Blocks::matchSetextLines() - Code Metrics - Inspection of "WiP" - s9e/TextFormatter - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Branch — wip/litedown (377511)

by Josh

created 2017-10-16 09:11 UTC

Blocks::matchSetextLines() C

↳ Parent: Blocks

Complexity

Conditions	8
Paths	7

Size

Total Lines	35
Code Lines	16

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	0
CRAP Score	72

Importance

Changes

Metric	Value
dl	0
loc	35
ccs	0
cts	22
cp	0
rs	5.3846
c	0
b	0
f	0
cc	8
eloc	16
nc	7
nop	0
crap	72

<?php

/**
* @package   s9e\TextFormatter
* @copyright Copyright (c) 2010-2017 The s9e Authors
* @license   http://www.opensource.org/licenses/mit-license.php The MIT License
*/
namespace s9e\TextFormatter\Plugins\Litedown\Parser;

use s9e\TextFormatter\Parser as Rules;

class Blocks extends AbstractParser
{
	/**
	* Close a list at given offset
	*
	* @param  array   $list
	* @param  integer $textBoundary
	* @return void
	*/
	protected function closeList(array $list, $textBoundary)
	{
		$this->parser->addEndTag('LIST', $textBoundary, 0)->pairWith($list['listTag']);
		$this->parser->addEndTag('LI',   $textBoundary, 0)->pairWith($list['itemTag']);

		if ($list['tight'])
		{
			foreach ($list['itemTags'] as $itemTag)
			{
				$itemTag->removeFlags(Rules::RULE_CREATE_PARAGRAPHS);
			}
		}
	}

	/**
	* Compute the amount of text to ignore at the start of a quote line
	*
	* @param  string  $str           Original quote markup
	* @param  integer $maxQuoteDepth Maximum quote depth
	* @return integer                Number of characters to ignore
	*/
	protected function computeQuoteIgnoreLen($str, $maxQuoteDepth)
	{
		$remaining = $str;
		while (--$maxQuoteDepth >= 0)
		{
			$remaining = preg_replace('/^ *> ?/', '', $remaining);
		}

		return strlen($str) - strlen($remaining);
	}

	/**
	* {@inheritdoc}
	*/
	protected function execute()
	{
		$this->matchSetextLines();

		$codeFence    = null;
		$codeIndent   = 4;
		$codeTag      = null;
		$lineIsEmpty  = true;
		$lists        = [];
		$listsCnt     = 0;
		$newContext   = false;
		$quotes       = [];
		$quotesCnt    = 0;
		$textBoundary = 0;

		$regexp = '/^(?:(?=[-*+\\d \\t>`~#_])((?: {0,3}> ?)+)?([ \\t]+)?(\\* *\\* *\\*[* ]*$|- *- *-[- ]*$|_ *_ *_[_ ]*$|=+$)?((?:[-*+]|\\d+\\.)[ \\t]+(?=\\S))?[ \\t]*(#{1,6}[ \\t]+|```+[^`\\n]*$|~~~+[^~\\n]*$)?)?/m';
		preg_match_all($regexp, $this->text, $matches, PREG_OFFSET_CAPTURE | PREG_SET_ORDER);

		foreach ($matches as $m)
		{
			$matchPos   = $m[0][1];
			$matchLen   = strlen($m[0][0]);
			$ignoreLen  = 0;
			$quoteDepth = 0;

			// If the last line was empty then this is not a continuation, and vice-versa
			$continuation = !$lineIsEmpty;

			// Capture the position of the end of the line and determine whether the line is empty
			$lfPos       = strpos($this->text, "\n", $matchPos);
			$lineIsEmpty = ($lfPos === $matchPos + $matchLen && empty($m[3][0]) && empty($m[4][0]) && empty($m[5][0]));

			// If the line is empty and it's the first empty line then we break current paragraph.
			$breakParagraph = ($lineIsEmpty && $continuation);

