|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
/** |
|
4
|
|
|
* Converts HTML to Markdown text |
|
5
|
|
|
* |
|
6
|
|
|
* @name ElkArte Forum |
|
7
|
|
|
* @copyright ElkArte Forum contributors |
|
8
|
|
|
* @license BSD http://opensource.org/licenses/BSD-3-Clause |
|
9
|
|
|
* |
|
10
|
|
|
* @version 1.0.10 |
|
11
|
|
|
* |
|
12
|
|
|
*/ |
|
13
|
|
|
|
|
14
|
|
|
if (!defined('ELK')) |
|
15
|
|
|
die('No access...'); |
|
16
|
|
|
|
|
17
|
|
|
/** |
|
18
|
|
|
* Converts HTML to Markdown text |
|
19
|
|
|
*/ |
|
20
|
|
|
class Html_2_Md |
|
21
|
|
|
{ |
|
22
|
|
|
/** |
|
23
|
|
|
* The value that will hold our dom object |
|
24
|
|
|
* @var object |
|
25
|
|
|
*/ |
|
26
|
|
|
public $doc; |
|
27
|
|
|
|
|
28
|
|
|
/** |
|
29
|
|
|
* The value that will hold if we are using the internal or external parser |
|
30
|
|
|
* @var boolean |
|
31
|
|
|
*/ |
|
32
|
|
|
private $_parser; |
|
33
|
|
|
|
|
34
|
|
|
/** |
|
35
|
|
|
* Line end character |
|
36
|
|
|
* @var string |
|
37
|
|
|
*/ |
|
38
|
|
|
public $line_end = "\n"; |
|
39
|
|
|
|
|
40
|
|
|
/** |
|
41
|
|
|
* Line break character |
|
42
|
|
|
* @var string |
|
43
|
|
|
*/ |
|
44
|
|
|
public $line_break = "\n\n"; |
|
45
|
|
|
|
|
46
|
|
|
/** |
|
47
|
|
|
* Wordwrap output, set to 0 to skip wrapping |
|
48
|
|
|
* @var int |
|
49
|
|
|
*/ |
|
50
|
|
|
public $body_width = 76; |
|
51
|
|
|
|
|
52
|
|
|
/** |
|
53
|
|
|
* Strip remaining tags, set to false to leave them in |
|
54
|
|
|
* @var boolean |
|
55
|
|
|
*/ |
|
56
|
|
|
public $strip_tags = true; |
|
57
|
|
|
|
|
58
|
|
|
/** |
|
59
|
|
|
* Regex to run on plain text to prevent markdown from erroneously converting |
|
60
|
|
|
* @var string[] |
|
61
|
|
|
*/ |
|
62
|
|
|
private $_textEscapeRegex = array(); |
|
63
|
|
|
|
|
64
|
|
|
/** |
|
65
|
|
|
* The passed html string to convert |
|
66
|
|
|
* @var string |
|
67
|
|
|
*/ |
|
68
|
|
|
public $html; |
|
69
|
|
|
|
|
70
|
|
|
/** |
|
71
|
|
|
* The markdown equivalent to the html string |
|
72
|
|
|
* @var string |
|
73
|
|
|
*/ |
|
74
|
|
|
public $markdown; |
|
75
|
|
|
|
|
76
|
|
|
/** |
|
77
|
|
|
* Gets everything started using the built in or external parser |
|
78
|
|
|
* |
|
79
|
|
|
* @param string $html string of html to convert to MD text |
|
80
|
|
|
*/ |
|
81
|
|
|
public function __construct($html) |
|
82
|
|
|
{ |
|
83
|
|
|
// Up front, remove whitespace between html tags |
|
84
|
|
|
$this->html = preg_replace('/(?:(?<=\>)|(?<=\/\>))(\s+)(?=\<\/?)/', '', $html); |
|
85
|
|
|
|
|
86
|
|
|
// The XML parser will not deal gracefully with these |
|
87
|
|
|
$this->html = strtr($this->html, array( |
|
88
|
|
|
'?<' => '|?|<', |
|
89
|
|
|
'?>' => '|?|>', |
|
90
|
|
|
'>?' => '>|?|', |
|
91
|
|
|
'<?' => '<|?|' |
|
92
|
|
|
)); |
|
93
|
|
|
|
|
94
|
|
|
// Set the dom parser to use and load the HTML to the parser |
|
95
|
|
|
$this->_set_parser(); |
|
96
|
|
|
|
|
97
|
|
|
// Initialize the regex array to escape text areas so markdown does |
|
98
|
|
|
// not interpret plain text as markdown syntax |
|
99
|
|
|
$this->_textEscapeRegex = array( |
|
|
|
|
|
|
100
|
|
|
// Things that may convert to an hr --- or - - - etc |
|
101
|
|
|
'([-*_])([ ]{0,2}\1){2,}' => '\\\\$0|', |
|
102
|
|
|
// or **stuff** => \*\*stuff\*\* |
|
103
|
|
|
'\*\*([^*\s]+)\*\*' => '\*\*$1\*\*', |
|
104
|
|
|
// or versions of *italic* __italic__ _italic_ |
|
105
|
|
|
'\*([^*\s]+)\*' => '\*$1\*', |
|
106
|
|
|
'__(?! |_)(.+)(?!<_| )__' => '\_\_$1\_\_', |
|
107
|
|
|
'_(?! |_)(.+)(?!<_| )_' => '\_$1\_', |
|
108
|
|
|
// nor `code` |
|
109
|
|
|
'`(.+)`' => '\`$1\`', |
|
110
|
|
|
// or links |
|
111
|
|
|
'\[(.+)\](\s*\()' => '\[$1\]$2', |
|
112
|
|
|
'\[(.+)\](\s*)\[(.*)\]' => '\[$1\]$2\[$3\]', |
|
113
|
|
|
); |
|
114
|
|
|
} |
|
115
|
|
|
|
|
116
|
|
|
/** |
|
117
|
|
|
* Set the DOM parser for class, loads the supplied HTML |
|
118
|
|
|
*/ |
|
119
|
|
|
private function _set_parser() |
|
120
|
|
|
{ |
|
121
|
|
|
// Using PHP built in functions ... |
|
122
|
|
|
if (class_exists('DOMDocument')) |
|
123
|
|
|
{ |
|
124
|
|
|
$this->_parser = true; |
|
125
|
|
|
$previous = libxml_use_internal_errors(true); |
|
126
|
|
|
|
|
127
|
|
|
// Set up basic parameters for DomDocument, including silencing structural errors |
|
128
|
|
|
$this->_setupDOMDocument(); |
|
129
|
|
|
|
|
130
|
|
|
// Set the error handle back to what it was, and flush |
|
131
|
|
|
libxml_use_internal_errors($previous); |
|
132
|
|
|
libxml_clear_errors(); |
|
133
|
|
|
} |
|
134
|
|
|
// Or using the external simple html parser |
|
135
|
|
|
else |
|
136
|
|
|
{ |
|
137
|
|
|
$this->_parser = false; |
|
138
|
|
|
require_once(EXTDIR . '/simple_html_dom.php'); |
|
139
|
|
|
$this->doc = str_get_html($this->html, true, true, 'UTF-8', false); |
|
140
|
|
|
} |
|
141
|
|
|
} |
|
142
|
|
|
|
|
143
|
|
|
/** |
|
144
|
|
|
* Loads the html body and sends it to the parsing loop to convert all |
|
145
|
|
|
* DOM nodes to markup |
|
146
|
|
|
*/ |
|
147
|
|
|
public function get_markdown() |
|
148
|
|
|
{ |
|
149
|
|
|
// For this html node, find all child elements and convert |
|
150
|
|
|
$body = $this->_getBody(); |
|
151
|
|
|
$this->_convert_childNodes($body); |
|
152
|
|
|
|
|
153
|
|
|
// Done replacing HTML elements, now get the converted DOM tree back into a string |
|
154
|
|
|
$this->markdown = ($this->_parser) ? $this->doc->saveHTML() : $this->doc->save(); |
|
155
|
|
|
|
|
156
|
|
|
// Using the internal DOM methods requires we need to do a little extra work |
