Verzeichnisstruktur phpBB-3.3.15
- Veröffentlicht
- 28.08.2024
So funktioniert es
|
Auf das letzte Element klicken. Dies geht jeweils ein Schritt zurück |
Auf das Icon klicken, dies öffnet das Verzeichnis. Nochmal klicken schließt das Verzeichnis. |
|
(Beispiel Datei-Icons)
|
Auf das Icon klicken um den Quellcode anzuzeigen |
Blocks.php
001 <?php
002
003 /**
004 * @package s9e\TextFormatter
005 * @copyright Copyright (c) 2010-2022 The s9e authors
006 * @license http://www.opensource.org/licenses/mit-license.php The MIT License
007 */
008 namespace s9e\TextFormatter\Plugins\Litedown\Parser\Passes;
009
010 use s9e\TextFormatter\Parser as Rules;
011
012 class Blocks extends AbstractPass
013 {
014 /**
015 * @var array
016 */
017 protected $setextLines = [];
018
019 /**
020 * {@inheritdoc}
021 */
022 public function parse()
023 {
024 $this->matchSetextLines();
025
026 $blocks = [];
027 $blocksCnt = 0;
028 $codeFence = null;
029 $codeIndent = 4;
030 $codeTag = null;
031 $lineIsEmpty = true;
032 $lists = [];
033 $listsCnt = 0;
034 $newContext = false;
035 $textBoundary = 0;
036
037 $regexp = '/^(?:(?=[-*+\\d \\t>`~#_])((?: {0,3}>(?:(?!!)|!(?![^\\n>]*?!<)) ?)+)?([ \\t]+)?(\\* *\\* *\\*[* ]*$|- *- *-[- ]*$|_ *_ *_[_ ]*$|=+$)?((?:[-*+]|\\d+\\.)[ \\t]+(?=\\S))?[ \\t]*(#{1,6}[ \\t]+|```+[^`\\n]*$|~~~+[^~\\n]*$)?)?/m';
038 preg_match_all($regexp, $this->text, $matches, PREG_OFFSET_CAPTURE | PREG_SET_ORDER);
039
040 foreach ($matches as $m)
041 {
042 $blockDepth = 0;
043 $blockMarks = [];
044 $ignoreLen = 0;
045 $matchLen = strlen($m[0][0]);
046 $matchPos = $m[0][1];
047
048 // If the last line was empty then this is not a continuation, and vice-versa
049 $continuation = !$lineIsEmpty;
050
051 // Capture the position of the end of the line and determine whether the line is empty
052 $lfPos = $this->text->indexOf("\n", $matchPos);
053 $lineIsEmpty = ($lfPos === $matchPos + $matchLen && empty($m[3][0]) && empty($m[4][0]) && empty($m[5][0]));
054
055 // If the line is empty and it's the first empty line then we break current paragraph.
056 $breakParagraph = ($lineIsEmpty && $continuation);
057
058 // Count block marks
059 if (!empty($m[1][0]))
060 {
061 $blockMarks = $this->getBlockMarks($m[1][0]);
062 $blockDepth = count($blockMarks);
063 $ignoreLen = strlen($m[1][0]);
064 if (isset($codeTag) && $codeTag->hasAttribute('blockDepth'))
065 {
066 $blockDepth = min($blockDepth, $codeTag->getAttribute('blockDepth'));
067 $ignoreLen = $this->computeBlockIgnoreLen($m[1][0], $blockDepth);
068 }
069
070 // Overwrite block markup
071 $this->text->overwrite($matchPos, $ignoreLen);
072 }
073
074 // Close supernumerary blocks
075 if ($blockDepth < $blocksCnt && !$continuation)
076 {
077 $newContext = true;
078 do
079 {
080 $startTag = array_pop($blocks);
081 $this->parser->addEndTag($startTag->getName(), $textBoundary, 0)
082 ->pairWith($startTag);
083 }
084 while ($blockDepth < --$blocksCnt);
085 }
086
087 // Open new blocks
088 if ($blockDepth > $blocksCnt && !$lineIsEmpty)
089 {
090 $newContext = true;
091 do
092 {
093 $tagName = ($blockMarks[$blocksCnt] === '>!') ? 'SPOILER' : 'QUOTE';
094 $blocks[] = $this->parser->addStartTag($tagName, $matchPos, 0, -999);
095 }
096 while ($blockDepth > ++$blocksCnt);
097 }
098
099 // Compute the width of the indentation
100 $indentWidth = 0;
101 $indentPos = 0;
102 if (!empty($m[2][0]) && !$codeFence)
103 {
104 $indentStr = $m[2][0];
105 $indentLen = strlen($indentStr);
106 do
107 {
108 if ($indentStr[$indentPos] === ' ')
109 {
110 ++$indentWidth;
111 }
112 else
113 {
114 $indentWidth = ($indentWidth + 4) & ~3;
115 }
116 }
117 while (++$indentPos < $indentLen && $indentWidth < $codeIndent);
118 }
119
120 // Test whether we're out of a code block
