source: trunk/include/emogrifier.class.php @ 25344

Revision 25344, 19.8 KB checked in by mistic100, 6 years ago (diff)

feature 2965: apply new template on pwg_mail()
TODO : review other mail functions, configuration GUI

Line 
1<?php
2/*
3UPDATES
4
5    2008-08-10  Fixed CSS comment stripping regex to add PCRE_DOTALL (changed from '/\/\*.*\*\//U' to '/\/\*.*\*\//sU')
6    2008-08-18  Added lines instructing DOMDocument to attempt to normalize HTML before processing
7    2008-10-20  Fixed bug with bad variable name... Thanks Thomas!
8    2008-03-02  Added licensing terms under the MIT License
9                Only remove unprocessable HTML tags if they exist in the array
10    2009-06-03  Normalize existing CSS (style) attributes in the HTML before we process the CSS.
11                Made it so that the display:none stripper doesn't require a trailing semi-colon.
12    2009-08-13  Added support for subset class values (e.g. "p.class1.class2").
13                Added better protection for bad css attributes.
14                Fixed support for HTML entities.
15    2009-08-17  Fixed CSS selector processing so that selectors are processed by precedence/specificity, and not just in order.
16    2009-10-29  Fixed so that selectors appearing later in the CSS will have precedence over identical selectors appearing earlier.
17    2009-11-04  Explicitly declared static functions static to get rid of E_STRICT notices.
18    2010-05-18  Fixed bug where full url filenames with protocols wouldn't get split improperly when we explode on ':'... Thanks Mark!
19                Added two new attribute selectors
20    2010-06-16  Added static caching for less processing overhead in situations where multiple emogrification takes place
21    2010-07-26  Fixed bug where '0' values were getting discarded because of php's empty() function... Thanks Scott!
22    2010-09-03  Added checks to invisible node removal to ensure that we don't try to remove non-existent child nodes of parents that have already been deleted
23    2011-04-08  Fixed errors in CSS->XPath conversion for adjacent sibling selectors and id/class combinations... Thanks Bob V.!
24    2011-06-08  Fixed an error where CSS @media types weren't being parsed correctly... Thanks Will W.!
25    2011-08-03  Fixed an error where an empty selector at the beginning of the CSS would cause a parse error on the next selector... Thanks Alexei T.!
26    2011-10-13  Fully fixed a bug introduced in 2011-06-08 where selectors at the beginning of the CSS would be parsed incorrectly... Thanks Thomas A.!
27    2011-10-26  Added an option to allow you to output emogrified code without extended characters being turned into HTML entities.
28                Moved static references to class attributes so they can be manipulated.
29                Added the ability to clear out the (formerly) static cache when CSS is reloaded.
30    2011-12-22  Fixed a bug that was overwriting existing inline styles from the original HTML... Thanks Sagi L.!
31    2012-01-31  Fixed a bug that was introduced with the 2011-12-22 revision... Thanks Sagi L. and M. Bąkowski!
32                Added extraction of <style> blocks within the HTML due to popular demand.
33                Added several new pseudo-selectors (first-child, last-child, nth-child, and nth-of-type).
34    2012-02-07  Fixed some recent code introductions to use class constants rather than global constants.
35                Fixed some recent code introductions to make it cleaner to read.
36    2012-05-01  Made removal of invisible nodes operate in a case-insensitive manner... Thanks Juha P.!
37    2013-10-10  Add preserveStyleTag option
38*/
39
40define('CACHE_CSS', 0);
41define('CACHE_SELECTOR', 1);
42define('CACHE_XPATH', 2);
43
44class Emogrifier {
45
46    // for calculating nth-of-type and nth-child selectors
47    const INDEX = 0;
48    const MULTIPLIER = 1;
49
50    private $html = '';
51    private $css = '';
52    private $unprocessableHTMLTags = array('wbr');
53    private $caches = array();
54
55    // this attribute applies to the case where you want to preserve your original text encoding.
56    // by default, emogrifier translates your text into HTML entities for two reasons:
57    // 1. because of client incompatibilities, it is better practice to send out HTML entities rather than unicode over email
58    // 2. it translates any illegal XML characters that DOMDocument cannot work with
59    // if you would like to preserve your original encoding, set this attribute to true.
