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
|
|
|
|
19
|
|
|
abstract class TemplateHelper |
20
|
|
|
{ |
21
|
|
|
/** |
22
|
|
|
* XSL namespace |
23
|
|
|
*/ |
24
|
|
|
const XMLNS_XSL = 'http://www.w3.org/1999/XSL/Transform'; |
25
|
|
|
|
26
|
|
|
/** |
27
|
|
|
* Return all attributes (literal or generated) that match given regexp |
28
|
|
|
* |
29
|
|
|
* @param DOMDocument $dom Document |
30
|
|
|
* @param string $regexp Regexp |
31
|
|
|
* @return array Array of DOMNode instances |
32
|
|
|
*/ |
33
|
|
|
public static function getAttributesByRegexp(DOMDocument $dom, $regexp) |
34
|
|
|
{ |
35
|
|
|
return NodeLocator::getAttributesByRegexp($dom, $regexp); |
36
|
|
|
} |
37
|
|
|
|
38
|
|
|
/** |
39
|
|
|
* Return all DOMNodes whose content is CSS |
40
|
|
|
* |
41
|
|
|
* @param DOMDocument $dom Document |
42
|
|
|
* @return array Array of DOMNode instances |
43
|
|
|
*/ |
44
|
|
|
public static function getCSSNodes(DOMDocument $dom) |
45
|
|
|
{ |
46
|
|
|
return NodeLocator::getCSSNodes($dom); |
47
|
|
|
} |
48
|
|
|
|
49
|
|
|
/** |
50
|
|
|
* Return all elements (literal or generated) that match given regexp |
51
|
|
|
* |
52
|
|
|
* @param DOMDocument $dom Document |
53
|
|
|
* @param string $regexp Regexp |
54
|
|
|
* @return array Array of DOMNode instances |
55
|
|
|
*/ |
56
|
|
|
public static function getElementsByRegexp(DOMDocument $dom, $regexp) |
57
|
|
|
{ |
58
|
|
|
return NodeLocator::getElementsByRegexp($dom, $regexp); |
59
|
|
|
} |
60
|
|
|
|
61
|
|
|
/** |
62
|
|
|
* Return all DOMNodes whose content is JavaScript |
63
|
|
|
* |
64
|
|
|
* @param DOMDocument $dom Document |
65
|
|
|
* @return array Array of DOMNode instances |
66
|
|
|
*/ |
67
|
|
|
public static function getJSNodes(DOMDocument $dom) |
68
|
|
|
{ |
69
|
|
|
return NodeLocator::getJSNodes($dom); |
70
|
|
|
} |
71
|
|
|
|
72
|
|
|
/** |
73
|
|
|
* Return all elements (literal or generated) that match given regexp |
74
|
|
|
* |
75
|
|
|
* Will return all <param/> descendants of <object/> and all attributes of <embed/> whose name |
76
|
|
|
* matches given regexp. This method will NOT catch <param/> elements whose 'name' attribute is |
77
|
|
|
* set via an <xsl:attribute/> |
78
|
|
|
* |
79
|
|
|
* @param DOMDocument $dom Document |
80
|
|
|
* @param string $regexp |
81
|
|
|
* @return array Array of DOMNode instances |
82
|
|
|
*/ |
83
|
|
|
public static function getObjectParamsByRegexp(DOMDocument $dom, $regexp) |
84
|
|
|
{ |
85
|
|
|
return NodeLocator::getObjectParamsByRegexp($dom, $regexp); |
86
|
|
|
} |
87
|
|
|
|
88
|
|
|
/** |
89
|
|
|
* Return a list of parameters in use in given XSL |
90
|
|
|
* |
91
|
|
|
* @param string $xsl XSL source |
92
|
|
|
* @return array Alphabetically sorted list of unique parameter names |
93
|
|
|
*/ |
94
|
|
|
public static function getParametersFromXSL($xsl) |
95
|
|
|
{ |
96
|
|
|
$paramNames = []; |
97
|
|
|
$xpath = new DOMXPath(TemplateLoader::load($xsl)); |
98
|
|
|
|
99
|
|
|
// Start by collecting XPath expressions in XSL elements |
100
|
|
|
$query = '//xsl:*/@match | //xsl:*/@select | //xsl:*/@test'; |
101
|
|
|
foreach ($xpath->query($query) as $attribute) |
102
|
|
|
{ |
103
|
|
|
$expr = $attribute->value; |
104
|
|
|