121 if (isset($codeTag) && !$codeFence && $indentWidth < $codeIndent && !$lineIsEmpty)
122 {
123 $newContext = true;
124 }
125
126 if ($newContext)
127 {
128 $newContext = false;
129
130 // Close the code block if applicable
131 if (isset($codeTag))
132 {
133 if ($textBoundary > $codeTag->getPos())
134 {
135 // Overwrite the whole block
136 $this->text->overwrite($codeTag->getPos(), $textBoundary - $codeTag->getPos());
137 $codeTag->pairWith($this->parser->addEndTag('CODE', $textBoundary, 0, -1));
138 }
139 else
140 {
141 // The code block is empty
142 $codeTag->invalidate();
143 }
144
145 $codeTag = null;
146 $codeFence = null;
147 }
148
149 // Close all the lists
150 foreach ($lists as $list)
151 {
152 $this->closeList($list, $textBoundary);
153 }
154 $lists = [];
155 $listsCnt = 0;
156
157 // Mark the block boundary
158 if ($matchPos)
159 {
160 $this->text->markBoundary($matchPos - 1);
161 }
162 }
163
164 if ($indentWidth >= $codeIndent)
165 {
166 if (isset($codeTag) || !$continuation)
167 {
168 // Adjust the amount of text being ignored
169 $ignoreLen += $indentPos;
170
171 if (!isset($codeTag))
172 {
173 // Create code block
174 $codeTag = $this->parser->addStartTag('CODE', $matchPos + $ignoreLen, 0, -999);
175 }
176
177 // Clear the captures to prevent any further processing
178 $m = [];
179 }
180 }
181 elseif (!isset($codeTag))
182 {
183 $hasListItem = !empty($m[4][0]);
184
185 if (!$indentWidth && !$continuation && !$hasListItem)
186 {
187 // Start of a new context
188 $listIndex = -1;
189 }
190 elseif ($continuation && !$hasListItem)
191 {
192 // Continuation of current list item or paragraph
193 $listIndex = $listsCnt - 1;
194 }
195 elseif (!$listsCnt)
196 {
197 // We're not inside of a list already, we can start one if there's a list item
198 $listIndex = ($hasListItem) ? 0 : -1;
199 }
200 else
201 {
202 // We're inside of a list but we need to compute the depth
203 $listIndex = 0;
204 while ($listIndex < $listsCnt && $indentWidth > $lists[$listIndex]['maxIndent'])
205 {
206 ++$listIndex;
207 }
208 }
209
210 // Close deeper lists
211 while ($listIndex < $listsCnt - 1)
212 {
213 $this->closeList(array_pop($lists), $textBoundary);
214 --$listsCnt;
215 }
216
217 // If there's no list item at current index, we'll need to either create one or
218 // drop down to previous index, in which case we have to adjust maxIndent
219 if ($listIndex === $listsCnt && !$hasListItem)
220 {
221 --$listIndex;
222 }
223
224 if ($hasListItem && $listIndex >= 0)
225 {
226 $breakParagraph = true;
227
228 // Compute the position and amount of text consumed by the item tag
229 $tagPos = $matchPos + $ignoreLen + $indentPos;
230 $tagLen = strlen($m[4][0]);
231
232 // Create a LI tag that consumes its markup
233 $itemTag = $this->parser->addStartTag('LI', $tagPos, $tagLen);
234
235 // Overwrite the markup
236 $this->text->overwrite($tagPos, $tagLen);
237
238 // If the list index is within current lists count it means this is not a new
239 // list and we have to close the last item. Otherwise, it's a new list that we
240 // have to create
241 if ($listIndex < $listsCnt)
242 {
243 $this->parser->addEndTag('LI', $textBoundary, 0)
244 ->pairWith($lists[$listIndex]['itemTag']);
245
246 // Record the item in the list
247 $lists[$listIndex]['itemTag'] = $itemTag;
248 $lists[$listIndex]['itemTags'][] = $itemTag;
249 }
250 else
251 {
252 ++$listsCnt;
253
254 if ($listIndex)
255 {
256 $minIndent = $lists[$listIndex - 1]['maxIndent'] + 1;
257 $maxIndent = max($minIndent, $listIndex * 4);
258 }
259 else
260 {
261 $minIndent = 0;
262 $maxIndent = $indentWidth;
263 }
264
265 // Create a 0-width LIST tag right before the item tag LI
266 $listTag = $this->parser->addStartTag('LIST', $tagPos, 0);
267
268 // Test whether the list item ends with a dot, as in "1."