60    public $preserveEncoding = false;
61   
62    // by default, emogrifier removes <style> tags, set preserveStyleTag to true to keep them
63    public $preserveStyleTag = false;
64
65    public function __construct($html = '', $css = '') {
66        $this->html = $html;
67        $this->css  = $css;
68        $this->clearCache();
69    }
70
71    public function setHTML($html = '') { $this->html = $html; }
72    public function setCSS($css = '') {
73        $this->css = $css;
74        $this->clearCache(CACHE_CSS);
75    }
76
77    public function clearCache($key = null) {
78        if (!is_null($key)) {
79            if (isset($this->caches[$key])) $this->caches[$key] = array();
80        } else {
81            $this->caches = array(
82                CACHE_CSS       => array(),
83                CACHE_SELECTOR  => array(),
84                CACHE_XPATH     => array(),
85            );
86        }
87    }
88
89    // there are some HTML tags that DOMDocument cannot process, and will throw an error if it encounters them.
90    // in particular, DOMDocument will complain if you try to use HTML5 tags in an XHTML document.
91    // these functions allow you to add/remove them if necessary.
92    // it only strips them from the code (does not remove actual nodes).
93    public function addUnprocessableHTMLTag($tag) { $this->unprocessableHTMLTags[] = $tag; }
94    public function removeUnprocessableHTMLTag($tag) {
95        if (($key = array_search($tag,$this->unprocessableHTMLTags)) !== false)
96            unset($this->unprocessableHTMLTags[$key]);
97    }
98
99    // applies the CSS you submit to the html you submit. places the css inline
100    public function emogrify() {
101        $body = $this->html;
102
103        // remove any unprocessable HTML tags (tags that DOMDocument cannot parse; this includes wbr and many new HTML5 tags)
104        if (count($this->unprocessableHTMLTags)) {
105            $unprocessableHTMLTags = implode('|',$this->unprocessableHTMLTags);
106            $body = preg_replace("/<\/?($unprocessableHTMLTags)[^>]*>/i",'',$body);
107        }
108
109        $encoding = mb_detect_encoding($body);
110        $body = mb_convert_encoding($body, 'HTML-ENTITIES', $encoding);
111
112        $xmldoc = new DOMDocument;
113        $xmldoc->encoding = $encoding;
114        $xmldoc->strictErrorChecking = false;
115        $xmldoc->formatOutput = true;
116        $xmldoc->loadHTML($body);
117        $xmldoc->normalizeDocument();
118
119        $xpath = new DOMXPath($xmldoc);
120
121        // before be begin processing the CSS file, parse the document and normalize all existing CSS attributes (changes 'DISPLAY: none' to 'display: none');
122        // we wouldn't have to do this if DOMXPath supported XPath 2.0.
123        // also store a reference of nodes with existing inline styles so we don't overwrite them
124        $vistedNodes = $vistedNodeRef = array();
125        $nodes = @$xpath->query('//*[@style]');
126        foreach ($nodes as $node) {
127            $normalizedOrigStyle = preg_replace('/[A-z\-]+(?=\:)/Se',"strtolower('\\0')", $node->getAttribute('style'));
128
129            // in order to not overwrite existing style attributes in the HTML, we have to save the original HTML styles
130            $nodeKey = md5($node->getNodePath());
131            if (!isset($vistedNodeRef[$nodeKey])) {
132                $vistedNodeRef[$nodeKey] = $this->cssStyleDefinitionToArray($normalizedOrigStyle);
133                $vistedNodes[$nodeKey]   = $node;
134            }
135
136            $node->setAttribute('style', $normalizedOrigStyle);
137        }
138
139        // grab any existing style blocks from the html and append them to the existing CSS
140        // (these blocks should be appended so as to have precedence over conflicting styles in the existing CSS)
141        $css = $this->css;
142        $nodes = @$xpath->query('//style');
143        foreach ($nodes as $node) {
144            // append the css
145            $css .= "\n\n{$node->nodeValue}";
146            // remove the <style> node
147            if (!$this->preserveStyleTag) {
148                $node->parentNode->removeChild($node);
149            }
150        }
151
152        // filter the CSS
153        $search = array(