$paramNames += array_flip(self::getParametersFromExpression($attribute, $expr)); |
105
|
|
|
} |
106
|
|
|
|
107
|
|
|
// Collect XPath expressions in attribute value templates |
108
|
|
|
$query = '//*[namespace-uri() != "' . self::XMLNS_XSL . '"]/@*[contains(., "{")]'; |
109
|
|
|
foreach ($xpath->query($query) as $attribute) |
110
|
|
|
{ |
111
|
|
|
foreach (AVTHelper::parse($attribute->value) as $token) |
112
|
|
|
{ |
113
|
|
|
if ($token[0] === 'expression') |
114
|
|
|
{ |
115
|
|
|
$expr = $token[1]; |
116
|
|
|
$paramNames += array_flip(self::getParametersFromExpression($attribute, $expr)); |
117
|
|
|
} |
118
|
|
|
} |
119
|
|
|
} |
120
|
|
|
|
121
|
|
|
// Sort the parameter names and return them in a list |
122
|
|
|
ksort($paramNames); |
123
|
|
|
|
124
|
|
|
return array_keys($paramNames); |
125
|
|
|
} |
126
|
|
|
|
127
|
|
|
/** |
128
|
|
|
* Return all DOMNodes whose content is an URL |
129
|
|
|
* |
130
|
|
|
* NOTE: it will also return HTML4 nodes whose content is an URI |
131
|
|
|
* |
132
|
|
|
* @param DOMDocument $dom Document |
133
|
|
|
* @return array Array of DOMNode instances |
134
|
|
|
*/ |
135
|
|
|
public static function getURLNodes(DOMDocument $dom) |
136
|
|
|
{ |
137
|
|
|
return NodeLocator::getURLNodes($dom); |
138
|
|
|
} |
139
|
|
|
|
140
|
|
|
/** |
141
|
|
|
* Highlight the source of a node inside of a template |
142
|
|
|
* |
143
|
|
|
* @param DOMNode $node Node to highlight |
144
|
|
|
* @param string $prepend HTML to prepend |
145
|
|
|
* @param string $append HTML to append |
146
|
|
|
* @return string Template's source, as HTML |
147
|
|
|
*/ |
148
|
|
|
public static function highlightNode(DOMNode $node, $prepend, $append) |
149
|
|
|
{ |
150
|
|
|
// Add a unique token to the node |
151
|
|
|
$uniqid = uniqid('_'); |
152
|
|
|
if ($node instanceof DOMAttr) |
153
|
|
|
{ |
154
|
|
|
$node->value .= $uniqid; |
155
|
|
|
} |
156
|
|
|
elseif ($node instanceof DOMElement) |
157
|
|
|
{ |
158
|
|
|
$node->setAttribute($uniqid, ''); |
159
|
|
|
} |
160
|
|
|
elseif ($node instanceof DOMCharacterData || $node instanceof DOMProcessingInstruction) |
161
|
|
|
{ |
162
|
|
|
$node->data .= $uniqid; |
163
|
|
|
} |
164
|
|
|
|
165
|
|
|
$dom = $node->ownerDocument; |
166
|
|
|
$dom->formatOutput = true; |
167
|
|
|
|
168
|
|
|
$docXml = TemplateLoader::innerXML($dom->documentElement); |
169
|
|
|
$docXml = trim(str_replace("\n ", "\n", $docXml)); |
170
|
|
|
|
171
|
|
|
$nodeHtml = htmlspecialchars(trim($dom->saveXML($node))); |
172
|
|
|
$docHtml = htmlspecialchars($docXml); |
173
|
|
|
|
174
|
|
|
// Enclose the node's representation in our hilighting HTML |
175
|
|
|
$html = str_replace($nodeHtml, $prepend . $nodeHtml . $append, $docHtml); |
176
|
|
|
|
177
|
|
|
// Remove the unique token from HTML and from the node |
178
|
|
|
if ($node instanceof DOMAttr) |
179
|
|
|
{ |
180
|
|
|
$node->value = substr($node->value, 0, -strlen($uniqid)); |
181
|
|
|
$html = str_replace($uniqid, '', $html); |
182
|
|
|
} |
183
|
|
|
elseif ($node instanceof DOMElement) |
184
|
|
|
{ |
185
|
|
|
$node->removeAttribute($uniqid); |
186
|
|
|
$html = str_replace(' ' . $uniqid . '=""', '', $html); |
187
|
|
|
} |
188
|
|
|
elseif ($node instanceof DOMCharacterData || $node instanceof DOMProcessingInstruction) |
