1
|
|
|
<?php |
2
|
|
|
|
3
|
|
|
/** |
4
|
|
|
* @package s9e\TextFormatter |
5
|
|
|
* @copyright Copyright (c) 2010-2018 The s9e Authors |
6
|
|
|
* @license http://www.opensource.org/licenses/mit-license.php The MIT License |
7
|
|
|
*/ |
8
|
|
|
namespace s9e\TextFormatter\Configurator\Helpers; |
9
|
|
|
|
10
|
|
|
use DOMAttr; |
11
|
|
|
use DOMCharacterData; |
12
|
|
|
use DOMDocument; |
13
|
|
|
use DOMElement; |
14
|
|
|
use DOMNode; |
15
|
|
|
use DOMProcessingInstruction; |
16
|
|
|
use DOMText; |
17
|
|
|
use DOMXPath; |
18
|
|
|
use RuntimeException; |
19
|
|
|
use s9e\TextFormatter\Configurator\Helpers\RegexpBuilder; |
20
|
|
|
|
21
|
|
|
abstract class TemplateHelper |
22
|
|
|
{ |
23
|
|
|
/** |
24
|
|
|
* XSL namespace |
25
|
|
|
*/ |
26
|
|
|
const XMLNS_XSL = 'http://www.w3.org/1999/XSL/Transform'; |
27
|
|
|
|
28
|
|
|
/** |
29
|
|
|
* Return all attributes (literal or generated) that match given regexp |
30
|
|
|
* |
31
|
|
|
* @param DOMDocument $dom Document |
32
|
|
|
* @param string $regexp Regexp |
33
|
|
|
* @return array Array of DOMNode instances |
34
|
|
|
*/ |
35
|
|
|
public static function getAttributesByRegexp(DOMDocument $dom, $regexp) |
36
|
|
|
{ |
37
|
|
|
$xpath = new DOMXPath($dom); |
38
|
|
|
$nodes = []; |
39
|
|
|
|
40
|
|
|
// Get literal attributes |
41
|
|
|
foreach ($xpath->query('//@*') as $attribute) |
42
|
|
|
{ |
43
|
|
|
if (preg_match($regexp, $attribute->name)) |
44
|
|
|
{ |
45
|
|
|
$nodes[] = $attribute; |
46
|
|
|
} |
47
|
|
|
} |
48
|
|
|
|
49
|
|
|
// Get generated attributes |
50
|
|
|
foreach ($xpath->query('//xsl:attribute') as $attribute) |
51
|
|
|
{ |
52
|
|
|
if (preg_match($regexp, $attribute->getAttribute('name'))) |
53
|
|
|
{ |
54
|
|
|
$nodes[] = $attribute; |
55
|
|
|
} |
56
|
|
|
} |
57
|
|
|
|
58
|
|
|
// Get attributes created with <xsl:copy-of/> |
59
|
|
|
foreach ($xpath->query('//xsl:copy-of') as $node) |
60
|
|
|
{ |
61
|
|
|
$expr = $node->getAttribute('select'); |
62
|
|
|
|
63
|
|
|
if (preg_match('/^@(\\w+)$/', $expr, $m) |
64
|
|
|
&& preg_match($regexp, $m[1])) |
65
|
|
|
{ |
66
|
|
|
$nodes[] = $node; |
67
|
|
|
} |
68
|
|
|
} |
69
|
|
|
|
70
|
|
|
return $nodes; |
71
|
|
|
} |
72
|
|
|
|
73
|
|
|
/** |
74
|
|
|
* Return all DOMNodes whose content is CSS |
75
|
|
|
* |
76
|
|
|
* @param DOMDocument $dom Document |
77
|
|
|
* @return array Array of DOMNode instances |
78
|
|
|
*/ |
79
|
|
|
public static function getCSSNodes(DOMDocument $dom) |
80
|
|
|
{ |
81
|
|
|
$regexp = '/^style$/i'; |
82
|
|
|
$nodes = array_merge( |
83
|
|
|
self::getAttributesByRegexp($dom, $regexp), |
84
|
|
|
self::getElementsByRegexp($dom, '/^style$/i') |
85
|
|
|
); |
86
|
|
|
|
87
|
|
|
return $nodes; |
88
|
|
|
} |
89
|
|
|
|
90
|
|
|
/** |
91
|
|
|
* Return all elements (literal or generated) that match given regexp |
92
|
|
|
* |
93
|
|
|
* @param DOMDocument $dom Document |
94
|
|
|
* @param string $regexp Regexp |
95
|
|
|
* @return array Array of DOMNode instances |
96
|
|
|
*/ |
97
|
|
|
public static function getElementsByRegexp(DOMDocument $dom, $regexp) |
98
|
|
|
{ |
99
|
|
|
$xpath = new DOMXPath($dom); |
100
|
|
|
$nodes = []; |
101
|
|
|
|
102
|
|
|
// Get literal attributes |
103
|
|
|
foreach ($xpath->query('//*') as $element) |
104
|
|
|
{ |
105
|
|
|