|
157
|
|
|
if ($this->_parser) |
|
158
|
|
|
{ |
|
159
|
|
|
$this->markdown = html_entity_decode(htmlspecialchars_decode($this->markdown, ENT_QUOTES), ENT_QUOTES, 'UTF-8'); |
|
160
|
|
|
} |
|
161
|
|
|
|
|
162
|
|
|
// Clean up any excess spacing etc |
|
163
|
|
|
$this->_clean_markdown(); |
|
164
|
|
|
|
|
165
|
|
|
// Wordwrap? |
|
166
|
|
|
if (!empty($this->body_width)) |
|
167
|
|
|
$this->markdown = $this->_utf8_wordwrap($this->markdown, $this->body_width, $this->line_end); |
|
168
|
|
|
|
|
169
|
|
|
return $this->markdown; |
|
170
|
|
|
} |
|
171
|
|
|
|
|
172
|
|
|
/** |
|
173
|
|
|
* Returns just the body of the HTML, as best possible, so we are not dealing with head |
|
174
|
|
|
* and above head markup |
|
175
|
|
|
* |
|
176
|
|
|
* @return object |
|
177
|
|
|
*/ |
|
178
|
|
|
private function _getBody() |
|
179
|
|
|
{ |
|
180
|
|
|
// If there is a head node, then off with his head! |
|
181
|
|
|
$this->_clipHead(); |
|
182
|
|
|
|
|
183
|
|
|
// The body of the HTML is where its at. |
|
184
|
|
|
if ($this->_parser) |
|
185
|
|
|
{ |
|
186
|
|
|
$body = $this->doc->getElementsByTagName('body')->item(0); |
|
187
|
|
|
} |
|
188
|
|
|
else |
|
189
|
|
|
{ |
|
190
|
|
|
if ($this->doc->find('body', 0) !== null) |
|
191
|
|
|
{ |
|
192
|
|
|
$body = $this->doc->find('body', 0); |
|
193
|
|
|
} |
|
194
|
|
|
elseif ($this->doc->find('html', 0) !== null) |
|
195
|
|
|
{ |
|
196
|
|
|
$body = $this->doc->find('html', 0); |
|
197
|
|
|
} |
|
198
|
|
|
else |
|
199
|
|
|
{ |
|
200
|
|
|
$body = $this->doc->root; |
|
201
|
|
|
} |
|
202
|
|
|
} |
|
203
|
|
|
|
|
204
|
|
|
return $body; |
|
205
|
|
|
} |
|
206
|
|
|
|
|
207
|
|
|
/** |
|
208
|
|
|
* Remove any <head node from the DOM |
|
209
|
|
|
*/ |
|
210
|
|
|
private function _clipHead() |
|
211
|
|
|
{ |
|
212
|
|
|
$head = ($this->_parser) ? $this->doc->getElementsByTagName('head')->item(0) : $this->doc->find('head', 0)->outertext; |
|
213
|
|
|
if ($head !== null) |
|
214
|
|
|
{ |
|
215
|
|
|
if ($this->_parser) |
|
216
|
|
|
{ |
|
217
|
|
|
$head->parentNode->removeChild($head); |
|
218
|
|
|
} |
|
219
|
|
|
else |
|
220
|
|
|
{ |
|
221
|
|
|
$this->doc->find('head', 0)->outertext = ''; |
|
222
|
|
|
} |
|
223
|
|
|
} |
|
224
|
|
|
} |
|
225
|
|
|
|
|
226
|
|
|
/** |
|
227
|
|
|
* Sets up processing parameters for DOMDocument to ensure that text is processed as UTF-8 |
|
228
|
|
|
*/ |
|
229
|
|
|
private function _setupDOMDocument() |
|
230
|
|
|
{ |
|
231
|
|
|
// If the html is already wrapped, remove it |
|
232
|
|
|
$this->html = $this->_returnBodyText($this->html); |
|
233
|
|
|
|
|
234
|
|
|
// Set up processing details |
|
235
|
|
|
$this->doc = new DOMDocument(); |
|
236
|
|
|
$this->doc->preserveWhiteSpace = false; |
|
237
|
|
|
$this->doc->encoding = 'UTF-8'; |
|
238
|
|
|
|
|
239
|
|
|
// Do what we can to ensure this is processed as UTF-8 |
|
240
|
|
|
$this->doc->loadHTML('<?xml encoding="UTF-8"><html><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8"/></head><body>' . $this->html . '</body></html>'); |
|
241
|
|
|
} |
|
242
|
|
|
|
|
243
|
|
|
/** |
|
244
|
|
|
* Normalize any spacing and excess blank lines that may have been generated |
|
245
|
|
|
*/ |
|
246
|
|
|
private function _clean_markdown() |
|
247
|
|
|
{ |
|
248
|
|
|
// We only want the content, no wrappers |
|
249
|
|
|
$this->markdown = $this->_returnBodyText($this->markdown); |
|
250
|
|
|
|
|
251
|
|
|
// Remove non breakable spaces that may be hiding in here |
|
252
|
|
|
$this->markdown = str_replace("\xC2\xA0\x20", ' ', $this->markdown); |
|
253
|
|
|
$this->markdown = str_replace("\xC2\xA0", ' ', $this->markdown); |
|
254
|
|
|
|
|
255
|
|
|
// Remove any "bonus" tags |
|
256
|
|
|
if ($this->strip_tags) |
|
257
|
|
|
$this->markdown = strip_tags($this->markdown); |
|
258
|
|
|
|
|
259
|
|
|
// Replace content that we "hide" from the XML parsers |
|
260
|
|
|
$this->markdown = strtr($this->markdown, array( |
|
261
|
|
|
'|?|>' => '?>', |
|
262
|
|
|
'|?|<' => '?<', |
|
263
|
|
|
'<|?|' => '<?', |
|
264
|
|
|
'>|?|' => '>?' |
|
265
|
|
|
)); |
|
266
|
|
|
|
|
267
|
|
|
// Strip the chaff and any excess blank lines we may have produced |
|
268
|
|
|
$this->markdown = trim($this->markdown); |
|
269
|
|
|
$this->markdown = preg_replace("~(\n(\s)?){3,}~", "\n\n", $this->markdown); |
|
270
|
|
|
$this->markdown = preg_replace("~(^\s\s\n){3,}~m", " \n \n", $this->markdown); |
|
271
|
|
|
$this->markdown = preg_replace("~(^\s\s\r?\n){3,}~m", " \n \n", $this->markdown); |
|
272
|
|
|
$this->markdown = preg_replace("~(^\s\s(?:\r?\n){2}){3,}~m", " \n \n", $this->markdown); |
|
273
|
|
|
} |
|
274
|
|
|
|
|
275
|
|
|
/** |
|
276
|
|
|
* Looks for the text inside of <body> and then <html>, returning just the inner |
|
277
|
|
|
* |
|
278
|
|
|
* @param $text |
|
279
|
|
|
* |
|
280
|
|
|
* @return string |
|
281
|
|
|
*/ |
|
282
|
|
|
private function _returnBodyText($text) |
|
283
|
|
|
{ |
|
284
|
|
|
if (preg_match('~<body>(.*)</body>~su', $text, $body)) |
|
285
|
|
|
return $body[1]; |
|
286
|
|
|
elseif (preg_match('~<html>(.*)</html>~su', $text, $body)) |
|
287
|
|
|
return $body[1]; |
|
288
|
|
|
|
|
289
|
|
|
return $text; |
|
290
|
|
|
} |
|
291
|
|
|
|
|
292
|
|
|
/** |
|
293
|
|
|
* For a given node, checks if it is anywhere nested inside of a code block |
|
294
|
|
|
* - Prevents converting anything that's inside a code block |
|
295
|
|
|
* |
|
296
|
|
|