189
|
|
|
{ |
190
|
|
|
$node->data .= $uniqid; |
191
|
|
|
$html = str_replace($uniqid, '', $html); |
192
|
|
|
} |
193
|
|
|
|
194
|
|
|
return $html; |
195
|
|
|
} |
196
|
|
|
|
197
|
|
|
/** |
198
|
|
|
* Load a template as an xsl:template node |
199
|
|
|
* |
200
|
|
|
* Will attempt to load it as XML first, then as HTML as a fallback. Either way, an xsl:template |
201
|
|
|
* node is returned |
202
|
|
|
* |
203
|
|
|
* @param string $template |
204
|
|
|
* @return DOMDocument |
205
|
|
|
*/ |
206
|
|
|
public static function loadTemplate($template) |
207
|
|
|
{ |
208
|
|
|
return TemplateLoader::load($template); |
209
|
|
|
} |
210
|
|
|
|
211
|
|
|
/** |
212
|
|
|
* Replace simple templates (in an array, in-place) with a common template |
213
|
|
|
* |
214
|
|
|
* In some situations, renderers can take advantage of multiple tags having the same template. In |
215
|
|
|
* any configuration, there's almost always a number of "simple" tags that are rendered as an |
216
|
|
|
* HTML element of the same name with no HTML attributes. For instance, the system tag "p" used |
217
|
|
|
* for paragraphs, "B" tags used for "b" HTML elements, etc... This method replaces those |
218
|
|
|
* templates with a common template that uses a dynamic element name based on the tag's name, |
219
|
|
|
* either its nodeName or localName depending on whether the tag is namespaced, and normalized to |
220
|
|
|
* lowercase using XPath's translate() function |
221
|
|
|
* |
222
|
|
|
* @param array<string> &$templates Associative array of [tagName => template] |
223
|
|
|
* @param integer $minCount |
224
|
|
|
* @return void |
225
|
|
|
*/ |
226
|
|
|
public static function replaceHomogeneousTemplates(array &$templates, $minCount = 3) |
227
|
|
|
{ |
228
|
|
|
// Prepare the XPath expression used for the element's name |
229
|
|
|
$expr = 'name()'; |
230
|
|
|
|
231
|
|
|
// Identify "simple" tags, whose template is one element of the same name. Their template |
232
|
|
|
// can be replaced with a dynamic template shared by all the simple tags |
233
|
|
|
$tagNames = []; |
234
|
|
|
foreach ($templates as $tagName => $template) |
235
|
|
|
{ |
236
|
|
|
// Generate the element name based on the tag's localName, lowercased |
237
|
|
|
$elName = strtolower(preg_replace('/^[^:]+:/', '', $tagName)); |
238
|
|
|
if ($template === '<' . $elName . '><xsl:apply-templates/></' . $elName . '>') |
239
|
|
|
{ |
240
|
|
|
$tagNames[] = $tagName; |
241
|
|
|
|
242
|
|
|
// Use local-name() if any of the tags are namespaced |
243
|
|
|
if (strpos($tagName, ':') !== false) |
244
|
|
|
{ |
245
|
|
|
$expr = 'local-name()'; |
246
|
|
|
} |
247
|
|
|
} |
248
|
|
|
} |
249
|
|
|
|
250
|
|
|
// We only bother replacing their template if there are at least $minCount simple tags. |
251
|
|
|
// Otherwise it only makes the stylesheet bigger |
252
|
|
|
if (count($tagNames) < $minCount) |
253
|
|
|
{ |
254
|
|
|
return; |
255
|
|
|
} |
256
|
|
|
|
257
|
|
|
// Generate a list of uppercase characters from the tags' names |
258
|
|
|
$chars = preg_replace('/[^A-Z]+/', '', count_chars(implode('', $tagNames), 3)); |
259
|
|
|
if ($chars > '') |
260
|
|
|
{ |
261
|
|
|
$expr = 'translate(' . $expr . ",'" . $chars . "','" . strtolower($chars) . "')"; |