if (preg_match($regexp, $element->localName)) |
106
|
|
|
{ |
107
|
|
|
$nodes[] = $element; |
108
|
|
|
} |
109
|
|
|
} |
110
|
|
|
|
111
|
|
|
// Get generated elements |
112
|
|
|
foreach ($xpath->query('//xsl:element') as $element) |
113
|
|
|
{ |
114
|
|
|
if (preg_match($regexp, $element->getAttribute('name'))) |
115
|
|
|
{ |
116
|
|
|
$nodes[] = $element; |
117
|
|
|
} |
118
|
|
|
} |
119
|
|
|
|
120
|
|
|
// Get elements created with <xsl:copy-of/> |
121
|
|
|
// NOTE: this method of creating elements is disallowed by default |
122
|
|
|
foreach ($xpath->query('//xsl:copy-of') as $node) |
123
|
|
|
{ |
124
|
|
|
$expr = $node->getAttribute('select'); |
125
|
|
|
|
126
|
|
|
if (preg_match('/^\\w+$/', $expr) |
127
|
|
|
&& preg_match($regexp, $expr)) |
128
|
|
|
{ |
129
|
|
|
$nodes[] = $node; |
130
|
|
|
} |
131
|
|
|
} |
132
|
|
|
|
133
|
|
|
return $nodes; |
134
|
|
|
} |
135
|
|
|
|
136
|
|
|
/** |
137
|
|
|
* Return all DOMNodes whose content is JavaScript |
138
|
|
|
* |
139
|
|
|
* @param DOMDocument $dom Document |
140
|
|
|
* @return array Array of DOMNode instances |
141
|
|
|
*/ |
142
|
|
|
public static function getJSNodes(DOMDocument $dom) |
143
|
|
|
{ |
144
|
|
|
$regexp = '/^(?>data-s9e-livepreview-postprocess$|on)/i'; |
145
|
|
|
$nodes = array_merge( |
146
|
|
|
self::getAttributesByRegexp($dom, $regexp), |
147
|
|
|
self::getElementsByRegexp($dom, '/^script$/i') |
148
|
|
|
); |
149
|
|
|
|
150
|
|
|
return $nodes; |
151
|
|
|
} |
152
|
|
|
|
153
|
|
|
/** |
154
|
|
|
* Return all elements (literal or generated) that match given regexp |
155
|
|
|
* |
156
|
|
|
* Will return all <param/> descendants of <object/> and all attributes of <embed/> whose name |
157
|
|
|
* matches given regexp. This method will NOT catch <param/> elements whose 'name' attribute is |
158
|
|
|
* set via an <xsl:attribute/> |
159
|
|
|
* |
160
|
|
|
* @param DOMDocument $dom Document |
161
|
|
|
* @param string $regexp |
162
|
|
|
* @return array Array of DOMNode instances |
163
|
|
|
*/ |
164
|
|
|
public static function getObjectParamsByRegexp(DOMDocument $dom, $regexp) |
165
|
|
|
{ |
166
|
|
|
$xpath = new DOMXPath($dom); |
167
|
|
|
$nodes = []; |
168
|
|
|
|
169
|
|
|
// Collect attributes from <embed/> elements |
170
|
|
|
foreach (self::getAttributesByRegexp($dom, $regexp) as $attribute) |
171
|
|
|
{ |
172
|
|
|
if ($attribute->nodeType === XML_ATTRIBUTE_NODE) |
173
|
|
|
{ |
174
|
|
|
if (strtolower($attribute->parentNode->localName) === 'embed') |
175
|
|
|
{ |
176
|
|
|
$nodes[] = $attribute; |
177
|
|
|
} |
178
|
|
|
} |
179
|
|
|
elseif ($xpath->evaluate('ancestor::embed', $attribute)) |
180
|
|
|
{ |
181
|
|
|
// Assuming <xsl:attribute/> or <xsl:copy-of/> |
182
|
|
|
$nodes[] = $attribute; |
183
|
|
|
} |
184
|
|
|
} |
185
|
|
|
|
186
|
|
|
// Collect <param/> descendants of <object/> elements |
187
|
|
|
foreach ($dom->getElementsByTagName('object') as $object) |
188
|
|
|
{ |
189
|
|
|
foreach ($object->getElementsByTagName('param') as $param) |
190
|
|
|
{ |
191
|
|
|
if (preg_match($regexp, $param->getAttribute('name'))) |
192
|
|
|
{ |
193
|
|
|
$nodes[] = $param; |
194
|
|
|
} |
195
|
|
|
} |
196
|
|
|
} |
197
|
|
|
|
198
|
|
|
return $nodes; |
199
|
|
|
} |
200
|
|
|
|
201
|
|
|
/** |
202
|
|
|
* Return a list of parameters in use in given XSL |
