1
|
|
|
<?php |
2
|
|
|
|
3
|
|
|
/** |
4
|
|
|
* @package s9e\TextFormatter |
5
|
|
|
* @copyright Copyright (c) 2010-2019 The s9e Authors |
6
|
|
|
* @license http://www.opensource.org/licenses/mit-license.php The MIT License |
7
|
|
|
*/ |
8
|
|
|
namespace s9e\TextFormatter\Plugins\HTMLElements; |
9
|
|
|
|
10
|
|
|
use InvalidArgumentException; |
11
|
|
|
use RuntimeException; |
12
|
|
|
use s9e\TextFormatter\Configurator\Helpers\RegexpBuilder; |
13
|
|
|
use s9e\TextFormatter\Configurator\Items\Tag; |
14
|
|
|
use s9e\TextFormatter\Configurator\Items\UnsafeTemplate; |
15
|
|
|
use s9e\TextFormatter\Configurator\JavaScript\Dictionary; |
16
|
|
|
use s9e\TextFormatter\Configurator\Validators\AttributeName; |
17
|
|
|
use s9e\TextFormatter\Configurator\Validators\TagName; |
18
|
|
|
use s9e\TextFormatter\Plugins\ConfiguratorBase; |
19
|
|
|
|
20
|
|
|
class Configurator extends ConfiguratorBase |
21
|
|
|
{ |
22
|
|
|
/** |
23
|
|
|
* @var array 2D array using HTML element names as keys, each value being an associative array |
24
|
|
|
* using HTML attribute names as keys and their alias as values. A special empty entry |
25
|
|
|
* is used to store the HTML element's alias |
26
|
|
|
*/ |
27
|
|
|
protected $aliases = []; |
28
|
|
|
|
29
|
|
|
/** |
30
|
|
|
* @var array Default filter of a few known attributes |
31
|
|
|
* |
32
|
|
|
* It doesn't make much sense to try to declare every known HTML attribute here. Validation is |
33
|
|
|
* not the purpose of this plugin. It does make sense however to declare URL attributes as such, |
34
|
|
|
* so that they are subject to our constraints (disallowed hosts, etc...) |
35
|
|
|
* |
36
|
|
|
* @see scripts/patchHTMLElementConfigurator.php |
37
|
|
|
*/ |
38
|
|
|
protected $attributeFilters = [ |
39
|
|
|
'action' => '#url', |
40
|
|
|
'cite' => '#url', |
41
|
|
|
'data' => '#url', |
42
|
|
|
'formaction' => '#url', |
43
|
|
|
'href' => '#url', |
44
|
|
|
'icon' => '#url', |
45
|
|
|
'longdesc' => '#url', |
46
|
|
|
'manifest' => '#url', |
47
|
|
|
'ping' => '#url', |
48
|
|
|
'poster' => '#url', |
49
|
|
|
'src' => '#url' |
50
|
|
|
]; |
51
|
|
|
|
52
|
|
|
/** |
53
|
|
|
* @var array Hash of allowed HTML elements. Element names are lowercased and used as keys for |
54
|
|
|
* this array |
55
|
|
|
*/ |
56
|
|
|
protected $elements = []; |
57
|
|
|
|
58
|
|
|
/** |
59
|
|
|
* @var string Namespace prefix of the tags produced by this plugin's parser |
60
|
|
|
*/ |
61
|
|
|
protected $prefix = 'html'; |
62
|
|
|
|
63
|
|
|
/** |
64
|
|
|
* {@inheritdoc} |
65
|
|
|
*/ |
66
|
|
|
protected $quickMatch = '<'; |
67
|
|
|
|
68
|
|
|
/** |
69
|
|
|
* @var array Blacklist of elements that are considered unsafe |
70
|
|
|
*/ |
71
|
|
|
protected $unsafeElements = [ |
72
|
|
|
'base', |
73
|
|
|
'embed', |
74
|