* @param object $node |
|
297
|
|
|
* @param boolean $parser flag for internal or external parser |
|
298
|
|
|
* |
|
299
|
|
|
* @return boolean |
|
300
|
|
|
*/ |
|
301
|
|
|
private static function _has_parent_code($node, $parser) |
|
302
|
|
|
{ |
|
303
|
|
|
$parent = $parser ? $node->parentNode : $node->parentNode(); |
|
304
|
|
|
while ($parent) |
|
305
|
|
|
{ |
|
306
|
|
|
if ($parent === null) |
|
307
|
|
|
return false; |
|
308
|
|
|
|
|
309
|
|
|
// Anywhere nested inside a code block we don't render tags |
|
310
|
|
|
$tag = $parser ? $parent->nodeName : $parent->nodeName(); |
|
311
|
|
|
if ($tag === 'code') |
|
312
|
|
|
return true; |
|
313
|
|
|
|
|
314
|
|
|
// Back out another level, until we are done |
|
315
|
|
|
$parent = $parser ? $parent->parentNode : $parent->parentNode(); |
|
316
|
|
|
} |
|
317
|
|
|
|
|
318
|
|
|
return false; |
|
319
|
|
|
} |
|
320
|
|
|
|
|
321
|
|
|
/** |
|
322
|
|
|
* Get the nesting level when inside a list |
|
323
|
|
|
* |
|
324
|
|
|
* @param object $node |
|
325
|
|
|
* @param boolean $parser flag for internal or external parser |
|
326
|
|
|
* |
|
327
|
|
|
* @return int |
|
328
|
|
|
*/ |
|
329
|
|
|
private static function _has_parent_list($node, $parser) |
|
330
|
|
|
{ |
|
331
|
|
|
$inlist = array('ul', 'ol'); |
|
332
|
|
|
$depth = 0; |
|
333
|
|
|
|
|
334
|
|
|
$parent = $parser ? $node->parentNode : $node->parentNode(); |
|
335
|
|
|
while ($parent) |
|
336
|
|
|
{ |
|
337
|
|
|
// Anywhere nested inside a list we need to get the depth |
|
338
|
|
|
$tag = $parser ? $parent->nodeName : $parent->nodeName(); |
|
339
|
|
|
if (in_array($tag, $inlist)) |
|
340
|
|
|
$depth++; |
|
341
|
|
|
|
|
342
|
|
|
// Back out another level |
|
343
|
|
|
$parent = $parser ? $parent->parentNode : $parent->parentNode(); |
|
344
|
|
|
} |
|
345
|
|
|
|
|
346
|
|
|
return $depth; |
|
347
|
|
|
} |
|
348
|
|
|
|
|
349
|
|
|
/** |
|
350
|
|
|
* Traverse each node to its base, then convert tags to markup on the way back out |
|
351
|
|
|
* |
|
352
|
|
|
* @param object $node |
|
353
|
|
|
*/ |
|
354
|
|
|
private function _convert_childNodes($node) |
|
355
|
|
|
{ |
|
356
|
|
|
if (self::_has_parent_code($node, $this->_parser)) |
|
357
|
|
|
return; |
|
358
|
|
|
|
|
359
|
|
|
// Keep traversing till we are at the base of this node |
|
360
|
|
|
if ($node->hasChildNodes()) |
|
361
|
|
|
{ |
|
362
|
|
|
$num = $this->_parser ? $node->childNodes->length : count($node->childNodes()); |
|
363
|
|
|
for ($i = 0; $i < $num; $i++) |
|
364
|
|
|
{ |
|
365
|
|
|
$child = $this->_parser ? $node->childNodes->item($i) : $node->childNodes($i); |
|
366
|
|
|
$this->_convert_childNodes($child); |
|
367
|
|
|
} |
|
368
|
|
|
} |
|
369
|
|
|
|
|
370
|
|
|
// At the root of this node, convert it to markdown |
|
371
|
|
|
$this->_convert_to_markdown($node); |
|
372
|
|
|
} |
|
373
|
|
|
|
|
374
|
|
|
/** |
|
375
|
|
|
* Convert the supplied node into its markdown equivalent |
|
376
|
|
|
* - Supports *some* markdown extra tags, namely: table, abbr & dl in a limited fashion |
|
377
|
|
|
* |
|
378
|
|
|
* @param object $node |
|
379
|
|
|
*/ |
|
380
|
|
|
private function _convert_to_markdown($node) |
|
381
|
|
|
{ |
|
382
|
|
|
// HTML tag we are dealing with |
|
383
|
|
|
$tag = $this->_get_name($node); |
|
384
|
|
|
|
|
385
|
|
|
// Based on the tag, determine how to convert |
|
386
|
|
|
switch ($tag) |
|
387
|
|
|
{ |
|
388
|
|
|
case 'a': |
|
389
|
|
|
$markdown = $this->line_end . $this->_convert_anchor($node); |
|
390
|
|
|
break; |
|
391
|
|
|
case 'abbr': |
|
392
|
|
|
$markdown = $this->_convert_abbr($node); |
|
393
|
|
|
break; |
|
394
|
|
|
case 'b': |
|
395
|
|
|
case 'strong': |
|
396
|
|
|
$markdown = '**' . $this->_get_value($node) . '**'; |
|
397
|
|
|
break; |
|
398
|
|
|
case 'blockquote': |
|
399
|
|
|
$markdown = $this->_convert_blockquote($node); |
|
400
|
|
|
break; |
|
401
|
|
|
case 'br': |
|
402
|
|
|
// DomDocument strips empty lines, this prevents that |
|
403
|
|
|
$markdown = "\xC2\xA0\xC2\xA0" . $this->line_break; |
|
404
|
|
|
break; |
|
405
|
|
|
case 'center': |
|
406
|
|
|
$markdown = $this->line_end . $this->_get_value($node) . $this->line_end; |
|
407
|
|
|
break; |
|
408
|
|
|
case 'code': |
|
409
|
|
|
$markdown = $this->_convert_code($node); |
|
410
|
|
|
break; |
|
411
|
|
|
case 'dt': |
|
412
|
|
|
$markdown = str_replace(array("\n", "\r", "\n\r"), '', $this->_get_value($node)) . $this->line_end; |
|
413
|
|
|
break; |
|
414
|
|
|
case 'dd': |
|
415
|
|
|
$markdown = ': ' . $this->_get_value($node) . $this->line_break; |
|
416
|
|
|
break; |
|
417
|
|
|
case 'dl': |
|
418
|
|
|
$markdown = trim($this->_get_value($node)) . $this->line_break; |
|
419
|
|
|
break; |
|
420
|
|
|
case 'em': |
|
421
|
|
|
case 'i': |
|
422
|
|
|
$markdown = '_' . $this->_get_value($node) . '_'; |
|
423
|
|
|
break; |
|
424
|
|
|
case 'hr': |
|
425
|
|
|
$markdown = $this->line_end . str_repeat('-', 3) . $this->line_end; |
|
426
|
|
|
break; |
|
427
|
|
|
case 'h1': |
|
428
|
|
|
case 'h2': |
|
429
|
|
|
case 'h3': |
|
430
|
|
|
case 'h4': |
|
431
|
|
|
case 'h5': |
|
432
|
|
|
case 'h6': |
|
433
|
|
|
$markdown = $this->_convert_header($tag, $this->_get_value($node)); |
|
434
|
|
|
break; |
|
435
|
|
|
case 'img': |
|
436
|
|
|
$markdown = $this->_convert_image($node); |
|
437
|
|
|
break; |
|
438
|
|
|
case 'ol': |
|
439
|
|
|
case 'ul': |
|
440
|
|
|
$markdown = rtrim($this->_get_value($node)) . $this->line_break; |
|
441
|
|
|