262
|
|
|
} |
263
|
|
|
|
264
|
|
|
// Prepare the common template |
265
|
|
|
$template = '<xsl:element name="{' . $expr . '}"><xsl:apply-templates/></xsl:element>'; |
266
|
|
|
|
267
|
|
|
// Replace the templates |
268
|
|
|
foreach ($tagNames as $tagName) |
269
|
|
|
{ |
270
|
|
|
$templates[$tagName] = $template; |
271
|
|
|
} |
272
|
|
|
} |
273
|
|
|
|
274
|
|
|
/** |
275
|
|
|
* Replace parts of a template that match given regexp |
276
|
|
|
* |
277
|
|
|
* Treats attribute values as plain text. Replacements within XPath expression is unsupported. |
278
|
|
|
* The callback must return an array with two elements. The first must be either of 'expression', |
279
|
|
|
* 'literal' or 'passthrough', and the second element depends on the first. |
280
|
|
|
* |
281
|
|
|
* - 'expression' indicates that the replacement must be treated as an XPath expression such as |
282
|
|
|
* '@foo', which must be passed as the second element. |
283
|
|
|
* - 'literal' indicates a literal (plain text) replacement, passed as its second element. |
284
|
|
|
* - 'passthrough' indicates that the replacement should the tag's content. It works differently |
285
|
|
|
* whether it is inside an attribute's value or a text node. Within an attribute's value, the |
286
|
|
|
* replacement will be the text content of the tag. Within a text node, the replacement |
287
|
|
|
* becomes an <xsl:apply-templates/> node. |
288
|
|
|
* |
289
|
|
|
* @param string $template Original template |
290
|
|
|
* @param string $regexp Regexp for matching parts that need replacement |
291
|
|
|
* @param callback $fn Callback used to get the replacement |
292
|
|
|
* @return string Processed template |
293
|
|
|
*/ |
294
|
|
|
public static function replaceTokens($template, $regexp, $fn) |
295
|
|
|
{ |
296
|
|
|
return TemplateModifier::replaceTokens($template, $regexp, $fn); |
297
|
|
|
} |
298
|
|
|
|
299
|
|
|
/** |
300
|
|
|
* Serialize a loaded template back into a string |
301
|
|
|
* |
302
|
|
|
* NOTE: removes the root node created by loadTemplate() |
303
|
|
|
* |
304
|
|
|
* @param DOMDocument $dom |
305
|
|
|
* @return string |
306
|
|
|
*/ |
307
|
|
|
public static function saveTemplate(DOMDocument $dom) |
308
|
|
|
{ |
309
|
|
|
return TemplateLoader::save($dom); |
310
|
|
|
} |
311
|
|
|
|
312
|
|
|
/** |
313
|
|
|
* Get a list of parameters from given XPath expression |
314
|
|
|
* |
315
|
|
|
* @param DOMNode $node Context node |
316
|
|
|
* @param string $expr XPath expression |
317
|
|
|
* @return string[] |
318
|
|
|
*/ |
319
|
|
|
protected static function getParametersFromExpression(DOMNode $node, $expr) |
320
|
|
|
{ |
321
|
|
|
$varNames = XPathHelper::getVariables($expr); |
322
|
|
|
$paramNames = []; |
323
|
|
|
$xpath = new DOMXPath($node->ownerDocument); |
324
|
|
|
foreach ($varNames as $name) |
325
|
|
|
{ |
326
|
|
|
// Test whether this is the name of a local variable |
327
|
|
|
$query = 'ancestor-or-self::*/preceding-sibling::xsl:variable[@name="' . $name . '"]'; |
328
|
|
|
if (!$xpath->query($query, $node)->length) |
329
|
|
|
{ |
330
|
|
|
$paramNames[] = $name; |
331
|
|
|
} |
332
|
|
|
} |
333
|
|
|
|
334
|
|
|
return $paramNames; |
335
|
|
|
} |
336
|
|
|
} |