|
|
'frame', |
75
|
|
|
'iframe', |
76
|
|
|
'meta', |
77
|
|
|
'object', |
78
|
|
|
'script' |
79
|
|
|
]; |
80
|
|
|
|
81
|
|
|
/** |
82
|
|
|
* @var array Blacklist of attributes that are considered unsafe, in addition of any attribute |
83
|
|
|
* whose name starts with "on" such as "onmouseover" |
84
|
|
|
*/ |
85
|
|
|
protected $unsafeAttributes = [ |
86
|
|
|
'style', |
87
|
|
|
'target' |
88
|
|
|
]; |
89
|
|
|
|
90
|
|
|
/** |
91
|
|
|
* Alias the HTML attribute of given HTML element to a given attribute name |
92
|
|
|
* |
93
|
|
|
* NOTE: will *not* create the target attribute |
94
|
|
|
* |
95
|
|
|
* @param string $elName Name of the HTML element |
96
|
|
|
* @param string $attrName Name of the HTML attribute |
97
|
|
|
* @param string $alias Alias |
98
|
|
|
* @return void |
99
|
|
|
*/ |
100
|
2 |
|
public function aliasAttribute($elName, $attrName, $alias) |
101
|
|
|
{ |
102
|
2 |
|
$elName = $this->normalizeElementName($elName); |
103
|
2 |
|
$attrName = $this->normalizeAttributeName($attrName); |
104
|
|
|
|
105
|
2 |
|
$this->aliases[$elName][$attrName] = AttributeName::normalize($alias); |
106
|
|
|
} |
107
|
|
|
|
108
|
|
|
/** |
109
|
|
|
* Alias an HTML element to a given tag name |
110
|
|
|
* |
111
|
|
|
* NOTE: will *not* create the target tag |
112
|
|
|
* |
113
|
|
|
* @param string $elName Name of the HTML element |
114
|
|
|
* @param string $tagName Name of the tag |
115
|
|
|
* @return void |
116
|
|
|
*/ |
117
|
3 |
|
public function aliasElement($elName, $tagName) |
118
|
|
|
{ |
119
|
3 |
|
$elName = $this->normalizeElementName($elName); |
120
|
|
|
|
121
|
3 |
|
$this->aliases[$elName][''] = TagName::normalize($tagName); |
122
|
|
|
} |
123
|
|
|
|
124
|
|
|
/** |
125
|
|
|
* Allow an HTML element to be used |
126
|
|
|
* |
127
|
|
|
* @param string $elName Name of the element |
128
|
|
|
* @return Tag Tag that represents this element |
129
|
|
|
*/ |
130
|
21 |
|
public function allowElement($elName) |
131
|
|
|
{ |
132
|
21 |
|
return $this->allowElementWithSafety($elName, false); |
133
|
|
|
} |
134
|
|
|
|
135
|
|
|
/** |
136
|
|
|
* Allow an unsafe HTML element to be used |
137
|
|
|
* |
138
|
|
|
* @param string $elName Name of the element |
139
|
|
|
* @return Tag Tag that represents this element |
140
|
|
|
*/ |
141
|
2 |
|
public function allowUnsafeElement($elName) |
142
|
|
|
{ |
143
|
2 |
|
return $this->allowElementWithSafety($elName, true); |
144
|
|
|
} |
145
|
|
|
|
146
|
|
|
/** |
147
|
|
|
* Allow a (potentially unsafe) HTML element to be used |
148
|
|
|
* |
149
|
|
|
* @param string $elName Name of the element |
150
|
|
|
* @param bool $allowUnsafe Whether to allow unsafe elements |
151
|
|
|
* @return Tag Tag that represents this element |
152
|
|
|
*/ |
153
|
23 |
|