break; |
|
442
|
|
|
case 'li': |
|
443
|
|
|
$markdown = $this->_convert_list($node); |
|
444
|
|
|
break; |
|
445
|
|
|
case 'p': |
|
446
|
|
|
if (!$node->hasChildNodes()) |
|
447
|
|
|
{ |
|
448
|
|
|
$markdown = str_replace("\n", ' ', $this->_get_value($node)) . $this->line_break; |
|
449
|
|
|
$markdown = $this->_escape_text($markdown); |
|
450
|
|
|
} |
|
451
|
|
|
else |
|
452
|
|
|
$markdown = rtrim($this->_get_value($node)) . $this->line_break; |
|
453
|
|
|
break; |
|
454
|
|
|
case 'pre': |
|
455
|
|
|
$markdown = $this->_get_value($node) . $this->line_break; |
|
456
|
|
|
break; |
|
457
|
|
|
case 'div': |
|
458
|
|
|
$markdown = $this->line_end . $this->_get_value($node) . $this->line_end; |
|
459
|
|
|
if (!$node->hasChildNodes()) |
|
460
|
|
|
$markdown = $this->_escape_text($markdown); |
|
461
|
|
|
break; |
|
462
|
|
|
//case '#text': |
|
463
|
|
|
// $markdown = $this->_escape_text($this->_get_value($node)); |
|
464
|
|
|
// break; |
|
465
|
|
|
case 'title': |
|
466
|
|
|
$markdown = '# ' . $this->_get_value($node) . $this->line_break; |
|
467
|
|
|
break; |
|
468
|
|
|
case 'table': |
|
469
|
|
|
$markdown = $this->_convert_table($node) . $this->line_break; |
|
470
|
|
|
break; |
|
471
|
|
|
case 'th': |
|
472
|
|
|
case 'tr': |
|
473
|
|
|
case 'td': |
|
474
|
|
|
case 'tbody': |
|
475
|
|
|
case 'tfoot': |
|
476
|
|
|
case 'thead': |
|
477
|
|
|
// Just skip over these as we handle them in the table tag itself |
|
478
|
|
|
$markdown = '~`skip`~'; |
|
479
|
|
|
break; |
|
480
|
|
|
case 'root': |
|
481
|
|
|
case 'span': |
|
482
|
|
|
case 'body': |
|
483
|
|
|
// Remove these tags and simply replace with the text inside the tags |
|
484
|
|
|
$markdown = $this->_get_innerHTML($node); |
|
485
|
|
|
break; |
|
486
|
|
|
default: |
|
487
|
|
|
// Don't know you or text, so just preserve whats there |
|
488
|
|
|
$markdown = $this->_get_outerHTML($node); |
|
489
|
|
|
} |
|
490
|
|
|
|
|
491
|
|
|
// Replace the node with our markdown replacement, or with the node itself if none was found |
|
492
|
|
|
if ($markdown !== '~`skip`~') |
|
493
|
|
|
{ |
|
494
|
|
|
if ($this->_parser) |
|
495
|
|
|
{ |
|
496
|
|
|
// Create a new text node with our markdown tag and replace the original node |
|
497
|
|
|
$markdown_node = $this->doc->createTextNode($markdown); |
|
498
|
|
|
$node->parentNode->replaceChild($markdown_node, $node); |
|
499
|
|
|
} |
|
500
|
|
|
else |
|
501
|
|
|
$node->outertext = $markdown; |
|
502
|
|
|
} |
|
503
|
|
|
} |
|
504
|
|
|
|
|
505
|
|
|
/** |
|
506
|
|
|
* Converts <abbr> tags to markdown (extra) |
|
507
|
|
|
* |
|
508
|
|
|
* html: <abbr title="Hyper Text Markup Language">HTML</abbr> |
|
509
|
|
|
* md: *[HTML]: Hyper Text Markup Language |
|
510
|
|
|
* |
|
511
|
|
|
* @param object $node |
|
512
|
|
|
* @return string |
|
513
|
|
|
*/ |
|
514
|
|
|
private function _convert_abbr($node) |
|
515
|
|
|
{ |
|
516
|
|
|
$title = $node->getAttribute('title'); |
|
517
|
|
|
$value = $this->_get_value($node); |
|
518
|
|
|
|
|
519
|
|
|
if (!empty($title)) |
|
520
|
|
|
$markdown = '*[' . $value . ']: ' . $title . $this->line_break; |
|
521
|
|
|
else |
|
522
|
|
|
$markdown = ''; |
|
523
|
|
|
|
|
524
|
|
|
return $markdown; |
|
525
|
|
|
} |
|
526
|
|
|
|
|
527
|
|
|
/** |
|
528
|
|
|
* Converts <a> tags to markdown |
|
529
|
|
|
* |
|
530
|
|
|
* html: <a href='http://somesite.com' title='Title'>Awesome Site</a> |
|
531
|
|
|
* md: [Awesome Site](http://somesite.com 'Title') |
|
532
|
|
|
* |
|
533
|
|
|
* @param object $node |
|
534
|
|
|
* @return string |
|
535
|
|
|
*/ |
|
536
|
|
|
private function _convert_anchor($node) |
|
537
|
|
|
{ |
|
538
|
|
|
global $txt; |
|
539
|
|
|
|
|
540
|
|
|
$href = htmlentities($node->getAttribute('href'), ENT_COMPAT, 'UTF-8', false); |
|
541
|
|
|
$title = $node->getAttribute('title'); |
|
542
|
|
|
$class = $node->getAttribute('class'); |
|
543
|
|
|
$value = $this->_get_value($node); |
|
544
|
|
|
|
|
545
|
|
|
// Provide a more compact [name] if none is given |
|
546
|
|
|
if ($value == $node->getAttribute('href') || empty($value)) |
|
547
|
|
|
$value = empty($title) ? $txt['link'] : $title; |
|
548
|
|
|
|
|
549
|
|
|
// Special processing just for our own footnotes |
|
550
|
|
|
if ($class === 'target' || $class === 'footnote_return') |
|
551
|
|
|
$markdown = $value; |
|
552
|
|
|
elseif (!empty($title)) |
|
553
|
|
|
$markdown = '[' . $value . '](' . $href . ' "' . $title . '")'; |
|
554
|
|
|
else |
|
555
|
|
|
$markdown = '[' . $value . '](' . $href . ')'; |
|
556
|
|
|
|
|
557
|
|
|
// Some links can be very long and if we wrap them they break |
|
558
|
|
|
$this->_check_link_lenght($markdown); |
|
559
|
|
|
|
|
560
|
|
|
return $markdown; |
|
561
|
|
|
} |
|
562
|
|
|
|
|
563
|
|
|
/** |
|
564
|
|
|
* Converts blockquotes to markdown > quote style |
|
565
|
|
|
* |
|
566
|
|
|
* html: <blockquote>quote</blockquote> |
|
567
|
|
|
* md: > quote |
|
568
|
|
|
* |
|
569
|
|
|
* @param object $node |
|
570
|
|
|
* @return string |
|
571
|
|
|
*/ |
|
572
|
|
|
private function _convert_blockquote($node) |
|
573
|
|
|
{ |
|
574
|
|
|
$markdown = ''; |
|
575
|
|
|
|
|
576
|
|
|
// All the contents of this block quote |
|
577
|
|
|
$value = $this->_get_value($node); |
|
578
|
|
|
$value = trim($value); |
|
579
|
|
|
|
|
580
|
|
|
// Go line by line |
|
581
|
|
|
$lines = preg_split('~\r\n|\r|\n~', $value); |
|
582
|
|
|
|
|
583
|
|
|
// Each line gets a '> ' in front of it, just like email quotes really |
