keys to simplify usage and manipulation
48	*/
49	$matches_new = array_map(
50	function ( $match ) {
51	return [
52	'full' => $match[0] ?? '',
53	'figure_class' => $match[1] ?? '',
54	'inner_code' => $match[2] ?? '',
55	'img_src' => $match[5] ?? '',
56	];
57	},
58	$matches
59	);
60
61	/**
62	* Filter matches from regex
63	*/
64	return apply_filters( 'isc_extract_images_from_html', $matches_new, $matches, $pattern );
65	}
66
67	/**
68	* Extract image ID from HTML
69	*
70	* @param string $html HTML to extract an image ID from.
71	* @return int Image ID.
72	*/
73	public function extract_image_id( string $html ): int {
74	if ( ! $html ) {
75	return 0;
76	}
77
78	$id = 0;
79
80	/**
81	* Image IDs seem to be in the following elements:
82	* - img tag with class "wp-image-123"
83	* - img tag with data-id="123"
84	*/
85	$success = preg_match( '#wp-image-(\d+)\|data-id=["\'](\d+)#is', $html, $matches_id );
86	if ( $success ) {
87	$id = $matches_id[1] ? intval( $matches_id[1] ) : intval( $matches_id[2] );
88	ISC_Log::log( sprintf( 'found ID "%s"', $id ) );
89	} else {
90	ISC_Log::log( sprintf( 'no ID found for "%s"', $html ) );
91	}
92
93	return $id;
94	}
95
96	/**
97	* Extract the image URL from HTML, namely, the value of the src attribute
98	*
99	* @param string $html HTML to extract an image URL from.
100	* @return string Image URL.
101	*/
102	public function extract_image_src( string $html ): string {
103	if ( ! $html ) {
104	return '';
105	}
106
107	$src = '';
108
109	$success = preg_match( '#src=["\']([^"\']+)["\']#is', $html, $matches_src );
110	if ( $success ) {
111	$src = $matches_src[1];
112	ISC_Log::log( sprintf( 'found src "%s"', $src ) );
113	} else {
114	ISC_Log::log( sprintf( 'no src found for "%s"', $html ) );
115	}
116
117	return $src;
118	}
119
120	/**
121	* Extract image URLs from HTML.
122	* One image URL per HTML attribute will be found.
123	*
124	* Limitations:
125	* - retrieves the first valid image URL from any HTML tag
126	* - if an IMG tag has a SRC attribute with a valid image URL, all other tags with valid URLs are ignored
127	* - technically, one could generate tags with different images from the Media Library, e.g., having a differently cut URL in a data-attribute
128	* that then shows dynamically using JavaScript
129	* this would cause one of the images not being listed with a source or a used image
130	* since I haven’t seen that in the wild, this case is deliberately ignored
131	*
132	* @param string $html Any HTML code.
133	* @return array List of image URLs.
134	*/
135	public static function extract_image_urls_from_html_tags( $html = '' ): array {
136	$urls = [];
137
138	if ( empty( $html ) ) {
139	ISC_Log::log( 'Exit due to missing empty HTML' );
140	return $urls;
141	}
142
143	ISC_Log::log( 'look for image IDs within the HTML' );
144
145	$types = implode( '\|', \ISC\Image_Sources\Image_Sources::get_allowed_extensions() );
146	/**
147	* Look for any URLs
148	* - starting with http, or https
149	* - ending with a valid image format or " or '
150	* - or any URL in the SRC attribute of IMG tags, where we ignore the extension
151	*
152	* What it does
153	* (?<!, ) – needed to prevent multiple images from `srcset` attribute to being considered here
154	* (http[s]?:) - start with http or https
155	* [^\'^"^ ] - stop at one of these chars since they don't belong into URLs: ", ', or space
156	* or looks for any URL in the SRC attribute
157	*
158	* Limitations
159	* - we don't include images that don't have a full path, e.g., url("../img/image.png") or url("image.png"); they are likely not in the Media library
160	*/
161	$pattern = '#((?<!, )((http[s]?:)[^\'^"^ ]\.(' . $types . ')))\|(<img[\x20\|\x9\|\xD\|\xA]+[^>][^>]src=["\'](.+)["\'].\/?>)#isU';
162
163	/**
164	* Match index
165	* 0 - full match (including first char)
166	* 1 - URLs with valid image file extension
167	* 2 - ignore
168	* 3 – protocol
169	* 4 - file format (e.g. "jpg")
170	* 5 – full img tag
171	* 6 - URLs from img tags – the file extension is ignored here
172	*/
173	preg_match_all( $pattern, $html, $matches );
174
175	// merge matches from both conditions and remove empty ones
176	$urls = array_filter( array_merge( $matches[1], $matches[6] ) );
177	// remove duplicate URLs
178	return array_values( array_unique( $urls ) );
179	}
180
181	/**
182	* Extract any image URLs from HTML.
183	*
184	* @param string $html Any HTML code.
185	*
186	* @return array List of image URLs.
187	*/
188	public static function extract_image_urls( string $html = '' ): array {
189	$urls = [];
190
191	if ( empty( $html ) ) {
192	\ISC_Log::log( 'Exit due to empty HTML' );
193	return $urls;
194	}
195
196	\ISC_Log::log( 'Looking for valid image URLs within HTML' );
197
198	// Get allowed image extensions as a pipe-separated string.
199	$types = implode( '\|', \ISC\Image_Sources\Image_Sources::get_allowed_extensions() );
200
201	/**
202	* The regex below matches:
203	* - URLs starting with "http://" or "https://"
204	* - Followed by one or more non-whitespace characters (non-greedy)
205	* - Ending with a dot and one of the allowed extensions (case-insensitive)
206	* - Optionally followed by a query string
207	* - The URL is expected to be wrapped in either:
208	* -- single or double quotes
209	* -- whitespace
210	* -- brackets "()"
211	*/
212	$pattern = '#https?://\S+?\.(?:' . $types . ')(?:\?\S*)?(?=["\'\s\)])#i';
213
214	preg_match_all( $pattern, $html, $matches );
215
216	if ( ! empty( $matches[0] ) ) {
217	// Remove duplicate URLs.
218	return array_values( array_unique( $matches[0] ) );
219	}
220
221	return $urls;
222	}
223	}
224