protected function allowElementWithSafety($elName, $allowUnsafe) |
154
|
|
|
{ |
155
|
23 |
|
$elName = $this->normalizeElementName($elName); |
156
|
22 |
|
$tagName = $this->prefix . ':' . $elName; |
157
|
|
|
|
158
|
22 |
|
if (!$allowUnsafe && in_array($elName, $this->unsafeElements)) |
159
|
|
|
{ |
160
|
1 |
|
throw new RuntimeException("'" . $elName . "' elements are unsafe and are disabled by default. Please use " . __CLASS__ . '::allowUnsafeElement() to bypass this security measure'); |
161
|
|
|
} |
162
|
|
|
|
163
|
|
|
// Retrieve or create the tag |
164
|
21 |
|
$tag = ($this->configurator->tags->exists($tagName)) |
165
|
1 |
|
? $this->configurator->tags->get($tagName) |
166
|
21 |
|
: $this->configurator->tags->add($tagName); |
167
|
|
|
|
168
|
|
|
// Rebuild this tag's template |
169
|
21 |
|
$this->rebuildTemplate($tag, $elName, $allowUnsafe); |
170
|
|
|
|
171
|
|
|
// Record the element name |
172
|
21 |
|
$this->elements[$elName] = 1; |
173
|
|
|
|
174
|
21 |
|
return $tag; |
175
|
|
|
} |
176
|
|
|
|
177
|
|
|
/** |
178
|
|
|
* Allow an attribute to be used in an HTML element |
179
|
|
|
* |
180
|
|
|
* @param string $elName Name of the element |
181
|
|
|
* @param string $attrName Name of the attribute |
182
|
|
|
* @return \s9e\Configurator\Items\Attribute |
|
|
|
|
183
|
|
|
*/ |
184
|
8 |
|
public function allowAttribute($elName, $attrName) |
185
|
|
|
{ |
186
|
8 |
|
return $this->allowAttributeWithSafety($elName, $attrName, false); |
187
|
|
|
} |
188
|
|
|
|
189
|
|
|
/** |
190
|
|
|
* Allow an unsafe attribute to be used in an HTML element |
191
|
|
|
* |
192
|
|
|
* @param string $elName Name of the element |
193
|
|
|
* @param string $attrName Name of the attribute |
194
|
|
|
* @return \s9e\Configurator\Items\Attribute |
195
|
|
|
*/ |
196
|
3 |
|
public function allowUnsafeAttribute($elName, $attrName) |
197
|
|
|
{ |
198
|
3 |
|
return $this->allowAttributeWithSafety($elName, $attrName, true); |
199
|
|
|
} |
200
|
|
|
|
201
|
|
|
/** |
202
|
|
|
* Allow a (potentially unsafe) attribute to be used in an HTML element |
203
|
|
|
* |
204
|
|
|
* @param string $elName Name of the element |
205
|
|
|
* @param string $attrName Name of the attribute |
206
|
|
|
* @param bool $allowUnsafe |
207
|
|
|
* @return \s9e\Configurator\Items\Attribute |
208
|
|
|
*/ |
209
|
11 |
|
protected function allowAttributeWithSafety($elName, $attrName, $allowUnsafe) |
210
|
|
|
{ |
211
|
11 |
|
$elName = $this->normalizeElementName($elName); |
212
|
11 |
|
$attrName = $this->normalizeAttributeName($attrName); |
213
|
10 |
|
$tagName = $this->prefix . ':' . $elName; |
214
|
|
|
|
215
|
10 |
|
if (!isset($this->elements[$elName])) |
216
|
|
|
{ |
217
|
1 |
|
throw new RuntimeException("Element '" . $elName . "' has not been allowed"); |
218
|
|
|
} |
219
|
|
|
|
220
|
9 |
|