203
|
|
|
* |
204
|
|
|
* @param string $xsl XSL source |
205
|
|
|
* @return array Alphabetically sorted list of unique parameter names |
206
|
|
|
*/ |
207
|
|
|
public static function getParametersFromXSL($xsl) |
208
|
|
|
{ |
209
|
|
|
$paramNames = []; |
210
|
|
|
|
211
|
|
|
// Wrap the XSL in boilerplate code because it might not have a root element |
212
|
|
|
$xsl = '<xsl:stylesheet xmlns:xsl="' . self::XMLNS_XSL . '">' |
213
|
|
|
. '<xsl:template>' |
214
|
|
|
. $xsl |
215
|
|
|
. '</xsl:template>' |
216
|
|
|
. '</xsl:stylesheet>'; |
217
|
|
|
|
218
|
|
|
$dom = new DOMDocument; |
219
|
|
|
$dom->loadXML($xsl); |
220
|
|
|
|
221
|
|
|
$xpath = new DOMXPath($dom); |
222
|
|
|
|
223
|
|
|
// Start by collecting XPath expressions in XSL elements |
224
|
|
|
$query = '//xsl:*/@match | //xsl:*/@select | //xsl:*/@test'; |
225
|
|
|
foreach ($xpath->query($query) as $attribute) |
226
|
|
|
{ |
227
|
|
|
foreach (XPathHelper::getVariables($attribute->value) as $varName) |
228
|
|
|
{ |
229
|
|
|
// Test whether this is the name of a local variable |
230
|
|
|
$varQuery = 'ancestor-or-self::*/' |
231
|
|
|
. 'preceding-sibling::xsl:variable[@name="' . $varName . '"]'; |
232
|
|
|
|
233
|
|
|
if (!$xpath->query($varQuery, $attribute)->length) |
234
|
|
|
{ |
235
|
|
|
$paramNames[] = $varName; |
236
|
|
|
} |
237
|
|
|
} |
238
|
|
|
} |
239
|
|
|
|
240
|
|
|
// Collecting XPath expressions in attribute value templates |
241
|
|
|
$query = '//*[namespace-uri() != "' . self::XMLNS_XSL . '"]' |
242
|
|
|
. '/@*[contains(., "{")]'; |
243
|
|
|
foreach ($xpath->query($query) as $attribute) |
244
|
|
|
{ |
245
|
|
|
$tokens = AVTHelper::parse($attribute->value); |
246
|
|
|
|
247
|
|
|
foreach ($tokens as $token) |
248
|
|
|
{ |
249
|
|
|
if ($token[0] !== 'expression') |
250
|
|
|
{ |
251
|
|
|
continue; |
252
|
|
|
} |
253
|
|
|
|
254
|
|
|
foreach (XPathHelper::getVariables($token[1]) as $varName) |
255
|
|
|
{ |
256
|
|
|
// Test whether this is the name of a local variable |
257
|
|
|
$varQuery = 'ancestor-or-self::*/' |
258
|
|
|
. 'preceding-sibling::xsl:variable[@name="' . $varName . '"]'; |
259
|
|
|
|
260
|
|
|
if (!$xpath->query($varQuery, $attribute)->length) |
261
|
|
|
{ |
262
|
|
|
$paramNames[] = $varName; |
263
|
|
|
} |
264
|
|
|
} |
265
|
|
|
} |
266
|
|
|
} |
267
|
|
|
|
268
|
|
|
// Dedupe and sort names |
269
|
|
|
$paramNames = array_unique($paramNames); |
270
|
|
|
sort($paramNames); |
271
|
|
|
|
272
|
|
|
return $paramNames; |
273
|
|
|
} |
274
|
|
|
|
275
|
|
|
/** |
276
|
|
|
* Return all DOMNodes whose content is an URL |
277
|
|
|
* |
278
|
|
|
* NOTE: it will also return HTML4 nodes whose content is an URI |
279
|
|
|
* |
280
|
|
|
* @param DOMDocument $dom Document |
281
|
|
|
* @return array Array of DOMNode instances |
282
|
|
|
*/ |
283
|
|
|
public static function getURLNodes(DOMDocument $dom) |
284
|
|
|
{ |
285
|
|
|
$regexp = '/(?>^(?>action|background|c(?>ite|lassid|odebase)|data|formaction|href|icon|longdesc|manifest|p(?>ing|luginspage|oster|rofile)|usemap)|src)$/i'; |
286
|
|
|
$nodes = self::getAttributesByRegexp($dom, $regexp); |
287
|
|
|
|
288
|
|
|
/** |
289
|
|
|
* @link http://helpx.adobe.com/flash/kb/object-tag-syntax-flash-professional.html |
290
|
|
|