|
584
|
|
|
foreach ($lines as $line) |
|
585
|
|
|
$markdown .= '> ' . ltrim($line, "\t") . $this->line_end; |
|
586
|
|
|
|
|
587
|
|
|
$markdown .= $this->line_end; |
|
588
|
|
|
|
|
589
|
|
|
return $markdown; |
|
590
|
|
|
} |
|
591
|
|
|
|
|
592
|
|
|
/** |
|
593
|
|
|
* Converts code tags to markdown span `code` or block code |
|
594
|
|
|
* Converts single line code to inline tick mark |
|
595
|
|
|
* Converts multi line to 4 space indented code |
|
596
|
|
|
* |
|
597
|
|
|
* html: <code>code</code> |
|
598
|
|
|
* md: `code` |
|
599
|
|
|
* |
|
600
|
|
|
* @param object $node |
|
601
|
|
|
* @return string |
|
602
|
|
|
*/ |
|
603
|
|
|
private function _convert_code($node) |
|
604
|
|
|
{ |
|
605
|
|
|
$value = $this->_get_innerHTML($node); |
|
606
|
|
|
|
|
607
|
|
|
// If we have a multi line code block, we are working outside to in, and need to convert the br's ourselves |
|
608
|
|
|
$value = preg_replace('~<br( /)?' . '>~', "\n", str_replace(' ', ' ', $value)); |
|
609
|
|
|
|
|
610
|
|
|
// If there are html tags in this code block, we need to disable strip tags |
|
611
|
|
|
// This is NOT the ideal way to handle this, needs something along the lines of preparse and unpreparse. |
|
612
|
|
|
if ($this->strip_tags && preg_match('~<[^<]+>~', $value)) |
|
613
|
|
|
$this->strip_tags = false; |
|
614
|
|
|
|
|
615
|
|
|
// Get the number of lines of code that we have |
|
616
|
|
|
$lines = preg_split('~\r\n|\r|\n~', $value); |
|
617
|
|
|
$total = count($lines); |
|
618
|
|
|
|
|
619
|
|
|
// If there's more than one line of code, use leading four space syntax |
|
620
|
|
|
if ($total > 1) |
|
621
|
|
|
{ |
|
622
|
|
|
$first_line = trim($lines[0]); |
|
623
|
|
|
$last_line = trim($lines[$total - 1]); |
|
624
|
|
|
|
|
625
|
|
|
// Remove any leading and trailing blank lines |
|
626
|
|
|
if (empty($first_line)) |
|
627
|
|
|
array_shift($lines); |
|
628
|
|
|
if (empty($last_line)) |
|
629
|
|
|
array_pop($lines); |
|
630
|
|
|
|
|
631
|
|
|
// Convert what remains |
|
632
|
|
|
$markdown = ''; |
|
633
|
|
|
foreach ($lines as $line) |
|
634
|
|
|
{ |
|
635
|
|
|
// Adjust the word wrapping since this has code tags, leave it up to |
|
636
|
|
|
// the email client to mess these up ;) |
|
637
|
|
|
$this->_check_link_lenght($markdown, 5); |
|
638
|
|
|
|
|
639
|
|
|
$markdown .= str_repeat(' ', 4) . $line . $this->line_end; |
|
640
|
|
|
} |
|
641
|
|
|
|
|
642
|
|
|
// The parser will encode, but we don't want that for our code block |
|
643
|
|
|
if ($this->_parser) |
|
644
|
|
|
$markdown = html_entity_decode($markdown, ENT_QUOTES, 'UTF-8'); |
|
645
|
|
|
} |
|
646
|
|
|
// Single line, back tick and move on |
|
647
|
|
|
else |
|
648
|
|
|
{ |
|
649
|
|
|
// Account for backticks in the single line code itself |
|
650
|
|
|
$ticks = $this->_has_ticks($node, $value); |
|
651
|
|
|
if (!empty($ticks)) |
|
652
|
|
|
{ |
|
653
|
|
|
// If the ticks were at the start/end of the word space it off |
|
654
|
|
|
if ($lines[0][0] == '`' || substr($lines[0], -1) == '`') |
|
655
|
|
|
$lines[0] = ' ' . $lines[0] . ' '; |
|
656
|
|
|
|
|
657
|
|
|
$markdown = $ticks . ($this->_parser ? html_entity_decode($lines[0], ENT_QUOTES, 'UTF-8') : $lines[0]) . $ticks; |
|
658
|
|
|
} |
|
659
|
|
|
else |
|
660
|
|
|
$markdown = '`' . ($this->_parser ? html_entity_decode($lines[0], ENT_QUOTES, 'UTF-8') : $lines[0]) . '`'; |
|
661
|
|
|
} |
|
662
|
|
|
|
|
663
|
|
|
return $markdown; |
|
664
|
|
|
} |
|
665
|
|
|
|
|
666
|
|
|
/** |
|
667
|
|
|
* Converts <h1> and <h2> headers to markdown-style headers in setex style, |
|
668
|
|
|
* all other headers are returned as atx style ### h3 |
|
669
|
|
|
* |
|
670
|
|
|
* html: <h1>header</h1> |
|
671
|
|
|
* md: header |
|
672
|
|
|
* ====== |
|
673
|
|
|
* |
|
674
|
|
|
* html: <h3>header</h3> |
|
675
|
|
|
* md: ###header |
|
676
|
|
|
* |
|
677
|
|
|
* @param int $level |
|
678
|
|
|
* @param string $content |
|
679
|
|
|
* @return string |
|
680
|
|
|
*/ |
|
681
|
|
|
private function _convert_header($level, $content) |
|
682
|
|
|
{ |
|
683
|
|
|
$level = (int) ltrim($level, 'h'); |
|
684
|
|
|
|
|
685
|
|
|
if ($level < 3) |
|
686
|
|
|
{ |
|
687
|
|
|
$length = Util::strlen($content); |
|
688
|
|
|
$underline = ($level === 1) ? '=' : '-'; |
|
689
|
|
|
$markdown = $content . $this->line_end . str_repeat($underline, $length) . $this->line_break; |
|
690
|
|
|
} |
|
691
|
|
|
else |
|
692
|
|
|
$markdown = str_repeat('#', $level) . ' ' . $content . $this->line_break; |
|
693
|
|
|
|
|
694
|
|
|
return $markdown; |
|
695
|
|
|
} |
|
696
|
|
|
|
|
697
|
|
|
/** |
|
698
|
|
|
* Converts <img> tags to markdown |
|
699
|
|
|
* |
|
700
|
|
|
* html: <img src='source' alt='alt' title='title' /> |
|
701
|
|
|
* md:  |
|
702
|
|
|
* |
|
703
|
|
|
* @param object $node |
|
704
|
|
|
* @return string |
|
705
|
|
|
*/ |
|
706
|
|
|
private function _convert_image($node) |
|
707
|
|
|
{ |
|
708
|
|
|
$src = $node->getAttribute('src'); |
|
709
|
|
|
$alt = $node->getAttribute('alt'); |
|
710
|
|
|
$title = $node->getAttribute('title'); |
|
711
|
|
|
|
|
712
|
|
|
if (!empty($title)) |
|
713
|
|
|
$markdown = ''; |
|
714
|
|
|
else |
|
715
|
|
|
$markdown = ''; |
|
716
|
|
|
|
|
717
|
|
|
// Adjust width if needed to maintain the image |
|
718
|
|
|
$this->_check_link_lenght($markdown); |
|
719
|
|
|
|
|
720
|
|
|
return $markdown; |
|
721
|
|
|
} |
|
722
|
|
|
|
|
723
|
|
|
/** |
|
724
|
|
|