			// Count quote marks
			if (!empty($m[1][0]))
			{
				$quoteDepth = substr_count($m[1][0], '>');
				$ignoreLen  = strlen($m[1][0]);
				if (isset($codeTag) && $codeTag->hasAttribute('quoteDepth'))
				{
					$quoteDepth = min($quoteDepth, $codeTag->getAttribute('quoteDepth'));
					$ignoreLen  = $this->computeQuoteIgnoreLen($m[1][0], $quoteDepth);
				}

				// Overwrite quote markup
				$this->overwrite($matchPos, $ignoreLen);
			}

			// Close supernumerary quotes
			if ($quoteDepth < $quotesCnt && !$continuation)
			{
				$newContext = true;

				do
				{
					$this->parser->addEndTag('QUOTE', $textBoundary, 0)
					             ->pairWith(array_pop($quotes));
				}
				while ($quoteDepth < --$quotesCnt);
			}

			// Open new quotes
			if ($quoteDepth > $quotesCnt && !$lineIsEmpty)
			{
				$newContext = true;

				do
				{
					$tag = $this->parser->addStartTag('QUOTE', $matchPos, 0, $quotesCnt - 999);
					$quotes[] = $tag;
				}
				while ($quoteDepth > ++$quotesCnt);
			}

			// Compute the width of the indentation
			$indentWidth = 0;
			$indentPos   = 0;
			if (!empty($m[2][0]) && !$codeFence)
			{
				$indentStr = $m[2][0];
				$indentLen = strlen($indentStr);
				do
				{
					if ($indentStr[$indentPos] === ' ')
					{
						++$indentWidth;
					}
					else
					{
						$indentWidth = ($indentWidth + 4) & ~3;
					}
				}
				while (++$indentPos < $indentLen && $indentWidth < $codeIndent);
			}

			// Test whether we're out of a code block
			if (isset($codeTag) && !$codeFence && $indentWidth < $codeIndent && !$lineIsEmpty)
			{
				$newContext = true;
			}

			if ($newContext)
			{
				$newContext = false;

				// Close the code block if applicable
				if (isset($codeTag))
				{
					if ($textBoundary > $codeTag->getPos())
					{
						// Overwrite the whole block
						$this->overwrite($codeTag->getPos(), $textBoundary - $codeTag->getPos());

						$endTag = $this->parser->addEndTag('CODE', $textBoundary, 0, -1);
						$endTag->pairWith($codeTag);
					}
					else
					{
						// The code block is empty
						$codeTag->invalidate();
					}

					$codeTag = null;
					$codeFence = null;
				}

				// Close all the lists
				foreach ($lists as $list)
				{
					$this->closeList($list, $textBoundary);
				}
				$lists    = [];
				$listsCnt = 0;

				// Mark the block boundary
				if ($matchPos)
				{
					$this->markBoundary($matchPos - 1);
				}
			}

			if ($indentWidth >= $codeIndent)
			{
				if (isset($codeTag) || !$continuation)
				{
					// Adjust the amount of text being ignored
					$ignoreLen += $indentPos;

					if (!isset($codeTag))
					{
						// Create code block
						$codeTag = $this->parser->addStartTag('CODE', $matchPos + $ignoreLen, 0, -999);
					}

					// Clear the captures to prevent any further processing
					$m = [];
				}
			}
			else
			{
				$hasListItem = !empty($m[4][0]);

				if (!$indentWidth && !$continuation && !$hasListItem)
				{
					// Start of a new context
					$listIndex = -1;
				}
				elseif ($continuation && !$hasListItem)
				{
					// Continuation of current list item or paragraph
					$listIndex = $listsCnt - 1;
				}
				elseif (!$listsCnt)
				{
					// We're not inside of a list already, we can start one if there's a list item
					// and it's either not in continuation of a paragraph or immediately after a
					// block
					if ($hasListItem && (!$continuation || $this->text[$matchPos - 1] === "\x17"))
					{
						// Start of a new list
						$listIndex = 0;
					}
					else
					{
						// We're in a normal paragraph
						$listIndex = -1;
					}
				}
				else
				{
					// We're inside of a list but we need to compute the depth
					$listIndex = 0;
					while ($listIndex < $listsCnt && $indentWidth > $lists[$listIndex]['maxIndent'])
					{
						++$listIndex;
					}
				}

