|
1
|
|
|
<?php |
|
2
|
|
|
/** @file |
|
3
|
|
|
* Traverse a DOM. |
|
4
|
|
|
*/ |
|
5
|
|
|
|
|
6
|
|
|
namespace QueryPath\CSS; |
|
7
|
|
|
|
|
8
|
|
|
use \QueryPath\CSS\DOMTraverser\Util; |
|
9
|
|
|
use \QueryPath\CSS\DOMTraverser\PseudoClass; |
|
10
|
|
|
use \QueryPath\CSS\DOMTraverser\PseudoElement; |
|
|
|
|
|
|
11
|
|
|
use QueryPath\CSS\SimpleSelector; |
|
12
|
|
|
use SplObjectStorage; |
|
13
|
|
|
|
|
14
|
|
|
/** |
|
15
|
|
|
* Traverse a DOM, finding matches to the selector. |
|
16
|
|
|
* |
|
17
|
|
|
* This traverses a DOMDocument and attempts to find |
|
18
|
|
|
* matches to the provided selector. |
|
19
|
|
|
* |
|
20
|
|
|
* \b How this works |
|
21
|
|
|
* |
|
22
|
|
|
* This performs a bottom-up search. On the first pass, |
|
23
|
|
|
* it attempts to find all of the matching elements for the |
|
24
|
|
|
* last simple selector in a selector. |
|
25
|
|
|
* |
|
26
|
|
|
* Subsequent passes attempt to eliminate matches from the |
|
27
|
|
|
* initial matching set. |
|
28
|
|
|
* |
|
29
|
|
|
* Example: |
|
30
|
|
|
* |
|
31
|
|
|
* Say we begin with the selector `foo.bar baz`. This is processed |
|
32
|
|
|
* as follows: |
|
33
|
|
|
* |
|
34
|
|
|
* - First, find all baz elements. |
|
35
|
|
|
* - Next, for any baz element that does not have foo as an ancestor, |
|
36
|
|
|
* eliminate it from the matches. |
|
37
|
|
|
* - Finally, for those that have foo as an ancestor, does that foo |
|
38
|
|
|
* also have a class baz? If not, it is removed from the matches. |
|
39
|
|
|
* |
|
40
|
|
|
* \b Extrapolation |
|
41
|
|
|
* |
|
42
|
|
|
* Partial simple selectors are almost always expanded to include an |
|
43
|
|
|
* element. |
|
44
|
|
|
* |
|
45
|
|
|
* Examples: |
|
46
|
|
|
* |
|
47
|
|
|
* - `:first` is expanded to `*:first` |
|
48
|
|
|
* - `.bar` is expanded to `*.bar`. |
|
49
|
|
|
* - `.outer .inner` is expanded to `*.outer *.inner` |
|
50
|
|
|
* |
|
51
|
|
|
* The exception is that IDs are sometimes not expanded, e.g.: |
|
52
|
|
|
* |
|
53
|
|
|
* - `#myElement` does not get expanded |
|
54
|
|
|
* - `#myElement .class` \i may be expanded to `*#myElement *.class` |
|
55
|
|
|
* (which will obviously not perform well). |
|
56
|
|
|
*/ |
|
57
|
|
|
class DOMTraverser implements Traverser |
|
58
|
|
|
{ |
|
59
|
|
|
protected $matches = []; |
|
60
|
|
|
protected $selector; |
|
61
|
|
|
protected $dom; |
|
62
|
|
|
protected $initialized = true; |
|
63
|
|
|
protected $psHandler; |
|
64
|
|
|
protected $scopeNode; |
|
65
|
|
|
|
|
66
|
|
|
/** |
|
67
|
|
|
* Build a new DOMTraverser. |
|
68
|
|
|
* |
|
69
|
|
|
* This requires a DOM-like object or collection of DOM nodes. |
|
70
|
|
|
* |
|
71
|
|
|
* @param \SPLObjectStorage $splos |
|
72
|
|
|
* @param bool $initialized |
|
73
|
|
|
* @param null $scopeNode |
|
|
|
|
|
|
74
|
|
|
*/ |
|
75
|
|
|
public function __construct(\SPLObjectStorage $splos, bool $initialized = false, $scopeNode = NULL) |
|
76
|
|
|
{ |
|
77
|
|
|
$this->psHandler = new PseudoClass(); |
|
78
|
|
|
$this->initialized = $initialized; |
|
79
|
|
|
|
|
80
|
|
|
// Re-use the initial splos |
|
81
|
|
|
$this->matches = $splos; |
|
82
|
|
|
|
|
83
|
|
|
if (count($splos) !== 0) { |
|
84
|
|
|
$splos->rewind(); |
|
85
|
|
|
$first = $splos->current(); |
|
86
|
|
|
if ($first instanceof \DOMDocument) { |
|
87
|
|
|
$this->dom = $first;//->documentElement; |
|
88
|
|
|
} else { |
|
89
|
|
|
$this->dom = $first->ownerDocument;//->documentElement; |
|
90
|
|
|
} |
|
91
|
|
|
|
|
92
|
|
|
$this->scopeNode = $scopeNode; |
|
93
|
|
|
if (empty($scopeNode)) { |
|
94
|
|
|
$this->scopeNode = $this->dom->documentElement; |
|
95
|
|
|
} |
|
96
|
|
|
} |
|
97
|
|
|
|
|
98
|
|
|
// This assumes a DOM. Need to also accomodate the case |
|
99
|
|
|
// where we get a set of elements. |
|
100
|
|
|
/* |
|
101
|
|
|
$this->dom = $dom; |
|
102
|
|
|
$this->matches = new \SplObjectStorage(); |
|
103
|
|
|
$this->matches->attach($this->dom); |
|
104
|
|
|
*/ |
|
105
|
|
|
} |
|
106
|
|
|
|
|
107
|
|
|
public function debug($msg) |
|
108
|
|
|
{ |
|
109
|
|
|
fwrite(STDOUT, PHP_EOL . $msg); |
|
110
|
|
|
} |
|
111
|
|
|
|
|
112
|
|
|
/** |
|
113
|
|
|
* Given a selector, find the matches in the given DOM. |
|
114
|
|
|
* |
|
115
|
|
|