* Converts ordered <ol> and unordered <ul> lists to markdown syntax |
|
725
|
|
|
* |
|
726
|
|
|
* html: <ul><li>one</li></ul> |
|
727
|
|
|
* md * one |
|
728
|
|
|
* |
|
729
|
|
|
* @param object $node |
|
730
|
|
|
* @return string |
|
731
|
|
|
*/ |
|
732
|
|
|
private function _convert_list($node) |
|
733
|
|
|
{ |
|
734
|
|
|
$list_type = $this->_parser ? $node->parentNode->nodeName : $node->parentNode()->nodeName(); |
|
735
|
|
|
$value = $this->_get_value($node); |
|
736
|
|
|
|
|
737
|
|
|
$loose = rtrim($value) !== $value; |
|
738
|
|
|
$depth = max(0, $this->_has_parent_list($node, $this->_parser) - 1); |
|
739
|
|
|
|
|
740
|
|
|
// Unordered lists get a simple bullet |
|
741
|
|
|
if ($list_type === 'ul') |
|
742
|
|
|
$markdown = str_repeat("\t", $depth) . '* ' . $value; |
|
743
|
|
|
// Ordered lists need a number |
|
744
|
|
|
else |
|
745
|
|
|
{ |
|
746
|
|
|
$number = $this->_get_list_position($node); |
|
747
|
|
|
$markdown = str_repeat("\t", $depth) . $number . '. ' . $value; |
|
748
|
|
|
} |
|
749
|
|
|
|
|
750
|
|
|
return $markdown . (!$loose ? $this->line_end : ''); |
|
751
|
|
|
} |
|
752
|
|
|
|
|
753
|
|
|
/** |
|
754
|
|
|
* Converts tables tags to markdown extra table syntax |
|
755
|
|
|
* |
|
756
|
|
|
* - Have to build top down vs normal inside out due to needing col numbers and widths |
|
757
|
|
|
* |
|
758
|
|
|
* @param object $node |
|
759
|
|
|
* @return string |
|
760
|
|
|
*/ |
|
761
|
|
|
private function _convert_table($node) |
|
762
|
|
|
{ |
|
763
|
|
|
$table_heading = $node->getElementsByTagName('th'); |
|
764
|
|
|
if ($this->_get_item($table_heading, 0) === null) |
|
765
|
|
|
return ''; |
|
766
|
|
|
|
|
767
|
|
|
$th_parent = ($table_heading) ? ($this->_parser ? $this->_get_item($table_heading, 0)->parentNode->nodeName : $this->_get_item($table_heading, 0)->parentNode()->nodeName()) : false; |
|
768
|
|
|
|
|
769
|
|
|
// Set up for a markdown table, then storm the castle |
|
770
|
|
|
$align = array(); |
|
771
|
|
|
$value = array(); |
|
772
|
|
|
$width = array(); |
|
773
|
|
|
$max = array(); |
|
774
|
|
|
$header = array(); |
|
775
|
|
|
$rows = array(); |
|
776
|
|
|
|
|
777
|
|
|
// We only markdown well formed tables ... |
|
778
|
|
|
if ($table_heading && $th_parent === 'tr') |
|
779
|
|
|
{ |
|
780
|
|
|
// Find out how many columns we are dealing with |
|
781
|
|
|
$th_num = $this->_get_length($table_heading); |
|
782
|
|
|
|
|
783
|
|
|
for ($col = 0; $col < $th_num; $col++) |
|
784
|
|
|
{ |
|
785
|
|
|
// Get the align and text for each th (html5 this is no longer valid) |
|
786
|
|
|
$th = $this->_get_item($table_heading, $col); |
|
787
|
|
|
$align_value = ($th !== null) ? strtolower($th->getAttribute('align')) : false; |
|
788
|
|
|
$align[0][$col] = $align_value === false ? 'left' : $align_value; |
|
789
|
|
|
$value[0][$col] = $this->_get_value($th); |
|
790
|
|
|
$width[0][$col] = Util::strlen($this->_get_value($th)); |
|
791
|
|
|
|
|
792
|
|
|
// Seed the max col width |
|
793
|
|
|
$max[$col] = $width[0][$col]; |
|
794
|
|
|
} |
|
795
|
|
|
|
|
796
|
|
|
// Get all of the rows |
|
797
|
|
|
$table_rows = $node->getElementsByTagName('tr'); |
|
798
|
|
|
$num_rows = $this->_get_length($table_rows); |
|
799
|
|
|
for ($row = 1; $row < $num_rows; $row++) |
|
800
|
|
|
{ |
|
801
|
|
|
// Start at row 1 and get all of the td's in this row |
|
802
|
|
|
$row_data = $this->_get_item($table_rows, $row)->getElementsByTagName('td'); |
|
803
|
|
|
|
|
804
|
|
|
// Simply use the th count as the number of columns, if its not right its not markdown-able anyway |
|
805
|
|
|
for ($col = 0; $col < $th_num; $col++) |
|
806
|
|
|
{ |
|
807
|
|
|
// Get the align and text for each td in this row |
|
808
|
|
|
$td = $this->_get_item($row_data, $col); |
|
809
|
|
|
$align_value = ($td !== null) ? strtolower($td->getAttribute('align')) : false; |
|
810
|
|
|
$align[$row][$col] = $align_value === false ? 'left' : $align_value; |
|
811
|
|
|
$value[$row][$col] = $this->_get_value($td); |
|
812
|
|
|
$width[$row][$col] = Util::strlen($this->_get_value($td)); |
|
813
|
|
|
|
|
814
|
|
|
// Keep track of the longest col cell as we go |
|
815
|
|
|
if ($width[$row][$col] > $max[$col]) |
|
816
|
|
|
$max[$col] = $width[$row][$col]; |
|
817
|
|
|
} |
|
818
|
|
|
} |
|
819
|
|
|
|
|
820
|
|
|
// Done collecting data, we can rebuild it, we can make it better than it was. Better...stronger...faster |
|
821
|
|
|
for ($row = 0; $row < $num_rows; $row++) |
|
822
|
|
|
{ |
|
823
|
|
|
$temp = array(); |
|
824
|
|
|
for ($col = 0; $col < $th_num; $col++) |
|
825
|
|
|
{ |
|
826
|
|
|
// Build the header row once |
|
827
|
|
|
if ($row === 0) |
|
828
|
|
|
$header[] = str_repeat('-', $max[$col]); |
|
829
|
|
|
|
|
830
|
|
|
// Build the data for each col, align/pad as needed |
|
831
|
|
|
$temp[] = $this->_align_row_content($align[$row][$col], $width[$row][$col], $value[$row][$col], $max[$col]); |
|
832
|
|
|
} |
|
833
|
|
|
|
|
834
|
|
|
// Join it all up so we have a nice looking row |
|
835
|
|
|
$rows[] = '| ' . implode(' | ', $temp) . ' |'; |
|
836
|
|
|
|
|
837
|
|
|
// Stuff in the header after the th row |
|
838
|
|
|
if ($row === 0) |
|
839
|
|
|
$rows[] = '| ' . implode(' | ', $header) . ' | '; |
|
840
|
|
|
} |
|
841
|
|
|
|
|
842
|
|
|
// Adjust the word wrapping since this has a table, will get mussed by email anyway |
|
843
|
|
|
$this->_check_link_lenght($rows[1], 2); |
|
844
|
|
|
|
|
845
|
|
|