* @link http://www.sitepoint.com/control-internet-explorer/ |
291
|
|
|
*/ |
292
|
|
|
foreach (self::getObjectParamsByRegexp($dom, '/^(?:dataurl|movie)$/i') as $param) |
293
|
|
|
{ |
294
|
|
|
$node = $param->getAttributeNode('value'); |
295
|
|
|
if ($node) |
296
|
|
|
{ |
297
|
|
|
$nodes[] = $node; |
298
|
|
|
} |
299
|
|
|
} |
300
|
|
|
|
301
|
|
|
return $nodes; |
302
|
|
|
} |
303
|
|
|
|
304
|
|
|
/** |
305
|
|
|
* Highlight the source of a node inside of a template |
306
|
|
|
* |
307
|
|
|
* @param DOMNode $node Node to highlight |
308
|
|
|
* @param string $prepend HTML to prepend |
309
|
|
|
* @param string $append HTML to append |
310
|
|
|
* @return string Template's source, as HTML |
311
|
|
|
*/ |
312
|
|
|
public static function highlightNode(DOMNode $node, $prepend, $append) |
313
|
|
|
{ |
314
|
|
|
// Add a unique token to the node |
315
|
|
|
$uniqid = uniqid('_'); |
316
|
|
|
if ($node instanceof DOMAttr) |
317
|
|
|
{ |
318
|
|
|
$node->value .= $uniqid; |
319
|
|
|
} |
320
|
|
|
elseif ($node instanceof DOMElement) |
321
|
|
|
{ |
322
|
|
|
$node->setAttribute($uniqid, ''); |
323
|
|
|
} |
324
|
|
|
elseif ($node instanceof DOMCharacterData |
325
|
|
|
|| $node instanceof DOMProcessingInstruction) |
326
|
|
|
{ |
327
|
|
|
$node->data .= $uniqid; |
328
|
|
|
} |
329
|
|
|
|
330
|
|
|
$dom = $node->ownerDocument; |
331
|
|
|
$dom->formatOutput = true; |
332
|
|
|
|
333
|
|
|
$docXml = self::innerXML($dom->documentElement); |
334
|
|
|
$docXml = trim(str_replace("\n ", "\n", $docXml)); |
335
|
|
|
|
336
|
|
|
$nodeHtml = htmlspecialchars(trim($dom->saveXML($node))); |
337
|
|
|
$docHtml = htmlspecialchars($docXml); |
338
|
|
|
|
339
|
|
|
// Enclose the node's representation in our hilighting HTML |
340
|
|
|
$html = str_replace($nodeHtml, $prepend . $nodeHtml . $append, $docHtml); |
341
|
|
|
|
342
|
|
|
// Remove the unique token from HTML and from the node |
343
|
|
|
if ($node instanceof DOMAttr) |
344
|
|
|
{ |
345
|
|
|
$node->value = substr($node->value, 0, -strlen($uniqid)); |
346
|
|
|
$html = str_replace($uniqid, '', $html); |
347
|
|
|
} |
348
|
|
|
elseif ($node instanceof DOMElement) |
349
|
|
|
{ |
350
|
|
|
$node->removeAttribute($uniqid); |
351
|
|
|
$html = str_replace(' ' . $uniqid . '=""', '', $html); |
352
|
|
|
} |
353
|
|
|
elseif ($node instanceof DOMCharacterData |
354
|
|
|
|| $node instanceof DOMProcessingInstruction) |
355
|
|
|
{ |
356
|
|
|
$node->data .= $uniqid; |
357
|
|
|
$html = str_replace($uniqid, '', $html); |
358
|
|
|
} |
359
|
|
|
|
360
|
|
|
return $html; |
361
|
|
|
} |
362
|
|
|
|
363
|
|
|
/** |
364
|
|
|
* Load a template as an xsl:template node |
365
|
|
|
* |
366
|
|
|
* Will attempt to load it as XML first, then as HTML as a fallback. Either way, an xsl:template |
367
|
|
|
* node is returned |
368
|
|
|
* |
369
|
|
|
* @param string $template |
370
|
|
|
* @return DOMDocument |
371
|
|
|
*/ |
372
|
|
|
public static function loadTemplate($template) |
373
|
|
|
{ |
374
|
|
|
$dom = self::loadTemplateAsXML($template); |
375
|
|
|
if ($dom) |
376
|
|
|
{ |
377
|
|
|
return $dom; |
378
|
|
|
} |
379
|
|
|
|
380
|
|
|
$dom = self::loadTemplateAsXML(self::fixEntities($template)); |
381
|
|
|
if ($dom) |
382
|
|
|
{ |
383
|
|
|
return $dom; |
384
|
|
|
} |
385
|
|
|
|
386
|
|
|