269 if (strpos($m[4][0], '.') !== false)
270 {
271 $listTag->setAttribute('type', 'decimal');
272
273 $start = (int) $m[4][0];
274 if ($start !== 1)
275 {
276 $listTag->setAttribute('start', $start);
277 }
278 }
279
280 // Record the new list depth
281 $lists[] = [
282 'listTag' => $listTag,
283 'itemTag' => $itemTag,
284 'itemTags' => [$itemTag],
285 'minIndent' => $minIndent,
286 'maxIndent' => $maxIndent,
287 'tight' => true
288 ];
289 }
290 }
291
292 // If we're in a list, on a non-empty line preceded with a blank line...
293 if ($listsCnt && !$continuation && !$lineIsEmpty)
294 {
295 // ...and this is not the first item of the list...
296 if (count($lists[0]['itemTags']) > 1 || !$hasListItem)
297 {
298 // ...every list that is currently open becomes loose
299 foreach ($lists as &$list)
300 {
301 $list['tight'] = false;
302 }
303 unset($list);
304 }
305 }
306
307 $codeIndent = ($listsCnt + 1) * 4;
308 }
309
310 if (isset($m[5]))
311 {
312 // Headers
313 if ($m[5][0][0] === '#')
314 {
315 $startLen = strlen($m[5][0]);
316 $startPos = $matchPos + $matchLen - $startLen;
317 $endLen = $this->getAtxHeaderEndTagLen($matchPos + $matchLen, $lfPos);
318 $endPos = $lfPos - $endLen;
319
320 $this->parser->addTagPair('H' . strspn($m[5][0], '#', 0, 6), $startPos, $startLen, $endPos, $endLen);
321
322 // Mark the start and the end of the header as boundaries
323 $this->text->markBoundary($startPos);
324 $this->text->markBoundary($lfPos);
325
326 if ($continuation)
327 {
328 $breakParagraph = true;
329 }
330 }
331 // Code fence
332 elseif ($m[5][0][0] === '`' || $m[5][0][0] === '~')
333 {
334 $tagPos = $matchPos + $ignoreLen;
335 $tagLen = $lfPos - $tagPos;
336
337 if (isset($codeTag) && $m[5][0] === $codeFence)
338 {
339 $codeTag->pairWith($this->parser->addEndTag('CODE', $tagPos, $tagLen, -1));
340 $this->parser->addIgnoreTag($textBoundary, $tagPos - $textBoundary);
341
342 // Overwrite the whole block
343 $this->text->overwrite($codeTag->getPos(), $tagPos + $tagLen - $codeTag->getPos());
344 $codeTag = null;
345 $codeFence = null;
346 }
347 elseif (!isset($codeTag))
348 {
349 // Create code block
350 $codeTag = $this->parser->addStartTag('CODE', $tagPos, $tagLen);
351 $codeFence = substr($m[5][0], 0, strspn($m[5][0], '`~'));
352 $codeTag->setAttribute('blockDepth', $blockDepth);
353
354 // Ignore the next character, which should be a newline
355 $this->parser->addIgnoreTag($tagPos + $tagLen, 1);
356
357 // Add the language if present, e.g. ```php
358 $lang = trim(trim($m[5][0], '`~'));
359 if ($lang !== '')
360 {
361 $codeTag->setAttribute('lang', $lang);
362 }
363 }
364 }
365 }
366 elseif (!empty($m[3][0]) && !$listsCnt && $this->text->charAt($matchPos + $matchLen) !== "\x17")
367 {
368 // Horizontal rule
369 $this->parser->addSelfClosingTag('HR', $matchPos + $ignoreLen, $matchLen - $ignoreLen);
370 $breakParagraph = true;
371
372 // Mark the end of the line as a boundary
373 $this->text->markBoundary($lfPos);
374 }
375 elseif (isset($this->setextLines[$lfPos]) && $this->setextLines[$lfPos]['blockDepth'] === $blockDepth && !$lineIsEmpty && !$listsCnt && !isset($codeTag))
376 {
377 // Setext-style header
378 $this->parser->addTagPair(
379 $this->setextLines[$lfPos]['tagName'],
380 $matchPos + $ignoreLen,
381 0,
382 $this->setextLines[$lfPos]['endPos'],
383 $this->setextLines[$lfPos]['endLen']
384 );
385
386 // Mark the end of the Setext line