// Return what we did so it can be swapped in |
|
846
|
|
|
return implode($this->line_end, $rows); |
|
847
|
|
|
} |
|
848
|
|
|
} |
|
849
|
|
|
|
|
850
|
|
|
/** |
|
851
|
|
|
* Helper function for getting a node object |
|
852
|
|
|
* |
|
853
|
|
|
* @param object $node |
|
854
|
|
|
* @param int $item |
|
855
|
|
|
*/ |
|
856
|
|
|
private function _get_item($node, $item) |
|
857
|
|
|
{ |
|
858
|
|
|
if ($this->_parser) |
|
859
|
|
|
return $node->item($item); |
|
860
|
|
|
else |
|
861
|
|
|
return $node[$item]; |
|
862
|
|
|
} |
|
863
|
|
|
|
|
864
|
|
|
/** |
|
865
|
|
|
* Helper function for getting a node length |
|
866
|
|
|
* |
|
867
|
|
|
* @param object|array $node |
|
868
|
|
|
* @return int |
|
869
|
|
|
*/ |
|
870
|
|
|
private function _get_length($node) |
|
871
|
|
|
{ |
|
872
|
|
|
if ($this->_parser) |
|
873
|
|
|
return $node->length; |
|
874
|
|
|
else |
|
875
|
|
|
return count($node); |
|
876
|
|
|
} |
|
877
|
|
|
|
|
878
|
|
|
/** |
|
879
|
|
|
* Helper function for getting a node value |
|
880
|
|
|
* |
|
881
|
|
|
* @param object $node |
|
882
|
|
|
* @return string |
|
883
|
|
|
*/ |
|
884
|
|
|
private function _get_value($node) |
|
885
|
|
|
{ |
|
886
|
|
|
if ($node === null) |
|
887
|
|
|
return ''; |
|
888
|
|
|
|
|
889
|
|
|
if ($this->_parser) |
|
890
|
|
|
return $node->nodeValue; |
|
891
|
|
|
else |
|
892
|
|
|
return html_entity_decode(htmlspecialchars_decode($node->innertext, ENT_QUOTES), ENT_QUOTES, 'UTF-8'); |
|
893
|
|
|
} |
|
894
|
|
|
|
|
895
|
|
|
/** |
|
896
|
|
|
* Helper function for getting a node name |
|
897
|
|
|
* |
|
898
|
|
|
* @param object $node |
|
899
|
|
|
* @return string |
|
900
|
|
|
*/ |
|
901
|
|
|
private function _get_name($node) |
|
902
|
|
|
{ |
|
903
|
|
|
if ($node === null) |
|
904
|
|
|
return ''; |
|
905
|
|
|
|
|
906
|
|
|
if ($this->_parser) |
|
907
|
|
|
return $node->nodeName; |
|
908
|
|
|
else |
|
909
|
|
|
return $node->nodeName(); |
|
910
|
|
|
} |
|
911
|
|
|
|
|
912
|
|
|
/** |
|
913
|
|
|
* Helper function for creating ol's |
|
914
|
|
|
* |
|
915
|
|
|
* - Returns the absolute number of an <li> inside an <ol> |
|
916
|
|
|
* |
|
917
|
|
|
* @param object $node |
|
918
|
|
|
* @return int |
|
919
|
|
|
*/ |
|
920
|
|
|
private function _get_list_position($node) |
|
921
|
|
|
{ |
|
922
|
|
|
$position = 1; |
|
923
|
|
|
|
|
924
|
|
|
// Get all of the list nodes inside this parent |
|
925
|
|
|
$list_node = $this->_parser ? $node->parentNode : $node->parentNode(); |
|
926
|
|
|
$total_nodes = $this->_parser ? $node->parentNode->childNodes->length : count($list_node->childNodes()); |
|
927
|
|
|
|
|
928
|
|
|
// Loop through all li nodes and find where we are in this list |
|
929
|
|
|
for ($i = 0; $i < $total_nodes; $i++) |
|
930
|
|
|
{ |
|
931
|
|
|
$current_node = $this->_parser ? $list_node->childNodes->item($i) : $list_node->childNodes($i); |
|
932
|
|
|
if ($current_node === $node) |
|
933
|
|
|
$position = $i + 1; |
|
934
|
|
|
} |
|
935
|
|
|
|
|
936
|
|
|
return $position; |
|
937
|
|
|
} |
|
938
|
|
|
|
|
939
|
|
|
/** |
|
940
|
|
|
* Helper function for table creation |
|
941
|
|
|
* |
|
942
|
|
|
* - Builds td's to a give width, aligned as needed |
|
943
|
|
|
* |
|
944
|
|
|
* @param string $align |
|
945
|
|
|
* @param int $width |
|
946
|
|
|
* @param string $content |
|
947
|
|
|
* @param int $max |
|
948
|
|
|
* @return string |
|
949
|
|
|
*/ |
|
950
|
|
|
private function _align_row_content($align, $width, $content, $max) |
|
951
|
|
|
{ |
|
952
|
|
|
switch ($align) |
|
953
|
|
|
{ |
|
954
|
|
|
default: |
|
955
|
|
|
case 'left': |
|
956
|
|
|
$content .= str_repeat(' ', $max - $width); |
|
957
|
|
|
break; |
|
958
|
|
|
case 'right': |
|
959
|
|
|
$content = str_repeat(' ', $max - $width) . $content; |
|
960
|
|
|
break; |
|
961
|
|
|
case 'center': |
|
962
|
|
|
$paddingNeeded = $max - $width; |
|
963
|
|
|
$left = floor($paddingNeeded / 2); |
|
964
|
|
|
$right = $paddingNeeded - $left; |
|
965
|
|
|
$content = str_repeat(' ', $left) . $content . str_repeat(' ', $right); |
|
966
|
|
|
break; |
|
967
|
|
|
} |
|
968
|
|
|
|
|
969
|
|
|
return $content; |
|
970
|
|
|
} |
|
971
|
|
|
|
|
972
|
|
|
/** |
|
973
|
|
|
* Gets the inner html of a node |
|
974
|
|
|
* |
|
975
|
|
|
* @param DOMNode|object $node |
|
976
|
|
|
* @return string |
|
977
|
|
|
*/ |
|
978
|
|
|
private function _get_innerHTML($node) |
|
979
|
|
|
{ |
|
980
|
|
|
if ($this->_parser) |
|
981
|
|
|
{ |
|
982
|
|
|
$doc = new DOMDocument(); |
|
983
|
|
|
$doc->appendChild($doc->importNode($node, true)); |
|
984
|
|
|
$html = trim($doc->saveHTML()); |
|
985
|
|
|
$tag = $node->nodeName; |
|
986
|
|
|
|
|
987
|
|
|
return preg_replace('@^<' . $tag . '[^>]*>|</' . $tag . '>$@', '', $html); |
|
988
|
|
|
} |
|
989
|
|
|
else |
|
990
|
|
|
return $node->innertext; |
|
991
|
|
|
} |
|
992
|
|
|
|
|
993
|
|
|
/** |
|
994
|
|
|
* Gets the outer html of a node |
|
995
|
|
|
* |
|
996
|
|
|
* @param DOMNode|object $node |
|
997
|
|
|
* @return string |
|
998
|
|
|
*/ |
|
999
|
|
|
private function _get_outerHTML($node) |
|
1000
|
|
|
{ |
|
1001
|
|
|
if ($this->_parser) |
|
1002
|
|
|
{ |
|
1003
|
|
|
if (version_compare(PHP_VERSION, '5.3.6') >= 0) |
|
1004
|
|
|
return htmlspecialchars_decode($this->doc->saveHTML($node)); |
|
1005
|
|
|
else |
|
1006
|
|
|
{ |
|
1007
|
|
|
// @todo remove when 5.3.6 min |
|
1008
|
|
|
$doc = new DOMDocument(); |
|
1009
|
|
|
$doc->appendChild($doc->importNode($node, true)); |
|
1010
|