154            '/\/\*.*\*\//sU', // get rid of css comment code
155            '/^\s*@import\s[^;]+;/misU', // strip out any import directives
156            '/^\s*@media\s[^{]+{\s*}/misU', // strip any empty media enclosures
157            '/^\s*@media\s+((aural|braille|embossed|handheld|print|projection|speech|tty|tv)\s*,*\s*)+{.*}\s*}/misU', // strip out all media types that are not 'screen' or 'all' (these don't apply to email)
158            '/^\s*@media\s[^{]+{(.*})\s*}/misU', // get rid of remaining media type enclosures
159        );
160
161        $replace = array(
162            '',
163            '',
164            '',
165            '',
166            '\\1',
167        );
168
169        $css = preg_replace($search, $replace, $css);
170
171        $csskey = md5($css);
172        if (!isset($this->caches[CACHE_CSS][$csskey])) {
173
174            // process the CSS file for selectors and definitions
175            preg_match_all('/(^|[^{}])\s*([^{]+){([^}]*)}/mis', $css, $matches, PREG_SET_ORDER);
176
177            $all_selectors = array();
178            foreach ($matches as $key => $selectorString) {
179                // if there is a blank definition, skip
180                if (!strlen(trim($selectorString[3]))) continue;
181
182                // else split by commas and duplicate attributes so we can sort by selector precedence
183                $selectors = explode(',',$selectorString[2]);
184                foreach ($selectors as $selector) {
185
186                    // don't process pseudo-elements and behavioral (dynamic) pseudo-classes; ONLY allow structural pseudo-classes
187                    if (strpos($selector, ':') !== false && !preg_match('/:\S+\-(child|type)\(/i', $selector)) continue;
188
189                    $all_selectors[] = array('selector' => trim($selector),
190                                             'attributes' => trim($selectorString[3]),
191                                             'line' => $key, // keep track of where it appears in the file, since order is important
192                    );
193                }
194            }
195
196            // now sort the selectors by precedence
197            usort($all_selectors, array($this,'sortBySelectorPrecedence'));
198
199            $this->caches[CACHE_CSS][$csskey] = $all_selectors;
200        }
201
202        foreach ($this->caches[CACHE_CSS][$csskey] as $value) {
203
204            // query the body for the xpath selector
205            $nodes = $xpath->query($this->translateCSStoXpath(trim($value['selector'])));
206
207            foreach($nodes as $node) {
208                // if it has a style attribute, get it, process it, and append (overwrite) new stuff
209                if ($node->hasAttribute('style')) {
210                    // break it up into an associative array
211                    $oldStyleArr = $this->cssStyleDefinitionToArray($node->getAttribute('style'));
212                    $newStyleArr = $this->cssStyleDefinitionToArray($value['attributes']);
213
214                    // new styles overwrite the old styles (not technically accurate, but close enough)
215                    $combinedArr = array_merge($oldStyleArr,$newStyleArr);
216                    $style = '';
217                    foreach ($combinedArr as $k => $v) $style .= (strtolower($k) . ':' . $v . ';');
218                } else {
219                    // otherwise create a new style
220                    $style = trim($value['attributes']);
221                }
222                $node->setAttribute('style', $style);
223            }
224        }
225
226        // now iterate through the nodes that contained inline styles in the original HTML
227        foreach ($vistedNodeRef as $nodeKey => $origStyleArr) {
228            $node = $vistedNodes[$nodeKey];
229            $currStyleArr = $this->cssStyleDefinitionToArray($node->getAttribute('style'));
230
231            $combinedArr = array_merge($currStyleArr, $origStyleArr);
232            $style = '';
233            foreach ($combinedArr as $k => $v) $style .= (strtolower($k) . ':' . $v . ';');
234
235            $node->setAttribute('style', $style);
236        }
237
238        // This removes styles from your email that contain display:none.