// If the template contains an XSL element, abort now. Otherwise, try reparsing it as HTML |
387
|
|
|
if (strpos($template, '<xsl:') !== false) |
388
|
|
|
{ |
389
|
|
|
$error = libxml_get_last_error(); |
390
|
|
|
|
391
|
|
|
throw new RuntimeException('Invalid XSL: ' . $error->message); |
392
|
|
|
} |
393
|
|
|
|
394
|
|
|
return self::loadTemplateAsHTML($template); |
395
|
|
|
} |
396
|
|
|
|
397
|
|
|
/** |
398
|
|
|
* Replace simple templates (in an array, in-place) with a common template |
399
|
|
|
* |
400
|
|
|
* In some situations, renderers can take advantage of multiple tags having the same template. In |
401
|
|
|
* any configuration, there's almost always a number of "simple" tags that are rendered as an |
402
|
|
|
* HTML element of the same name with no HTML attributes. For instance, the system tag "p" used |
403
|
|
|
* for paragraphs, "B" tags used for "b" HTML elements, etc... This method replaces those |
404
|
|
|
* templates with a common template that uses a dynamic element name based on the tag's name, |
405
|
|
|
* either its nodeName or localName depending on whether the tag is namespaced, and normalized to |
406
|
|
|
* lowercase using XPath's translate() function |
407
|
|
|
* |
408
|
|
|
* @param array<string> &$templates Associative array of [tagName => template] |
409
|
|
|
* @param integer $minCount |
410
|
|
|
* @return void |
411
|
|
|
*/ |
412
|
|
|
public static function replaceHomogeneousTemplates(array &$templates, $minCount = 3) |
413
|
|
|
{ |
414
|
|
|
$tagNames = []; |
415
|
|
|
|
416
|
|
|
// Prepare the XPath expression used for the element's name |
417
|
|
|
$expr = 'name()'; |
418
|
|
|
|
419
|
|
|
// Identify "simple" tags, whose template is one element of the same name. Their template |
420
|
|
|
// can be replaced with a dynamic template shared by all the simple tags |
421
|
|
|
foreach ($templates as $tagName => $template) |
422
|
|
|
{ |
423
|
|
|
// Generate the element name based on the tag's localName, lowercased |
424
|
|
|
$elName = strtolower(preg_replace('/^[^:]+:/', '', $tagName)); |
425
|
|
|
|
426
|
|
|
if ($template === '<' . $elName . '><xsl:apply-templates/></' . $elName . '>') |
427
|
|
|
{ |
428
|
|
|
$tagNames[] = $tagName; |
429
|
|
|
|
430
|
|
|
// Use local-name() if any of the tags are namespaced |
431
|
|
|
if (strpos($tagName, ':') !== false) |
432
|
|
|
{ |
433
|
|
|
$expr = 'local-name()'; |
434
|
|
|
} |
435
|
|
|
} |
436
|
|
|
} |
437
|
|
|
|
438
|
|
|
// We only bother replacing their template if there are at least $minCount simple tags. |
439
|
|
|
// Otherwise it only makes the stylesheet bigger |
440
|
|
|
if (count($tagNames) < $minCount) |
441
|
|
|
{ |
442
|
|
|
return; |
443
|
|
|
} |
444
|
|
|
|
445
|
|
|
// Generate a list of uppercase characters from the tags' names |
446
|
|
|
$chars = preg_replace('/[^A-Z]+/', '', count_chars(implode('', $tagNames), 3)); |
447
|
|
|
|
448
|
|
|
if (is_string($chars) && $chars !== '') |
449
|
|
|
{ |
450
|
|
|
$expr = 'translate(' . $expr . ",'" . $chars . "','" . strtolower($chars) . "')"; |
451
|
|
|
} |
452
|
|
|
|
453
|
|
|
// Prepare the common template |
454
|
|
|
$template = '<xsl:element name="{' . $expr . '}">' |
455
|
|
|
. '<xsl:apply-templates/>' |
456
|
|
|
. '</xsl:element>'; |
457
|
|
|
|
458
|
|
|
// Replace the templates |