if (!$allowUnsafe) |
221
|
|
|
{ |
222
|
6 |
|
if (substr($attrName, 0, 2) === 'on' |
223
|
6 |
|
|| in_array($attrName, $this->unsafeAttributes)) |
224
|
|
|
{ |
225
|
2 |
|
throw new RuntimeException("'" . $attrName . "' attributes are unsafe and are disabled by default. Please use " . __CLASS__ . '::allowUnsafeAttribute() to bypass this security measure'); |
226
|
|
|
} |
227
|
|
|
} |
228
|
|
|
|
229
|
7 |
|
$tag = $this->configurator->tags->get($tagName); |
230
|
7 |
|
if (!isset($tag->attributes[$attrName])) |
231
|
|
|
{ |
232
|
7 |
|
$attribute = $tag->attributes->add($attrName); |
233
|
7 |
|
$attribute->required = false; |
234
|
|
|
|
235
|
7 |
|
if (isset($this->attributeFilters[$attrName])) |
236
|
|
|
{ |
237
|
1 |
|
$filterName = $this->attributeFilters[$attrName]; |
238
|
1 |
|
$filter = $this->configurator->attributeFilters->get($filterName); |
239
|
|
|
|
240
|
1 |
|
$attribute->filterChain->append($filter); |
241
|
|
|
} |
242
|
|
|
} |
243
|
|
|
|
244
|
|
|
// Rebuild this tag's template |
245
|
7 |
|
$this->rebuildTemplate($tag, $elName, $allowUnsafe); |
246
|
|
|
|
247
|
7 |
|
return $tag->attributes[$attrName]; |
248
|
|
|
} |
249
|
|
|
|
250
|
|
|
/** |
251
|
|
|
* Validate and normalize an element name |
252
|
|
|
* |
253
|
|
|
* Accepts any name that would be valid, regardless of whether this element exists in HTML5. |
254
|
|
|
* Might be slightly off as the HTML5 specs don't seem to require it to start with a letter but |
255
|
|
|
* our implementation does. |
256
|
|
|
* |
257
|
|
|
* @link http://dev.w3.org/html5/spec/syntax.html#syntax-tag-name |
258
|
|
|
* |
259
|
|
|
* @param string $elName Original element name |
260
|
|
|
* @return string Normalized element name, in lowercase |
261
|
|
|
*/ |
262
|
28 |
|
protected function normalizeElementName($elName) |
263
|
|
|
{ |
264
|
28 |
|
if (!preg_match('#^[a-z][a-z0-9]*$#Di', $elName)) |
265
|
|
|
{ |
266
|
1 |
|
throw new InvalidArgumentException ("Invalid element name '" . $elName . "'"); |
267
|
|
|
} |
268
|
|
|
|
269
|
27 |
|
return strtolower($elName); |
270
|
|
|
} |
271
|
|
|
|
272
|
|
|
/** |
273
|
|
|
* Validate and normalize an attribute name |
274
|
|
|
* |
275
|
|
|
* More restrictive than the specs but allows all HTML5 attributes and more. |
276
|
|
|
* |
277
|
|
|
* @param string $attrName Original attribute name |
278
|
|
|
* @return string Normalized attribute name, in lowercase |
279
|
|
|
*/ |
280
|
13 |
|
protected function normalizeAttributeName($attrName) |
281
|
|
|
{ |
282
|
13 |
|
if (!preg_match('#^[a-z][-\\w]*$#Di', $attrName)) |
283
|
|
|
{ |
284
|
1 |
|
throw new InvalidArgumentException ("Invalid attribute name '" . $attrName . "'"); |
285
|
|
|
} |
286
|
|
|
|
287
|
12 |
|
return strtolower($attrName); |
288
|
|
|
} |
289
|
|
|
|
290
|
|
|
/** |