* This is the main function for querying the DOM using a CSS |
|
116
|
|
|
* selector. |
|
117
|
|
|
* |
|
118
|
|
|
* @param string $selector |
|
119
|
|
|
* The selector. |
|
120
|
|
|
* @return DOMTraverser a list of matched |
|
121
|
|
|
* DOMNode objects. |
|
122
|
|
|
* @throws ParseException |
|
123
|
|
|
*/ |
|
124
|
|
|
public function find($selector) : DOMTraverser |
|
125
|
|
|
{ |
|
126
|
|
|
// Setup |
|
127
|
|
|
$handler = new Selector(); |
|
128
|
|
|
$parser = new Parser($selector, $handler); |
|
129
|
|
|
$parser->parse(); |
|
130
|
|
|
$this->selector = $handler; |
|
131
|
|
|
|
|
132
|
|
|
//$selector = $handler->toArray(); |
|
133
|
|
|
$found = $this->newMatches(); |
|
134
|
|
|
foreach ($handler as $selectorGroup) { |
|
135
|
|
|
// Initialize matches if necessary. |
|
136
|
|
|
if ($this->initialized) { |
|
137
|
|
|
$candidates = $this->matches; |
|
138
|
|
|
} else { |
|
139
|
|
|
$candidates = $this->initialMatch($selectorGroup[0], $this->matches); |
|
140
|
|
|
} |
|
141
|
|
|
|
|
142
|
|
|
/** @var \DOMElement $candidate */ |
|
143
|
|
|
foreach ($candidates as $candidate) { |
|
144
|
|
|
// fprintf(STDOUT, "Testing %s against %s.\n", $candidate->tagName, $selectorGroup[0]); |
|
145
|
|
|
if ($this->matchesSelector($candidate, $selectorGroup)) { |
|
146
|
|
|
// $this->debug('Attaching ' . $candidate->nodeName); |
|
147
|
|
|
$found->attach($candidate); |
|
148
|
|
|
} |
|
149
|
|
|
} |
|
150
|
|
|
} |
|
151
|
|
|
$this->setMatches($found); |
|
152
|
|
|
|
|
153
|
|
|
return $this; |
|
|
|
|
|
|
154
|
|
|
} |
|
155
|
|
|
|
|
156
|
|
|
public function matches() |
|
157
|
|
|
{ |
|
158
|
|
|
return $this->matches; |
|
|
|
|
|
|
159
|
|
|
} |
|
160
|
|
|
|
|
161
|
|
|
/** |
|
162
|
|
|
* Check whether the given node matches the given selector. |
|
163
|
|
|
* |
|
164
|
|
|
* A selector is a group of one or more simple selectors combined |
|
165
|
|
|
* by combinators. This determines if a given selector |
|
166
|
|
|
* matches the given node. |
|
167
|
|
|
* |
|
168
|
|
|
* @attention |
|
169
|
|
|
* Evaluation of selectors is done recursively. Thus the length |
|
170
|
|
|
* of the selector is limited to the recursion depth allowed by |
|
171
|
|
|
* the PHP configuration. This should only cause problems for |
|
172
|
|
|
* absolutely huge selectors or for versions of PHP tuned to |
|
173
|
|
|
* strictly limit recursion depth. |
|
174
|
|
|
* |
|
175
|
|
|
* @param \DOMElement $node |
|
176
|
|
|
* The DOMNode to check. |
|
177
|
|
|
* @param $selector |
|
178
|
|
|
* @return boolean |
|
179
|
|
|
* A boolean TRUE if the node matches, false otherwise. |
|
180
|
|
|
*/ |
|
181
|
|
|
public function matchesSelector(\DOMElement $node, $selector) |
|
182
|
|
|
{ |
|
183
|
|
|
return $this->matchesSimpleSelector($node, $selector, 0); |
|
184
|
|
|
} |
|
185
|
|
|
|
|
186
|
|
|
/** |
|
187
|
|
|
* Performs a match check on a SimpleSelector. |
|
188
|
|
|
* |
|
189
|
|
|
* Where matchesSelector() does a check on an entire selector, |
|
190
|
|
|
* this checks only a simple selector (plus an optional |
|
191
|
|
|
* combinator). |
|
192
|
|
|
* |
|
193
|
|
|
* @param \DOMElement $node |
|
194
|
|
|
* @param $selectors |
|
195
|
|
|
* @param $index |
|
196
|
|
|
* @return boolean |
|
197
|
|
|
* A boolean TRUE if the node matches, false otherwise. |
|
198
|
|
|
* @throws NotImplementedException |
|
199
|
|
|
*/ |
|
200
|
|
|
public function matchesSimpleSelector(\DOMElement $node, $selectors, $index) |
|
201
|
|
|
{ |
|
202
|
|
|
$selector = $selectors[$index]; |
|
203
|
|
|
// Note that this will short circuit as soon as one of these |
|
204
|
|
|
// returns FALSE. |
|
205
|
|
|
$result = $this->matchElement($node, $selector->element, $selector->ns) |
|
206
|
|
|
&& $this->matchAttributes($node, $selector->attributes) |
|
207
|
|
|
&& $this->matchId($node, $selector->id) |
|
208
|
|
|
&& $this->matchClasses($node, $selector->classes) |
|
209
|
|
|
&& $this->matchPseudoClasses($node, $selector->pseudoClasses) |
|
210
|
|
|
&& $this->matchPseudoElements($node, $selector->pseudoElements); |
|
211
|
|
|
|
|
212
|
|
|
$isNextRule = isset($selectors[++$index]); |
|
213
|
|
|
// If there is another selector, we process that if there a match |
|
214
|
|
|
// hasn't been found. |