239        // We need to look for display:none, but we need to do a case-insensitive search. Since DOMDocument only supports XPath 1.0,
240        // lower-case() isn't available to us. We've thus far only set attributes to lowercase, not attribute values. Consequently, we need
241        // to translate() the letters that would be in 'NONE' ("NOE") to lowercase.
242        $nodes = $xpath->query('//*[contains(translate(translate(@style," ",""),"NOE","noe"),"display:none")]');
243        // The checks on parentNode and is_callable below ensure that if we've deleted the parent node,
244        // we don't try to call removeChild on a nonexistent child node
245        if ($nodes->length > 0)
246            foreach ($nodes as $node)
247                if ($node->parentNode && is_callable(array($node->parentNode,'removeChild')))
248                        $node->parentNode->removeChild($node);
249
250        if ($this->preserveEncoding) {
251            return mb_convert_encoding($xmldoc->saveHTML(), $encoding, 'HTML-ENTITIES');
252        } else {
253            return $xmldoc->saveHTML();
254        }
255    }
256
257    private function sortBySelectorPrecedence($a, $b) {
258        $precedenceA = $this->getCSSSelectorPrecedence($a['selector']);
259        $precedenceB = $this->getCSSSelectorPrecedence($b['selector']);
260
261        // we want these sorted ascendingly so selectors with lesser precedence get processed first and
262        // selectors with greater precedence get sorted last
263        return ($precedenceA == $precedenceB) ? ($a['line'] < $b['line'] ? -1 : 1) : ($precedenceA < $precedenceB ? -1 : 1);
264    }
265
266    private function getCSSSelectorPrecedence($selector) {
267        $selectorkey = md5($selector);
268        if (!isset($this->caches[CACHE_SELECTOR][$selectorkey])) {
269            $precedence = 0;
270            $value = 100;
271            $search = array('\#','\.',''); // ids: worth 100, classes: worth 10, elements: worth 1
272
273            foreach ($search as $s) {
274                if (trim($selector == '')) break;
275                $num = 0;
276                $selector = preg_replace('/'.$s.'\w+/','',$selector,-1,$num);
277                $precedence += ($value * $num);
278                $value /= 10;
279            }
280            $this->caches[CACHE_SELECTOR][$selectorkey] = $precedence;
281        }
282
283        return $this->caches[CACHE_SELECTOR][$selectorkey];
284    }
285
286    // right now we support all CSS 1 selectors and most CSS2/3 selectors.
287    // http://plasmasturm.org/log/444/
288    private function translateCSStoXpath($css_selector) {
289
290        $css_selector = trim($css_selector);
291        $xpathkey = md5($css_selector);
292        if (!isset($this->caches[CACHE_XPATH][$xpathkey])) {
293            // returns an Xpath selector
294            $search = array(
295                               '/\s+>\s+/', // Matches any element that is a child of parent.
296                               '/\s+\+\s+/', // Matches any element that is an adjacent sibling.
297                               '/\s+/', // Matches any element that is a descendant of an parent element element.