459
|
|
|
foreach ($tagNames as $tagName) |
460
|
|
|
{ |
461
|
|
|
$templates[$tagName] = $template; |
462
|
|
|
} |
463
|
|
|
} |
464
|
|
|
|
465
|
|
|
/** |
466
|
|
|
* Replace parts of a template that match given regexp |
467
|
|
|
* |
468
|
|
|
* Treats attribute values as plain text. Replacements within XPath expression is unsupported. |
469
|
|
|
* The callback must return an array with two elements. The first must be either of 'expression', |
470
|
|
|
* 'literal' or 'passthrough', and the second element depends on the first. |
471
|
|
|
* |
472
|
|
|
* - 'expression' indicates that the replacement must be treated as an XPath expression such as |
473
|
|
|
* '@foo', which must be passed as the second element. |
474
|
|
|
* - 'literal' indicates a literal (plain text) replacement, passed as its second element. |
475
|
|
|
* - 'passthrough' indicates that the replacement should the tag's content. It works differently |
476
|
|
|
* whether it is inside an attribute's value or a text node. Within an attribute's value, the |
477
|
|
|
* replacement will be the text content of the tag. Within a text node, the replacement |
478
|
|
|
* becomes an <xsl:apply-templates/> node. |
479
|
|
|
* |
480
|
|
|
* @param string $template Original template |
481
|
|
|
* @param string $regexp Regexp for matching parts that need replacement |
482
|
|
|
* @param callback $fn Callback used to get the replacement |
483
|
|
|
* @return string Processed template |
484
|
|
|
*/ |
485
|
|
|
public static function replaceTokens($template, $regexp, $fn) |
486
|
|
|
{ |
487
|
|
|
$dom = self::loadTemplate($template); |
488
|
|
|
$xpath = new DOMXPath($dom); |
489
|
|
|
|
490
|
|
|
foreach ($xpath->query('//@*') as $attribute) |
491
|
|
|
{ |
492
|
|
|
self::replaceTokensInAttribute($attribute, $regexp, $fn); |
493
|
|
|
} |
494
|
|
|
foreach ($xpath->query('//text()') as $node) |
495
|
|
|
{ |
496
|
|
|
self::replaceTokensInText($node, $regexp, $fn); |
497
|
|
|
} |
498
|
|
|
|
499
|
|
|
return self::saveTemplate($dom); |
500
|
|
|
} |
501
|
|
|
|
502
|
|
|
/** |
503
|
|
|
* Replace parts of an attribute that match given regexp |
504
|
|
|
* |
505
|
|
|
* @param DOMAttr $attribute Attribute |
506
|
|
|
* @param string $regexp Regexp for matching parts that need replacement |
507
|
|
|
* @param callback $fn Callback used to get the replacement |
508
|
|
|
* @return void |
509
|
|
|
*/ |
510
|
|
|
protected static function replaceTokensInAttribute(DOMAttr $attribute, $regexp, $fn) |
511
|
|
|
{ |
512
|
|
|
$attrValue = preg_replace_callback( |
513
|
|
|
$regexp, |
514
|
|
|
function ($m) use ($fn, $attribute) |
515
|
|
|
{ |
516
|
|
|
$replacement = $fn($m, $attribute); |
517
|
|
|
if ($replacement[0] === 'expression') |
518
|
|
|
{ |
519
|
|
|
return '{' . $replacement[1] . '}'; |
520
|
|
|
} |
521
|
|
|
elseif ($replacement[0] === 'passthrough') |
522
|
|
|
{ |
523
|
|
|
return '{.}'; |
524
|
|
|
} |
525
|
|
|
else |
526
|
|
|
{ |
527
|
|
|
return $replacement[1]; |
528
|
|
|
} |
529
|
|
|
}, |
530
|
|
|
$attribute->value |
531
|
|
|
); |
532
|
|
|
$attribute->value = htmlspecialchars($attrValue, ENT_COMPAT, 'UTF-8'); |
533
|
|
|
} |
534
|
|
|
|
535
|
|
|
/** |
536
|
|
|
* Replace parts of a text node that match given regexp |
537
|
|
|
* |
538
|
|
|