|
215
|
|
|
/* |
|
216
|
|
|
if ($isNextRule && $selectors[$index]->combinator == SimpleSelector::anotherSelector) { |
|
217
|
|
|
// We may need to re-initialize the match set for the next selector. |
|
218
|
|
|
if (!$this->initialized) { |
|
219
|
|
|
$this->initialMatch($selectors[$index]); |
|
220
|
|
|
} |
|
221
|
|
|
if (!$result) fprintf(STDOUT, "Element: %s, Next selector: %s\n", $node->tagName, $selectors[$index]); |
|
222
|
|
|
return $result || $this->matchesSimpleSelector($node, $selectors, $index); |
|
223
|
|
|
} |
|
224
|
|
|
// If we have a match and we have a combinator, we need to |
|
225
|
|
|
// recurse up the tree. |
|
226
|
|
|
else*/ |
|
227
|
|
|
if ($isNextRule && $result) { |
|
228
|
|
|
$result = $this->combine($node, $selectors, $index); |
|
229
|
|
|
} |
|
230
|
|
|
|
|
231
|
|
|
return $result; |
|
232
|
|
|
} |
|
233
|
|
|
|
|
234
|
|
|
/** |
|
235
|
|
|
* Combine the next selector with the given match |
|
236
|
|
|
* using the next combinator. |
|
237
|
|
|
* |
|
238
|
|
|
* If the next selector is combined with another |
|
239
|
|
|
* selector, that will be evaluated too, and so on. |
|
240
|
|
|
* So if this function returns TRUE, it means that all |
|
241
|
|
|
* child selectors are also matches. |
|
242
|
|
|
* |
|
243
|
|
|
* @param DOMNode $node |
|
|
|
|
|
|
244
|
|
|
* The DOMNode to test. |
|
245
|
|
|
* @param array $selectors |
|
246
|
|
|
* The array of simple selectors. |
|
247
|
|
|
* @param int $index |
|
248
|
|
|
* The index of the current selector. |
|
249
|
|
|
* @return boolean |
|
250
|
|
|
* TRUE if the next selector(s) match. |
|
251
|
|
|
*/ |
|
252
|
|
|
public function combine(\DOMElement $node, $selectors, $index) |
|
253
|
|
|
{ |
|
254
|
|
|
$selector = $selectors[$index]; |
|
255
|
|
|
//$this->debug(implode(' ', $selectors)); |
|
256
|
|
|
switch ($selector->combinator) { |
|
257
|
|
|
case SimpleSelector::ADJACENT: |
|
258
|
|
|
return $this->combineAdjacent($node, $selectors, $index); |
|
259
|
|
|
case SimpleSelector::SIBLING: |
|
260
|
|
|
return $this->combineSibling($node, $selectors, $index); |
|
261
|
|
|
case SimpleSelector::DIRECT_DESCENDANT: |
|
262
|
|
|
return $this->combineDirectDescendant($node, $selectors, $index); |
|
263
|
|
|
case SimpleSelector::ANY_DESCENDANT: |
|
264
|
|
|
return $this->combineAnyDescendant($node, $selectors, $index); |
|
265
|
|
|
case SimpleSelector::ANOTHER_SELECTOR: |
|
266
|
|
|
// fprintf(STDOUT, "Next selector: %s\n", $selectors[$index]); |
|
267
|
|
|
return $this->matchesSimpleSelector($node, $selectors, $index);; |
|
268
|
|
|
} |
|
269
|
|
|
|
|
270
|
|
|
return false; |
|
271
|
|
|
} |
|
272
|
|
|
|
|
273
|
|
|
/** |
|
274
|
|
|
* Process an Adjacent Sibling. |
|
275
|
|
|
* |
|
276
|
|
|
* The spec does not indicate whether Adjacent should ignore non-Element |
|
277
|
|
|
* nodes, so we choose to ignore them. |
|
278
|
|
|
* |
|
279
|
|
|
* @param DOMNode $node |
|
280
|
|
|
* A DOM Node. |
|
281
|
|
|
* @param array $selectors |
|
282
|
|
|
* The selectors array. |
|
283
|
|
|
* @param int $index |
|
284
|
|
|
* The current index to the operative simple selector in the selectors |
|
285
|
|
|
* array. |
|
286
|
|
|
* @return boolean |
|
287
|
|
|
* TRUE if the combination matches, FALSE otherwise. |
|
288
|
|
|
*/ |
|
289
|
|
|
public function combineAdjacent($node, $selectors, $index) |
|
290
|
|
|
{ |
|
291
|
|
|
while (!empty($node->previousSibling)) { |
|
292
|
|
|
$node = $node->previousSibling; |
|
293
|
|
|
if ($node->nodeType == XML_ELEMENT_NODE) { |
|
294
|
|
|
//$this->debug(sprintf('Testing %s against "%s"', $node->tagName, $selectors[$index])); |
|
295
|
|
|
return $this->matchesSimpleSelector($node, $selectors, $index); |
|
296
|
|
|
} |
|
297
|
|
|
} |
|
298
|
|
|
|
|
299
|
|
|
return false; |
|
300
|
|
|
} |
|
301
|
|
|
|
|
302
|
|
|
/** |
|
303
|
|
|
* Check all siblings. |
|
304
|
|
|
* |
|
305
|
|
|
* According to the spec, this only tests elements LEFT of the provided |
|
306
|
|
|
* node. |
|
307
|
|
|
* |
|
308
|
|
|
* @param DOMNode $node |
|
309
|
|
|
* A DOM Node. |
|
310
|
|
|
* @param array $selectors |
|
311
|
|
|
* The selectors array. |
|
312
|
|
|
* @param int $index |
|
313
|
|
|
* The current index to the operative simple selector in the selectors |