|
|
$html = $doc->saveHTML(); |
|
1011
|
|
|
|
|
1012
|
|
|
// We just want the html of the inserted node, it *may* be wrapped |
|
1013
|
|
|
$html = $this->_returnBodyText($html); |
|
1014
|
|
|
|
|
1015
|
|
|
// Clean it up |
|
1016
|
|
|
$html = rtrim($html, "\n"); |
|
1017
|
|
|
return html_entity_decode(htmlspecialchars_decode($html, ENT_QUOTES), ENT_QUOTES, 'UTF-8'); |
|
1018
|
|
|
} |
|
1019
|
|
|
} |
|
1020
|
|
|
else |
|
1021
|
|
|
return $node->outertext; |
|
1022
|
|
|
} |
|
1023
|
|
|
|
|
1024
|
|
|
/** |
|
1025
|
|
|
* Escapes markup looking text in html to prevent accidental assignment |
|
1026
|
|
|
* |
|
1027
|
|
|
* <p>*stuff*</p> should not convert to *stuff* but \*stuff\* since its not to |
|
1028
|
|
|
* be converted by md to html as <strong>stuff</strong> |
|
1029
|
|
|
* |
|
1030
|
|
|
* @param string $value |
|
1031
|
|
|
* @return string |
|
1032
|
|
|
*/ |
|
1033
|
|
|
private function _escape_text($value) |
|
1034
|
|
|
{ |
|
1035
|
|
|
// Search and replace ... |
|
1036
|
|
|
foreach ($this->_textEscapeRegex as $regex => $replacement) |
|
1037
|
|
|
$value = preg_replace('~' . $regex . '~', $replacement, $value); |
|
1038
|
|
|
|
|
1039
|
|
|
return $value; |
|
1040
|
|
|
} |
|
1041
|
|
|
|
|
1042
|
|
|
/** |
|
1043
|
|
|
* If inline code contains backticks ` as part of its content, we need to wrap them so |
|
1044
|
|
|
* when markdown is run we don't interpret the ` as additional code blocks |
|
1045
|
|
|
* |
|
1046
|
|
|
* @param object $node |
|
1047
|
|
|
* @param string $value |
|
1048
|
|
|
* @return string |
|
1049
|
|
|
*/ |
|
1050
|
|
|
private function _has_ticks($node, $value) |
|
1051
|
|
|
{ |
|
1052
|
|
|
$ticks = ''; |
|
1053
|
|
|
$code_parent = $this->_parser ? $node->parentNode->nodeName : $node->parentNode()->nodeName(); |
|
1054
|
|
|
|
|
1055
|
|
|
// Inside of a pre, we don't do anything |
|
1056
|
|
|
if ($code_parent === 'pre') |
|
1057
|
|
|
return $value; |
|
1058
|
|
|
|
|
1059
|
|
|
// If we have backticks in code, then we back tick the ticks |
|
1060
|
|
|
// e.g. <code>`bla`</code> will become `` `bla` `` so markdown will deal with it properly |
|
1061
|
|
|
preg_match_all('~`+~', $value, $matches); |
|
1062
|
|
|
if (!empty($matches[0])) |
|
1063
|
|
|
{ |
|
1064
|
|
|
// Yup ticks in the hair |
|
1065
|
|
|
$ticks = '`'; |
|
1066
|
|
|
rsort($matches[0]); |
|
1067
|
|
|
|
|
1068
|
|
|
// Backtick as many as needed so markdown will work |
|
1069
|
|
|
while (true) |
|
1070
|
|
|
{ |
|
1071
|
|
|
if (!in_array($ticks, $matches[0])) |
|
1072
|
|
|
break; |
|
1073
|
|
|
$ticks .= '`'; |
|
1074
|
|
|
} |
|
1075
|
|
|
} |
|
1076
|
|
|
|
|
1077
|
|
|
return $ticks; |
|
1078
|
|
|
} |
|
1079
|
|
|
|
|
1080
|
|
|
/** |
|
1081
|
|
|
* Helper function to adjust wrapping width for long-ish links |
|
1082
|
|
|
* |
|
1083
|
|
|
* @param string $markdown |
|
1084
|
|
|
* @param bool|int $buffer |
|
1085
|
|
|
* @return int |
|
1086
|
|
|
*/ |
|
1087
|
|
|
private function _check_link_lenght($markdown, $buffer = false) |
|
1088
|
|
|
{ |
|
1089
|
|
|
// Some links can be very long and if we wrap them they break |
|
1090
|
|
|
$line_strlen = Util::strlen($markdown) + (!empty($buffer) ? (int) $buffer : 0); |
|
1091
|
|
|
if ($line_strlen > $this->body_width) |
|
1092
|
|
|
{ |
|
1093
|
|
|
$this->body_width = $line_strlen; |
|
1094
|
|
|
} |
|
1095
|
|
|
} |
|
1096
|
|
|
|
|
1097
|
|
|
/** |
|
1098
|
|
|
* Breaks a string up so its no more than width characters long |
|
1099
|
|
|
* |
|
1100
|
|
|
* - Will break at word boundaries |
|
1101
|
|
|
* - If no natural space is found will break mid-word |
|
1102
|
|
|
* |
|
1103
|
|
|
* @param string $string |
|
1104
|
|
|
* @param int $width |
|
1105
|
|
|
* @param string $break |
|
1106
|
|
|
* @return string |
|
1107
|
|
|
*/ |
|
1108
|
|
|
private function _utf8_wordwrap($string, $width = 75, $break = "\n") |
|
1109
|
|
|
{ |
|
1110
|
|
|
$strings = explode($break, $string); |
|
1111
|
|
|
$lines = array(); |
|
1112
|
|
|
|
|
1113
|
|
|
foreach ($strings as $string) |
|
1114
|
|
|
{ |
|
1115
|
|
|
$in_quote = isset($string[0]) && $string[0] === '>'; |
|
1116
|
|
|
while (!empty($string)) |
|
1117
|
|
|
{ |
|
1118
|
|
|
// Get the next #width characters before a break (space, punctuation tab etc) |
|
1119
|
|
|
if (preg_match('~^(.{1,' . $width . '})(?:\s|$|,|\.)~u', $string, $matches)) |
|
1120
|
|
|
{ |
|
1121
|
|
|
// Add the #width to the output and set up for the next pass |
|
1122
|
|
|
$lines[] = ($in_quote && $matches[1][0] !== '>' ? '> ' : '') . ltrim($matches[1], ' '); |
|
1123
|
|
|
$string = Util::substr($string, Util::strlen($matches[1])); |
|
1124
|
|
|
} |
|
1125
|
|
|
// Humm just a long word with no place to break so we simply cut it after width characters |
|
1126
|
|
|
else |
|
1127
|
|
|
{ |
|
1128
|
|
|
$lines[] = ($in_quote && $string[0] !== '>' ? '> ' : '') . Util::substr($string, 0, $width); |
|
1129
|
|
|
$string = Util::substr($string, $width); |
|
1130
|
|
|
} |
|
1131
|
|
|
} |
|
1132
|
|
|
} |
|
1133
|
|
|
|
|
1134
|
|
|
// Join it all the shortened sections up on our break characters |
|
1135
|
|
|
return implode($break, $lines); |
|
1136
|
|
|
} |
|
1137
|
|
|
} |
|
1138
|
|
|
|
elkarte /
Elkarte
Our type inference engine has found an assignment to a property that is incompatible with the declared type of that property.
Either this assignment is in error or the assigned type should be added to the documentation/type hint for that property..