291
|
|
|
* Rebuild a tag's template |
292
|
|
|
* |
293
|
|
|
* @param Tag $tag Source tag |
294
|
|
|
* @param string $elName Name of the HTML element created by the template |
295
|
|
|
* @param bool $allowUnsafe Whether to allow unsafe markup |
296
|
|
|
* @return void |
297
|
|
|
*/ |
298
|
21 |
|
protected function rebuildTemplate(Tag $tag, $elName, $allowUnsafe) |
299
|
|
|
{ |
300
|
21 |
|
$template = '<' . $elName . '>'; |
301
|
21 |
|
foreach ($tag->attributes as $attrName => $attribute) |
302
|
|
|
{ |
303
|
7 |
|
$template .= '<xsl:copy-of select="@' . $attrName . '"/>'; |
304
|
|
|
} |
305
|
21 |
|
$template .= '<xsl:apply-templates/></' . $elName . '>'; |
306
|
|
|
|
307
|
21 |
|
if ($allowUnsafe) |
308
|
|
|
{ |
309
|
5 |
|
$template = new UnsafeTemplate($template); |
310
|
|
|
} |
311
|
|
|
|
312
|
21 |
|
$tag->setTemplate($template); |
313
|
|
|
} |
314
|
|
|
|
315
|
|
|
/** |
316
|
|
|
* Generate this plugin's config |
317
|
|
|
* |
318
|
|
|
* @return array|null |
319
|
|
|
*/ |
320
|
7 |
|
public function asConfig() |
321
|
|
|
{ |
322
|
7 |
|
if (empty($this->elements) && empty($this->aliases)) |
323
|
|
|
{ |
324
|
1 |
|
return; |
325
|
|
|
} |
326
|
|
|
|
327
|
|
|
/** |
328
|
|
|
* Regexp used to match an attributes definition (name + value if applicable) |
329
|
|
|
* |
330
|
|
|
* @link http://dev.w3.org/html5/spec/syntax.html#attributes-0 |
331
|
|
|
*/ |
332
|
6 |
|
$attrRegexp = '[a-z][-a-z0-9]*(?>\\s*=\\s*(?>"[^"]*"|\'[^\']*\'|[^\\s"\'=<>`]+))?'; |
333
|
6 |
|
$tagRegexp = RegexpBuilder::fromList(array_merge( |
334
|
6 |
|
array_keys($this->aliases), |
335
|
6 |
|
array_keys($this->elements) |
336
|
|
|
)); |
337
|
|
|
|
338
|
6 |
|
$endTagRegexp = '/(' . $tagRegexp . ')'; |
339
|
6 |
|
$startTagRegexp = '(' . $tagRegexp . ')((?>\\s+' . $attrRegexp . ')*+)\\s*/?'; |
340
|
|
|
|
341
|
6 |
|
$regexp = '#<(?>' . $endTagRegexp . '|' . $startTagRegexp . ')\\s*>#i'; |
342
|
|
|
|
343
|
|
|
$config = [ |
344
|
6 |
|
'quickMatch' => $this->quickMatch, |
345
|
6 |
|
'prefix' => $this->prefix, |
346
|
6 |
|
'regexp' => $regexp |
347
|
|
|
]; |
348
|
|
|
|
349
|
6 |
|
if (!empty($this->aliases)) |
350
|
|
|
{ |
351
|
|
|
// Preserve the aliases array's keys in JavaScript |
352
|
4 |
|
$config['aliases'] = new Dictionary; |
353
|
4 |
|
foreach ($this->aliases as $elName => $aliases) |
354
|
|
|
{ |
355
|
4 |
|
$config['aliases'][$elName] = new Dictionary($aliases); |
356
|
|
|
} |
357
|
|
|
} |
358
|
|
|
|
359
|
6 |
|
return $config; |
360
|
|
|
} |
361
|
|
|
|
362
|
|
|
/** |
363
|
|
|
* {@inheritdoc} |
364
|
|
|
*/ |
365
|
2 |
|
public function getJSHints() |
366
|
|
|
{ |
367
|
2 |
|
return ['HTMLELEMENTS_HAS_ALIASES' => (int) !empty($this->aliases)]; |
368
|
|
|
} |
369
|
|
|
} |
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