|
314
|
|
|
* array. |
|
315
|
|
|
* @return boolean |
|
316
|
|
|
* TRUE if the combination matches, FALSE otherwise. |
|
317
|
|
|
*/ |
|
318
|
|
|
public function combineSibling($node, $selectors, $index) |
|
319
|
|
|
{ |
|
320
|
|
|
while (!empty($node->previousSibling)) { |
|
321
|
|
|
$node = $node->previousSibling; |
|
322
|
|
|
if ($node->nodeType == XML_ELEMENT_NODE && $this->matchesSimpleSelector($node, $selectors, $index)) { |
|
323
|
|
|
return true; |
|
324
|
|
|
} |
|
325
|
|
|
} |
|
326
|
|
|
|
|
327
|
|
|
return false; |
|
328
|
|
|
} |
|
329
|
|
|
|
|
330
|
|
|
/** |
|
331
|
|
|
* Handle a Direct Descendant combination. |
|
332
|
|
|
* |
|
333
|
|
|
* Check whether the given node is a rightly-related descendant |
|
334
|
|
|
* of its parent node. |
|
335
|
|
|
* |
|
336
|
|
|
* @param DOMNode $node |
|
337
|
|
|
* A DOM Node. |
|
338
|
|
|
* @param array $selectors |
|
339
|
|
|
* The selectors array. |
|
340
|
|
|
* @param int $index |
|
341
|
|
|
* The current index to the operative simple selector in the selectors |
|
342
|
|
|
* array. |
|
343
|
|
|
* @return boolean |
|
344
|
|
|
* TRUE if the combination matches, FALSE otherwise. |
|
345
|
|
|
*/ |
|
346
|
|
|
public function combineDirectDescendant($node, $selectors, $index) |
|
347
|
|
|
{ |
|
348
|
|
|
$parent = $node->parentNode; |
|
349
|
|
|
if (empty($parent)) { |
|
350
|
|
|
return false; |
|
351
|
|
|
} |
|
352
|
|
|
|
|
353
|
|
|
return $this->matchesSimpleSelector($parent, $selectors, $index); |
|
354
|
|
|
} |
|
355
|
|
|
|
|
356
|
|
|
/** |
|
357
|
|
|
* Handle Any Descendant combinations. |
|
358
|
|
|
* |
|
359
|
|
|
* This checks to see if there are any matching routes from the |
|
360
|
|
|
* selector beginning at the present node. |
|
361
|
|
|
* |
|
362
|
|
|
* @param DOMNode $node |
|
363
|
|
|
* A DOM Node. |
|
364
|
|
|
* @param array $selectors |
|
365
|
|
|
* The selectors array. |
|
366
|
|
|
* @param int $index |
|
367
|
|
|
* The current index to the operative simple selector in the selectors |
|
368
|
|
|
* array. |
|
369
|
|
|
* @return boolean |
|
370
|
|
|
* TRUE if the combination matches, FALSE otherwise. |
|
371
|
|
|
*/ |
|
372
|
|
|
public function combineAnyDescendant($node, $selectors, $index) |
|
373
|
|
|
{ |
|
374
|
|
|
while (!empty($node->parentNode)) { |
|
375
|
|
|
$node = $node->parentNode; |
|
376
|
|
|
|
|
377
|
|
|
// Catch case where element is child of something |
|
378
|
|
|
// else. This should really only happen with a |
|
379
|
|
|
// document element. |
|
380
|
|
|
if ($node->nodeType != XML_ELEMENT_NODE) { |
|
381
|
|
|
continue; |
|
382
|
|
|
} |
|
383
|
|
|
|
|
384
|
|
|
if ($this->matchesSimpleSelector($node, $selectors, $index)) { |
|
385
|
|
|
return true; |
|
386
|
|
|
} |
|
387
|
|
|
} |
|
388
|
|
|
} |
|
389
|
|
|
|
|
390
|
|
|
/** |
|
391
|
|
|
* Get the intial match set. |
|
392
|
|
|
* |
|
393
|
|
|
* This should only be executed when not working with |
|
394
|
|
|
* an existing match set. |
|
395
|
|
|
* @param \QueryPath\CSS\SimpleSelector $selector |
|
396
|
|
|
* @param SplObjectStorage $matches |
|
397
|
|
|
* @return SplObjectStorage |
|
398
|
|
|
*/ |
|
399
|
|
|
protected function initialMatch(SimpleSelector $selector, SplObjectStorage $matches) : SplObjectStorage |
|
400
|
|
|
{ |
|
401
|
|
|
$element = $selector->element; |
|
402
|
|
|
|
|
403
|
|
|
// If no element is specified, we have to start with the |
|
404
|
|
|
// entire document. |
|
405
|
|
|
if ($element === NULL) { |
|
406
|
|
|
$element = '*'; |
|
407
|
|
|
} |
|
408
|
|
|
|
|
409
|
|
|
// We try to do some optimization here to reduce the |
|
410
|
|
|
// number of matches to the bare minimum. This will |
|
411
|
|
|
// reduce the subsequent number of operations that |
|
412
|
|
|
// must be performed in the query. |
|
413
|
|
|
|
|
414
|
|
|
// Experimental: ID queries use XPath to match, since |
|
415
|
|
|
// this should give us only a single matched element |
|
416
|
|
|
// to work with. |
|
417
|
|
|
if (/*$element == '*' &&*/ |
|
418
|
|
|
!empty($selector->id)) { |
|
419
|
|
|
$initialMatches = $this->initialMatchOnID($selector, $matches); |
|
420
|
|
|
} // If a namespace is set, find the namespace matches. |
|
421
|