387 $this->text->markBoundary($this->setextLines[$lfPos]['endPos'] + $this->setextLines[$lfPos]['endLen']);
388 }
389
390 if ($breakParagraph)
391 {
392 $this->parser->addParagraphBreak($textBoundary);
393 $this->text->markBoundary($textBoundary);
394 }
395
396 if (!$lineIsEmpty)
397 {
398 $textBoundary = $lfPos;
399 }
400
401 if ($ignoreLen)
402 {
403 $this->parser->addIgnoreTag($matchPos, $ignoreLen, 1000);
404 }
405 }
406 }
407
408 /**
409 * Close a list at given offset
410 *
411 * @param array $list
412 * @param integer $textBoundary
413 * @return void
414 */
415 protected function closeList(array $list, $textBoundary)
416 {
417 $this->parser->addEndTag('LIST', $textBoundary, 0)->pairWith($list['listTag']);
418 $this->parser->addEndTag('LI', $textBoundary, 0)->pairWith($list['itemTag']);
419
420 if ($list['tight'])
421 {
422 foreach ($list['itemTags'] as $itemTag)
423 {
424 $itemTag->removeFlags(Rules::RULE_CREATE_PARAGRAPHS);
425 }
426 }
427 }
428
429 /**
430 * Compute the amount of text to ignore at the start of a block line
431 *
432 * @param string $str Original block markup
433 * @param integer $maxBlockDepth Maximum block depth
434 * @return integer Number of characters to ignore
435 */
436 protected function computeBlockIgnoreLen($str, $maxBlockDepth)
437 {
438 $remaining = $str;
439 while (--$maxBlockDepth >= 0)
440 {
441 $remaining = preg_replace('/^ *>!? ?/', '', $remaining);
442 }
443
444 return strlen($str) - strlen($remaining);
445 }
446
447 /**
448 * Return the length of the markup at the end of an ATX header
449 *
450 * @param integer $startPos Start of the header's text
451 * @param integer $endPos End of the header's text
452 * @return integer
453 */
454 protected function getAtxHeaderEndTagLen($startPos, $endPos)
455 {
456 $content = substr($this->text, $startPos, $endPos - $startPos);
457 preg_match('/[ \\t]*#*[ \\t]*$/', $content, $m);
458
459 return strlen($m[0]);
460 }
461
462 /**
463 * Capture and return block marks from given string
464 *
465 * @param string $str Block markup, composed of ">", "!" and whitespace
466 * @return string[]
467 */
468 protected function getBlockMarks($str)
469 {
470 preg_match_all('(>!?)', $str, $m);
471
472 return $m[0];
473 }
474
475 /**
476 * Capture and store lines that contain a Setext-tyle header
477 *
478 * @return void
479 */
480 protected function matchSetextLines()
481 {
482 if ($this->text->indexOf('-') === false && $this->text->indexOf('=') === false)
483 {
484 return;
485 }
486
487 // Capture the any series of - or = alone on a line, optionally preceded with the
488 // angle brackets notation used in block markup
489 $regexp = '/^(?=[-=>])(?:>!? ?)*(?=[-=])(?:-+|=+) *$/m';
490 if (!preg_match_all($regexp, $this->text, $matches, PREG_OFFSET_CAPTURE))
491 {
492 return;
493 }
494
495 foreach ($matches[0] as list($match, $matchPos))
496 {
497 // Compute the position of the end tag. We start on the LF character before the
498 // match and keep rewinding until we find a non-space character
499 $endPos = $matchPos - 1;
500 while ($endPos > 0 && $this->text->charAt($endPos - 1) === ' ')
501 {
502 --$endPos;
503 }
504
505 // Store at the offset of the LF character
506 $this->setextLines[$matchPos - 1] = [
507 'endLen' => $matchPos + strlen($match) - $endPos,
508 'endPos' => $endPos,
509 'blockDepth' => substr_count($match, '>'),
510 'tagName' => ($match[0] === '=') ? 'H1' : 'H2'
511 ];
512 }
513 }
514 }