* @param DOMText $node Text node |
539
|
|
|
* @param string $regexp Regexp for matching parts that need replacement |
540
|
|
|
* @param callback $fn Callback used to get the replacement |
541
|
|
|
* @return void |
542
|
|
|
*/ |
543
|
|
|
protected static function replaceTokensInText(DOMText $node, $regexp, $fn) |
544
|
|
|
{ |
545
|
|
|
// Grab the node's parent so that we can rebuild the text with added variables right |
546
|
|
|
// before the node, using DOM's insertBefore(). Technically, it would make more sense |
547
|
|
|
// to create a document fragment, append nodes then replace the node with the fragment |
548
|
|
|
// but it leads to namespace redeclarations, which looks ugly |
549
|
|
|
$parentNode = $node->parentNode; |
550
|
|
|
$dom = $node->ownerDocument; |
551
|
|
|
|
552
|
|
|
preg_match_all($regexp, $node->textContent, $matches, PREG_SET_ORDER | PREG_OFFSET_CAPTURE); |
553
|
|
|
$lastPos = 0; |
554
|
|
|
foreach ($matches as $m) |
555
|
|
|
{ |
556
|
|
|
$pos = $m[0][1]; |
557
|
|
|
|
558
|
|
|
// Catch-up to current position |
559
|
|
|
$text = substr($node->textContent, $lastPos, $pos - $lastPos); |
560
|
|
|
$parentNode->insertBefore($dom->createTextNode($text), $node); |
561
|
|
|
$lastPos = $pos + strlen($m[0][0]); |
562
|
|
|
|
563
|
|
|
// Get the replacement for this token |
564
|
|
|
$replacement = $fn(array_column($m, 0), $node); |
565
|
|
|
if ($replacement[0] === 'expression') |
566
|
|
|
{ |
567
|
|
|
$newNode = $dom->createElementNS(self::XMLNS_XSL, 'xsl:value-of'); |
568
|
|
|
$newNode->setAttribute('select', $replacement[1]); |
569
|
|
|
} |
570
|
|
|
elseif ($replacement[0] === 'passthrough') |
571
|
|
|
{ |
572
|
|
|
$newNode = $dom->createElementNS(self::XMLNS_XSL, 'xsl:apply-templates'); |
573
|
|
|
} |
574
|
|
|
else |
575
|
|
|
{ |
576
|
|
|
$newNode = $dom->createTextNode($replacement[1]); |
577
|
|
|
} |
578
|
|
|
$parentNode->insertBefore($newNode, $node); |
579
|
|
|
} |
580
|
|
|
|
581
|
|
|
// Append the rest of the text |
582
|
|
|
$text = substr($node->textContent, $lastPos); |
583
|
|
|
$parentNode->insertBefore($dom->createTextNode($text), $node); |
584
|
|
|
|
585
|
|
|
// Now remove the old text node |
586
|
|
|
$parentNode->removeChild($node); |
587
|
|
|
} |
588
|
|
|
|
589
|
|
|
/** |
590
|
|
|
* Serialize a loaded template back into a string |
591
|
|
|
* |
592
|
|
|
* NOTE: removes the root node created by loadTemplate() |
593
|
|
|
* |
594
|
|
|
* @param DOMDocument $dom |
595
|
|
|
* @return string |
596
|
|
|
*/ |
597
|
|
|
public static function saveTemplate(DOMDocument $dom) |
598
|
|
|
{ |
599
|
|
|
return self::innerXML($dom->documentElement); |
600
|
|
|
} |
601
|
|
|
|
602
|
|
|
/** |
603
|
|
|
* Replace HTML entities and unescaped ampersands in given template |
604
|
|
|
* |
605
|
|
|
* @param string $template |
606
|
|
|
* @return string |
607
|
|
|
*/ |
608
|
|
|
protected static function fixEntities($template) |
609
|
|
|
{ |
610
|
|
|
return preg_replace_callback( |
611
|
|
|
'(&(?!quot;|amp;|apos;|lt;|gt;)\\w+;)', |
612
|
|
|
function ($m) |
613
|
|
|
{ |
614
|
|
|
return html_entity_decode($m[0], ENT_NOQUOTES, 'UTF-8'); |
615
|
|
|
}, |
616
|
|
|
preg_replace('(&(?![A-Za-z0-9]+;|#\\d+;|#x[A-Fa-f0-9]+;))', '&', $template) |
617
|
|
|
); |
618
|
|
|
} |
619
|
|
|
|
620
|