|
|
elseif (!empty($selector->ns)) { |
|
422
|
|
|
$initialMatches = $this->initialMatchOnElementNS($selector, $matches); |
|
423
|
|
|
} |
|
424
|
|
|
// If the element is a wildcard, using class can |
|
425
|
|
|
// substantially reduce the number of elements that |
|
426
|
|
|
// we start with. |
|
427
|
|
|
elseif ($element === '*' && !empty($selector->classes)) { |
|
428
|
|
|
$initialMatches = $this->initialMatchOnClasses($selector, $matches); |
|
429
|
|
|
} else { |
|
430
|
|
|
$initialMatches = $this->initialMatchOnElement($selector, $matches); |
|
431
|
|
|
} |
|
432
|
|
|
|
|
433
|
|
|
return $initialMatches; |
|
434
|
|
|
} |
|
435
|
|
|
|
|
436
|
|
|
/** |
|
437
|
|
|
* Shortcut for finding initial match by ID. |
|
438
|
|
|
* |
|
439
|
|
|
* If the element is set to '*' and an ID is |
|
440
|
|
|
* set, then this should be used to find by ID, |
|
441
|
|
|
* which will drastically reduce the amount of |
|
442
|
|
|
* comparison operations done in PHP. |
|
443
|
|
|
* @param \QueryPath\CSS\SimpleSelector $selector |
|
444
|
|
|
* @param SplObjectStorage $matches |
|
445
|
|
|
* @return SplObjectStorage |
|
446
|
|
|
*/ |
|
447
|
|
|
protected function initialMatchOnID(SimpleSelector $selector, SplObjectStorage $matches) : SplObjectStorage |
|
448
|
|
|
{ |
|
449
|
|
|
$id = $selector->id; |
|
450
|
|
|
$found = $this->newMatches(); |
|
451
|
|
|
|
|
452
|
|
|
// Issue #145: DOMXPath will through an exception if the DOM is |
|
453
|
|
|
// not set. |
|
454
|
|
|
if (!($this->dom instanceof \DOMDocument)) { |
|
455
|
|
|
return $found; |
|
456
|
|
|
} |
|
457
|
|
|
$baseQuery = ".//*[@id='{$id}']"; |
|
458
|
|
|
$xpath = new \DOMXPath($this->dom); |
|
459
|
|
|
|
|
460
|
|
|
// Now we try to find any matching IDs. |
|
461
|
|
|
/** @var \DOMElement $node */ |
|
462
|
|
|
foreach ($matches as $node) { |
|
463
|
|
|
if ($node->getAttribute('id') === $id) { |
|
464
|
|
|
$found->attach($node); |
|
465
|
|
|
} |
|
466
|
|
|
|
|
467
|
|
|
$nl = $this->initialXpathQuery($xpath, $node, $baseQuery); |
|
468
|
|
|
if (!empty($nl) && $nl instanceof \DOMNodeList) { |
|
469
|
|
|
$this->attachNodeList($nl, $found); |
|
470
|
|
|
} |
|
471
|
|
|
} |
|
472
|
|
|
// Unset the ID selector. |
|
473
|
|
|
$selector->id = NULL; |
|
474
|
|
|
|
|
475
|
|
|
return $found; |
|
476
|
|
|
} |
|
477
|
|
|
|
|
478
|
|
|
/** |
|
479
|
|
|
* Shortcut for setting the intial match. |
|
480
|
|
|
* |
|
481
|
|
|
* This shortcut should only be used when the initial |
|
482
|
|
|
* element is '*' and there are classes set. |
|
483
|
|
|
* |
|
484
|
|
|
* In any other case, the element finding algo is |
|
485
|
|
|
* faster and should be used instead. |
|
486
|
|
|
* @param \QueryPath\CSS\SimpleSelector $selector |
|
487
|
|
|
* @param $matches |
|
488
|
|
|
* @return \SplObjectStorage |
|
489
|
|
|
*/ |
|
490
|
|
|
protected function initialMatchOnClasses(SimpleSelector $selector, SplObjectStorage $matches) : \SplObjectStorage |
|
491
|
|
|
{ |
|
492
|
|
|
$found = $this->newMatches(); |
|
493
|
|
|
|
|
494
|
|
|
// Issue #145: DOMXPath will through an exception if the DOM is |
|
495
|
|
|
// not set. |
|
496
|
|
|
if (!($this->dom instanceof \DOMDocument)) { |
|
497
|
|
|
return $found; |
|
498
|
|
|
} |
|
499
|
|
|
$baseQuery = './/*[@class]'; |
|
500
|
|
|
$xpath = new \DOMXPath($this->dom); |
|
501
|
|
|
|
|
502
|
|
|
// Now we try to find any matching IDs. |
|
503
|
|
|
/** @var \DOMElement $node */ |
|
504
|
|
|
foreach ($matches as $node) { |
|
505
|
|
|
// Refactor me! |
|
506
|
|
|
if ($node->hasAttribute('class')) { |
|
507
|
|
|
$intersect = array_intersect($selector->classes, explode(' ', $node->getAttribute('class'))); |
|
508
|
|
|
if (count($intersect) === count($selector->classes)) { |
|
509
|
|
|
$found->attach($node); |
|
510
|
|
|
} |
|
511
|
|
|
} |
|
512
|
|
|
|
|
513
|
|
|
$nl = $this->initialXpathQuery($xpath, $node, $baseQuery); |
|
514
|
|
|
/** @var \DOMElement $subNode */ |
|
515
|
|
|
foreach ($nl as $subNode) { |
|
516
|
|
|
$classes = $subNode->getAttribute('class'); |
|
517
|
|
|
$classArray = explode(' ', $classes); |
|
518
|
|
|
|
|
519
|
|
|
$intersect = array_intersect($selector->classes, $classArray); |
|
520
|
|
|
if (count($intersect) === count($selector->classes)) { |