298                               '/([^\/]+):first-child/i', // first-child pseudo-selector
299                               '/([^\/]+):last-child/i', // last-child pseudo-selector
300                               '/(\w)\[(\w+)\]/', // Matches element with attribute
301                               '/(\w)\[(\w+)\=[\'"]?(\w+)[\'"]?\]/', // Matches element with EXACT attribute
302                               '/(\w+)?\#([\w\-]+)/e', // Matches id attributes
303                               '/(\w+|[\*\]])?((\.[\w\-]+)+)/e', // Matches class attributes
304
305            );
306            $replace = array(
307                               '/',
308                               '/following-sibling::*[1]/self::',
309                               '//',
310                               '*[1]/self::\\1',
311                               '*[last()]/self::\\1',
312                               '\\1[@\\2]',
313                               '\\1[@\\2="\\3"]',
314                               "(strlen('\\1') ? '\\1' : '*').'[@id=\"\\2\"]'",
315                               "(strlen('\\1') ? '\\1' : '*').'[contains(concat(\" \",@class,\" \"),concat(\" \",\"'.implode('\",\" \"))][contains(concat(\" \",@class,\" \"),concat(\" \",\"',explode('.',substr('\\2',1))).'\",\" \"))]'",
316            );
317
318            $css_selector = '//'.preg_replace($search, $replace, $css_selector);
319
320            // advanced selectors are going to require a bit more advanced emogrification
321            // if we required PHP 5.3 we could do this with closures
322            $css_selector = preg_replace_callback('/([^\/]+):nth-child\(\s*(odd|even|[+\-]?\d|[+\-]?\d?n(\s*[+\-]\s*\d)?)\s*\)/i', array($this, 'translateNthChild'), $css_selector);
323            $css_selector = preg_replace_callback('/([^\/]+):nth-of-type\(\s*(odd|even|[+\-]?\d|[+\-]?\d?n(\s*[+\-]\s*\d)?)\s*\)/i', array($this, 'translateNthOfType'), $css_selector);
324
325            $this->caches[CACHE_SELECTOR][$xpathkey] = $css_selector;
326        }
327        return $this->caches[CACHE_SELECTOR][$xpathkey];
328    }
329
330    private function translateNthChild($match) {
331
332        $result = $this->parseNth($match);
333
334        if (isset($result[self::MULTIPLIER])) {
335            if ($result[self::MULTIPLIER] < 0) {
336                $result[self::MULTIPLIER] = abs($result[self::MULTIPLIER]);
337                return sprintf("*[(last() - position()) mod %u = %u]/self::%s", $result[self::MULTIPLIER], $result[self::INDEX], $match[1]);
338            } else {
339                return sprintf("*[position() mod %u = %u]/self::%s", $result[self::MULTIPLIER], $result[self::INDEX], $match[1]);
340            }
341        } else {
342            return sprintf("*[%u]/self::%s", $result[self::INDEX], $match[1]);
343        }
344    }
345
346    private function translateNthOfType($match) {
347
348        $result = $this->parseNth($match);
349
350        if (isset($result[self::MULTIPLIER])) {
351            if ($result[self::MULTIPLIER] < 0) {
352                $result[self::MULTIPLIER] = abs($result[self::MULTIPLIER]);
353                return sprintf("%s[(last() - position()) mod %u = %u]", $match[1], $result[self::MULTIPLIER], $result[self::INDEX]);
354            } else {
355                return sprintf("%s[position() mod %u = %u]", $match[1], $result[self::MULTIPLIER], $result[self::INDEX]);
356            }
357        } else {
358            return sprintf("%s[%u]", $match[1], $result[self::INDEX]);
359        }
360    }
361
362    private function parseNth($match) {
363
364        if (in_array(strtolower($match[2]), array('even','odd'))) {
365            $index = strtolower($match[2]) == 'even' ? 0 : 1;
366            return array(self::MULTIPLIER => 2, self::INDEX => $index);
367        // if there is a multiplier
368        } else if (stripos($match[2], 'n') === false) {
369            $index = intval(str_replace(' ', '', $match[2]));
370            return array(self::INDEX => $index);
371        } else {
372
373            if (isset($match[3])) {
374                $multiple_term = str_replace($match[3], '', $match[2]);
375                $index = intval(str_replace(' ', '', $match[3]));
376            } else {
377                $multiple_term = $match[2];
378                $index = 0;
379            }
380
381            $multiplier = str_ireplace('n', '', $multiple_term);
382
383            if (!strlen($multiplier)) $multiplier = 1;
384            elseif ($multiplier == 0) return array(self::INDEX => $index);
385            else $multiplier = intval($multiplier);
386
387            while ($index < 0) $index += abs($multiplier);
388
389            return array(self::MULTIPLIER => $multiplier, self::INDEX => $index);
390        }
391    }
392
393    private function cssStyleDefinitionToArray($style) {
394        $definitions = explode(';',$style);
395        $retArr = array();
396        foreach ($definitions as $def) {
397            if (empty($def) || strpos($def, ':') === false) continue;
398            list($key,$value) = explode(':',$def,2);
399            if (empty($key) || strlen(trim($value)) === 0) continue;
400            $retArr[trim($key)] = trim($value);
401        }
402        return $retArr;
403    }
404}
Note: See TracBrowser for help on using the repository browser.