|
|
/** |
621
|
|
|
* Get the XML content of an element |
622
|
|
|
* |
623
|
|
|
* @param DOMElement $element |
624
|
|
|
* @return string |
625
|
|
|
*/ |
626
|
|
|
protected static function innerXML(DOMElement $element) |
627
|
|
|
{ |
628
|
|
|
// Serialize the XML then remove the outer element |
629
|
|
|
$xml = $element->ownerDocument->saveXML($element); |
630
|
|
|
|
631
|
|
|
$pos = 1 + strpos($xml, '>'); |
632
|
|
|
$len = strrpos($xml, '<') - $pos; |
633
|
|
|
|
634
|
|
|
// If the template is empty, return an empty string |
635
|
|
|
if ($len < 1) |
636
|
|
|
{ |
637
|
|
|
return ''; |
638
|
|
|
} |
639
|
|
|
|
640
|
|
|
$xml = substr($xml, $pos, $len); |
641
|
|
|
|
642
|
|
|
return $xml; |
643
|
|
|
} |
644
|
|
|
|
645
|
|
|
/** |
646
|
|
|
* Load given HTML template in a DOM document |
647
|
|
|
* |
648
|
|
|
* @param string $template Original template |
649
|
|
|
* @return DOMDocument |
650
|
|
|
*/ |
651
|
|
|
protected static function loadTemplateAsHTML($template) |
652
|
|
|
{ |
653
|
|
|
$dom = new DOMDocument; |
654
|
|
|
$html = '<?xml version="1.0" encoding="utf-8" ?><html><body><div>' . $template . '</div></body></html>'; |
655
|
|
|
|
656
|
|
|
$useErrors = libxml_use_internal_errors(true); |
657
|
|
|
$dom->loadHTML($html); |
658
|
|
|
self::removeInvalidAttributes($dom); |
659
|
|
|
libxml_use_internal_errors($useErrors); |
660
|
|
|
|
661
|
|
|
// Now dump the thing as XML then reload it with the proper root element |
662
|
|
|
$xml = '<?xml version="1.0" encoding="utf-8" ?><xsl:template xmlns:xsl="' . self::XMLNS_XSL . '">' . self::innerXML($dom->documentElement->firstChild->firstChild) . '</xsl:template>'; |
663
|
|
|
|
664
|
|
|
$useErrors = libxml_use_internal_errors(true); |
665
|
|
|
$dom->loadXML($xml); |
666
|
|
|
libxml_use_internal_errors($useErrors); |
667
|
|
|
|
668
|
|
|
return $dom; |
669
|
|
|
} |
670
|
|
|
|
671
|
|
|
/** |
672
|
|
|
* Load given XSL template in a DOM document |
673
|
|
|
* |
674
|
|
|
* @param string $template Original template |
675
|
|
|
* @return bool|DOMDocument DOMDocument on success, FALSE otherwise |
676
|
|
|
*/ |
677
|
|
|
protected static function loadTemplateAsXML($template) |
678
|
|
|
{ |
679
|
|
|
$xml = '<?xml version="1.0" encoding="utf-8" ?><xsl:template xmlns:xsl="' . self::XMLNS_XSL . '">' . $template . '</xsl:template>'; |
680
|
|
|
|
681
|
|
|
$useErrors = libxml_use_internal_errors(true); |
682
|
|
|
$dom = new DOMDocument; |
683
|
|
|
$success = $dom->loadXML($xml); |
684
|
|
|
self::removeInvalidAttributes($dom); |
685
|
|
|
libxml_use_internal_errors($useErrors); |
686
|
|
|
|
687
|
|
|
return ($success) ? $dom : false; |
688
|
|
|
} |
689
|
|
|
|
690
|
|
|
/** |
691
|
|
|
* Remove attributes with an invalid name from given DOM document |
692
|
|
|
* |
693
|
|
|
* @param DOMDocument $dom |
694
|
|
|
* @return void |
695
|
|
|
*/ |
696
|
|
|
protected static function removeInvalidAttributes(DOMDocument $dom) |
697
|
|
|
{ |
698
|
|
|
$xpath = new DOMXPath($dom); |
699
|
|
|
foreach ($xpath->query('//@*') as $attribute) |
700
|
|
|
{ |
701
|
|
|
if (!preg_match('(^(?:[-\\w]+:)?(?!\\d)[-\\w]+$)D', $attribute->nodeName)) |
702
|
|
|
{ |
703
|
|
|
$attribute->parentNode->removeAttributeNode($attribute); |
704
|
|
|
} |
705
|
|
|
} |
706
|
|
|
} |
707
|
|
|
} |