|
521
|
|
|
$found->attach($subNode); |
|
522
|
|
|
} |
|
523
|
|
|
} |
|
524
|
|
|
} |
|
525
|
|
|
|
|
526
|
|
|
// Unset the classes selector. |
|
527
|
|
|
$selector->classes = []; |
|
528
|
|
|
|
|
529
|
|
|
return $found; |
|
530
|
|
|
} |
|
531
|
|
|
|
|
532
|
|
|
/** |
|
533
|
|
|
* Internal xpath query. |
|
534
|
|
|
* |
|
535
|
|
|
* This is optimized for very specific use, and is not a general |
|
536
|
|
|
* purpose function. |
|
537
|
|
|
* @param \DOMXPath $xpath |
|
538
|
|
|
* @param \DOMElement $node |
|
539
|
|
|
* @param string $query |
|
540
|
|
|
* @return \DOMNodeList |
|
541
|
|
|
*/ |
|
542
|
|
|
private function initialXpathQuery(\DOMXPath $xpath, \DOMElement $node, string $query) : \DOMNodeList |
|
543
|
|
|
{ |
|
544
|
|
|
// This works around a bug in which the document element |
|
545
|
|
|
// does not correctly search with the $baseQuery. |
|
546
|
|
|
if ($node->isSameNode($this->dom->documentElement)) { |
|
547
|
|
|
$query = mb_substr($query, 1); |
|
548
|
|
|
} |
|
549
|
|
|
|
|
550
|
|
|
return $xpath->query($query, $node); |
|
551
|
|
|
} |
|
552
|
|
|
|
|
553
|
|
|
/** |
|
554
|
|
|
* Shortcut for setting the initial match. |
|
555
|
|
|
* |
|
556
|
|
|
* @param $selector |
|
557
|
|
|
* @param $matches |
|
558
|
|
|
* @return \SplObjectStorage |
|
559
|
|
|
*/ |
|
560
|
|
|
protected function initialMatchOnElement(SimpleSelector $selector, SplObjectStorage $matches) : SplObjectStorage |
|
561
|
|
|
{ |
|
562
|
|
|
$element = $selector->element; |
|
563
|
|
|
if (NULL === $element) { |
|
564
|
|
|
$element = '*'; |
|
565
|
|
|
} |
|
566
|
|
|
$found = $this->newMatches(); |
|
567
|
|
|
/** @var \DOMDocument $node */ |
|
568
|
|
|
foreach ($matches as $node) { |
|
569
|
|
|
// Capture the case where the initial element is the root element. |
|
570
|
|
|
if ($node->tagName === $element |
|
571
|
|
|
|| ($element === '*' && $node->parentNode instanceof \DOMDocument)) { |
|
572
|
|
|
$found->attach($node); |
|
573
|
|
|
} |
|
574
|
|
|
$nl = $node->getElementsByTagName($element); |
|
575
|
|
|
if (!empty($nl) && $nl instanceof \DOMNodeList) { |
|
576
|
|
|
$this->attachNodeList($nl, $found); |
|
577
|
|
|
} |
|
578
|
|
|
} |
|
579
|
|
|
|
|
580
|
|
|
$selector->element = NULL; |
|
581
|
|
|
|
|
582
|
|
|
return $found; |
|
583
|
|
|
} |
|
584
|
|
|
|
|
585
|
|
|
/** |
|
586
|
|
|
* Get elements and filter by namespace. |
|
587
|
|
|
* @param \QueryPath\CSS\SimpleSelector $selector |
|
588
|
|
|
* @param SplObjectStorage $matches |
|
589
|
|
|
* @return SplObjectStorage |
|
590
|
|
|
*/ |
|
591
|
|
|
protected function initialMatchOnElementNS(SimpleSelector $selector, SplObjectStorage $matches) : SplObjectStorage |
|
592
|
|
|
{ |
|
593
|
|
|
$ns = $selector->ns; |
|
594
|
|
|
|
|
595
|
|
|
$elements = $this->initialMatchOnElement($selector, $matches); |
|
596
|
|
|
|
|
597
|
|
|
// "any namespace" matches anything. |
|
598
|
|
|
if ($ns === '*') { |
|
599
|
|
|
return $elements; |
|
600
|
|
|
} |
|
601
|
|
|
|
|
602
|
|
|
// Loop through and make a list of items that need to be filtered |
|
603
|
|
|
// out, then filter them. This is required b/c ObjectStorage iterates |
|
604
|
|
|
// wrongly when an item is detached in an access loop. |
|
605
|
|
|
$detach = []; |
|
606
|
|
|
foreach ($elements as $node) { |
|
607
|
|
|
// This lookup must be done PER NODE. |
|
608
|
|
|
$nsuri = $node->lookupNamespaceURI($ns); |
|
609
|
|
|
if (empty($nsuri) || $node->namespaceURI !== $nsuri) { |
|
610
|
|
|
$detach[] = $node; |
|
611
|
|
|
} |
|
612
|
|
|
} |
|
613
|
|
|
foreach ($detach as $rem) { |
|
614
|
|
|
$elements->detach($rem); |
|
615
|
|
|
} |
|
616
|
|
|
$selector->ns = NULL; |
|
617
|
|
|
|
|
618
|
|
|
return $elements; |
|
619
|
|
|
} |
|
620
|
|
|
|
|
621
|
|
|
/** |
|
622
|
|
|
* Checks to see if the DOMNode matches the given element selector. |
|
623
|
|
|
* |
|
624
|
|
|
* This handles the following cases: |
|
625
|
|
|
* |
|
626
|
|
|
* - element (foo) |
|
627
|
|
|
* - namespaced element (ns|foo) |
|
628
|
|
|
* - namespaced wildcard (ns|*) |
|
629
|
|
|
* - wildcard (* or *|*) |
|
630
|
|
|
* @param \DOMElement $node |
|
631
|
|
|
* @param $element |
|
632
|
|
|
* @param null $ns |
|
|
|
|
|
|
633
|
|
|
* @return bool |
|
634
|
|
|
*/ |
|
635
|
|
|
protected function matchElement(\DOMElement $node, $element, $ns = NULL) : bool |
|
636
|
|
|
{ |
|
637
|
|
|
if (empty($element)) { |
|
638
|
|
|
return true; |
|
639
|
|
|
} |
|
640
|
|
|
|
|
641
|
|
|
// Handle namespace. |
|
642
|
|
|
if (!empty($ns) && $ns !== '*') { |
|
643
|
|
|
// Check whether we have a matching NS URI. |
|
644
|
|
|
$nsuri = $node->lookupNamespaceURI($ns); |
|
645
|
|
|
if (empty($nsuri) || $node->namespaceURI !== $nsuri) { |
|
646
|
|
|
return false; |
|
647
|
|
|
} |
|
648
|
|
|
} |
|
649
|
|
|
|
|
650
|
|
|
// Compare local name to given element name. |
|
651
|
|
|
return $element === '*' || $node->localName === $element; |
|
652
|
|
|
} |
|
653
|
|
|
|
|
654
|
|
|
/** |
|
655
|
|
|
* Checks to see if the given DOMNode matches an "any element" (*). |
|
656
|
|
|
* |
|
657
|
|
|
* This does not handle namespaced whildcards. |
|
658
|
|
|
*/ |
|
659
|
|
|
/* |
|
660
|
|
|
protected function matchAnyElement($node) { |
|
661
|
|
|
$ancestors = $this->ancestors($node); |
|
662
|
|
|
|
|
663
|
|
|
return count($ancestors) > 0; |
|
664
|
|
|
} |
|
665
|
|
|
*/ |
|
666
|
|
|
|
|
667
|
|
|
/** |
|
668
|
|
|
* Get a list of ancestors to the present node. |
|
669
|
|
|
*/ |
|
670
|
|
|
protected function ancestors($node) |
|
671
|
|
|
{ |
|
672
|
|
|
$buffer = []; |
|
673
|
|
|
$parent = $node; |
|
674
|
|
|
while (($parent = $parent->parentNode) !== NULL) { |
|
675
|
|
|
$buffer[] = $parent; |
|
676
|
|
|
} |
|
677
|
|
|
|
|
678
|
|
|
return $buffer; |
|
679
|
|
|
} |
|
680
|
|
|
|
|
681
|
|
|
/** |
|
682
|
|
|
* Check to see if DOMNode has all of the given attributes. |
|
683
|
|
|
* |
|
684
|
|
|
* This can handle namespaced attributes, including namespace |
|
685
|
|
|
* wildcards. |
|
686
|
|
|
* @param \DOMElement $node |
|
687
|
|
|
* @param $attributes |
|
688
|
|
|
* @return bool |
|
689
|
|
|
*/ |
|
690
|
|
|
protected function matchAttributes(\DOMElement $node, $attributes) : bool |
|
691
|
|
|
{ |
|
692
|
|
|
if (empty($attributes)) { |
|
693
|
|
|
return true; |
|
694
|
|
|
} |
|
695
|
|
|
|
|
696
|
|
|
foreach ($attributes as $attr) { |
|
697
|
|
|
$val = isset($attr['value']) ? $attr['value'] : NULL; |
|
698
|
|
|
|
|
699
|
|
|
// Namespaced attributes. |
|
700
|
|
|
if (isset($attr['ns']) && $attr['ns'] !== '*') { |
|
701
|
|
|
$nsuri = $node->lookupNamespaceURI($attr['ns']); |
|
702
|
|
|
if (empty($nsuri) || !$node->hasAttributeNS($nsuri, $attr['name'])) { |
|
703
|
|
|
return false; |
|
704
|
|
|
} |
|
705
|
|
|
$matches = Util::matchesAttributeNS($node, $attr['name'], $nsuri, $val, $attr['op']); |
|
706
|
|
|
} elseif (isset($attr['ns']) && $attr['ns'] === '*' && $node->hasAttributes()) { |
|
707
|
|
|
// Cycle through all of the attributes in the node. Note that |
|
708
|
|
|
// these are DOMAttr objects. |
|
709
|
|
|
$matches = false; |
|
710
|
|
|
$name = $attr['name']; |
|
711
|
|
|
foreach ($node->attributes as $attrNode) { |
|
712
|
|
|
if ($attrNode->localName === $name) { |
|
713
|
|
|
$nsuri = $attrNode->namespaceURI; |
|
714
|
|
|
$matches = Util::matchesAttributeNS($node, $name, $nsuri, $val, $attr['op']); |
|
715
|
|
|
} |
|
716
|
|
|
} |
|
717
|
|
|
} // No namespace. |
|
718
|
|
|
else { |
|
719
|
|
|
$matches = Util::matchesAttribute($node, $attr['name'], $val, $attr['op']); |
|
720
|
|
|
} |
|
721
|
|
|
|
|
722
|
|
|
if (!$matches) { |
|
723
|
|
|
return false; |
|
724
|
|
|
} |
|
725
|
|
|
} |
|
726
|
|
|
|
|
727
|
|
|
return true; |
|
728
|
|
|
} |
|
729
|
|
|
|
|
730
|
|
|
/** |
|
731
|
|
|
* Check that the given DOMNode has the given ID. |
|
732
|
|
|
* @param \DOMElement $node |
|
733
|
|
|
* @param $id |
|
734
|
|
|
* @return bool |
|
735
|
|
|
*/ |
|
736
|
|
|
protected function matchId(\DOMElement $node, $id) : bool |
|
737
|
|
|
{ |
|
738
|
|
|
if (empty($id)) { |
|
739
|
|
|
return true; |
|
740
|
|
|
} |
|
741
|
|
|
|
|
742
|
|
|
return $node->hasAttribute('id') && $node->getAttribute('id') === $id; |
|
743
|
|
|
} |
|
744
|
|
|
|
|
745
|
|
|
/** |
|
746
|
|
|
* Check that the given DOMNode has all of the given classes. |
|
747
|
|
|
* @param \DOMElement $node |
|
748
|
|
|
* @param $classes |
|
749
|
|
|
* @return bool |
|
750
|
|
|
*/ |
|
751
|
|
|
protected function matchClasses(\DOMElement $node, $classes) : bool |
|
752
|
|
|
{ |
|
753
|
|
|
if (empty($classes)) { |
|
754
|
|
|
return true; |
|
755
|
|
|
} |
|
756
|
|
|
|
|
757
|
|
|
if (!$node->hasAttribute('class')) { |
|
758
|
|
|
return false; |
|
759
|
|
|
} |
|
760
|
|
|
|
|
761
|
|
|
$eleClasses = preg_split('/\s+/', $node->getAttribute('class')); |
|
762
|
|
|
if (empty($eleClasses)) { |
|
763
|
|
|
return false; |
|
764
|
|
|
} |
|
765
|
|
|
|
|
766
|
|
|
// The intersection should match the given $classes. |
|
767
|
|
|
$missing = array_diff($classes, array_intersect($classes, $eleClasses)); |
|
768
|
|
|
|
|
769
|
|
|
return count($missing) === 0; |
|
770
|
|
|
} |
|
771
|
|
|
|
|
772
|
|
|
/** |
|
773
|
|
|
* @param \DOMElement $node |
|
774
|
|
|
* @param $pseudoClasses |
|
775
|
|
|
* @return bool |
|
776
|
|
|
* @throws NotImplementedException |
|
777
|
|
|
* @throws ParseException |
|
778
|
|
|
*/ |
|
779
|
|
|
protected function matchPseudoClasses(\DOMElement $node, $pseudoClasses): bool |
|
780
|
|
|
{ |
|
781
|
|
|
$ret = true; |
|
782
|
|
|
foreach ($pseudoClasses as $pseudoClass) { |
|
783
|
|
|
$name = $pseudoClass['name']; |
|
784
|
|
|
// Avoid E_STRICT violation. |
|
785
|
|
|
$value = $pseudoClass['value'] ?? NULL; |
|
786
|
|
|
$ret &= $this->psHandler->elementMatches($name, $node, $this->scopeNode, $value); |
|
|
|
|
|
|
787
|
|
|
} |
|
788
|
|
|
|
|
789
|
|
|
return $ret; |
|
|
|
|
|
|
790
|
|
|
} |
|
791
|
|
|
|
|
792
|
|
|
/** |
|
793
|
|
|
* Test whether the given node matches the pseudoElements. |
|
794
|
|
|
* |
|
795
|
|
|
* If any pseudo-elements are passed, this will test to see |
|
796
|
|
|
* <i>if conditions obtain that would allow the pseudo-element |
|
797
|
|
|
* to be created</i>. This does not modify the match in any way. |
|
798
|
|
|
* @param \DOMElement $node |
|
799
|
|
|
* @param $pseudoElements |
|
800
|
|
|
* @return bool |
|
801
|
|
|
* @throws NotImplementedException |
|
802
|
|
|
*/ |
|
803
|
|
|
protected function matchPseudoElements(\DOMElement $node, $pseudoElements) : bool |
|
804
|
|
|
{ |
|
805
|
|
|
if (empty($pseudoElements)) { |
|
806
|
|
|
return true; |
|
807
|
|
|
} |
|
808
|
|
|
|
|
809
|
|
|
foreach ($pseudoElements as $pse) { |
|
810
|
|
|
switch ($pse) { |
|
811
|
|
|
case 'first-line': |
|
812
|
|
|
case 'first-letter': |
|
813
|
|
|
case 'before': |
|
814
|
|
|
case 'after': |
|
815
|
|
|
return strlen($node->textContent) > 0; |
|
816
|
|
|
case 'selection': |
|
817
|
|
|
throw new \QueryPath\CSS\NotImplementedException("::$pse is not implemented."); |
|
818
|
|
|
} |
|
819
|
|
|
} |
|
820
|
|
|
|
|
821
|
|
|
return false; |
|
822
|
|
|
} |
|
823
|
|
|
|
|
824
|
|
|
protected function newMatches() |
|
825
|
|
|
{ |
|
826
|
|
|
return new \SplObjectStorage(); |
|
827
|
|
|
} |
|
828
|
|
|
|
|
829
|
|
|
/** |
|
830
|
|
|
* Get the internal match set. |
|
831
|
|
|
* Internal utility function. |
|
832
|
|
|
*/ |
|
833
|
|
|
protected function getMatches() |
|
834
|
|
|
{ |
|
835
|
|
|
return $this->matches(); |
|
836
|
|
|
} |
|
837
|
|
|
|
|
838
|
|
|
/** |
|
839
|
|
|
* Set the internal match set. |
|
840
|
|
|
* |
|
841
|
|
|
* Internal utility function. |
|
842
|
|
|
*/ |
|
843
|
|
|
protected function setMatches($matches) |
|
844
|
|
|
{ |
|
845
|
|
|
$this->matches = $matches; |
|
846
|
|
|
} |
|
847
|
|
|
|
|
848
|
|
|
/** |
|
849
|
|
|
* Attach all nodes in a node list to the given \SplObjectStorage. |
|
850
|
|
|
* @param \DOMNodeList $nodeList |
|
851
|
|
|
* @param \SplObjectStorage $splos |
|
852
|
|
|
*/ |
|
853
|
|
|
public function attachNodeList(\DOMNodeList $nodeList, \SplObjectStorage $splos) |
|
854
|
|
|
{ |
|
855
|
|
|
foreach ($nodeList as $item) { |
|
856
|
|
|
$splos->attach($item); |
|
857
|
|
|
} |
|
858
|
|
|
} |
|
859
|
|
|
|
|
860
|
|
|
public function getDocument() |
|
861
|
|
|
{ |
|
862
|
|
|
return $this->dom; |
|
863
|
|
|
} |
|
864
|
|
|
|
|
865
|
|
|
} |
|
866
|
|
|
|
The issue could also be caused by a filter entry in the build configuration. If the path has been excluded in your configuration, e.g.
excluded_paths: ["lib/*"], you can move it to the dependency path list as follows:For further information see https://scrutinizer-ci.com/docs/tools/php/php-scrutinizer/#list-dependency-paths