				// Close deeper lists
				while ($listIndex < $listsCnt - 1)
				{
					$this->closeList(array_pop($lists), $textBoundary);
					--$listsCnt;
				}

				// If there's no list item at current index, we'll need to either create one or
				// drop down to previous index, in which case we have to adjust maxIndent
				if ($listIndex === $listsCnt && !$hasListItem)
				{
					--$listIndex;
				}

				if ($hasListItem && $listIndex >= 0)
				{
					$breakParagraph = true;

					// Compute the position and amount of text consumed by the item tag
					$tagPos = $matchPos + $ignoreLen + $indentPos;
					$tagLen = strlen($m[4][0]);

					// Create a LI tag that consumes its markup
					$itemTag = $this->parser->addStartTag('LI', $tagPos, $tagLen);

					// Overwrite the markup
					$this->overwrite($tagPos, $tagLen);

					// If the list index is within current lists count it means this is not a new
					// list and we have to close the last item. Otherwise, it's a new list that we
					// have to create
					if ($listIndex < $listsCnt)
					{
						$this->parser->addEndTag('LI', $textBoundary, 0)
						             ->pairWith($lists[$listIndex]['itemTag']);

						// Record the item in the list
						$lists[$listIndex]['itemTag']    = $itemTag;
						$lists[$listIndex]['itemTags'][] = $itemTag;
					}
					else
					{
						++$listsCnt;

						if ($listIndex)
						{
							$minIndent = $lists[$listIndex - 1]['maxIndent'] + 1;
							$maxIndent = max($minIndent, $listIndex * 4);
						}
						else
						{
							$minIndent = 0;
							$maxIndent = $indentWidth;
						}

						// Create a 0-width LIST tag right before the item tag LI
						$listTag = $this->parser->addStartTag('LIST', $tagPos, 0);

						// Test whether the list item ends with a dot, as in "1."
						if (strpos($m[4][0], '.') !== false)
						{
							$listTag->setAttribute('type', 'decimal');

							$start = (int) $m[4][0];
							if ($start !== 1)
							{
								$listTag->setAttribute('start', $start);
							}
						}

						// Record the new list depth
						$lists[] = [
							'listTag'   => $listTag,
							'itemTag'   => $itemTag,
							'itemTags'  => [$itemTag],
							'minIndent' => $minIndent,
							'maxIndent' => $maxIndent,
							'tight'     => true
						];
					}
				}

				// If we're in a list, on a non-empty line preceded with a blank line...
				if ($listsCnt && !$continuation && !$lineIsEmpty)
				{
					// ...and this is not the first item of the list...
					if (count($lists[0]['itemTags']) > 1 || !$hasListItem)
					{
						// ...every list that is currently open becomes loose
						foreach ($lists as &$list)
						{
							$list['tight'] = false;
						}
						unset($list);
					}
				}

				$codeIndent = ($listsCnt + 1) * 4;
			}

			if (isset($m[5]))
			{
				// Headers
				if ($m[5][0][0] === '#')
				{
					$startTagLen = strlen($m[5][0]);
					$startTagPos = $matchPos + $matchLen - $startTagLen;
					$endTagLen   = $this->getAtxHeaderEndTagLen($matchPos + $matchLen, $lfPos);
					$endTagPos   = $lfPos - $endTagLen;

					$this->parser->addTagPair('H' . strspn($m[5][0], '#', 0, 6), $startTagPos, $startTagLen, $endTagPos, $endTagLen);

					// Mark the start and the end of the header as boundaries
					$this->markBoundary($startTagPos);
					$this->markBoundary($lfPos);

					if ($continuation)
					{
						$breakParagraph = true;
					}
				}
				// Code fence
				elseif ($m[5][0][0] === '`' || $m[5][0][0] === '~')
				{
					$tagPos = $matchPos + $ignoreLen;
					$tagLen = $lfPos - $tagPos;

					if (isset($codeTag) && $m[5][0] === $codeFence)
					{
						$endTag = $this->parser->addEndTag('CODE', $tagPos, $tagLen, -1);
						$endTag->pairWith($codeTag);

						$this->parser->addIgnoreTag($textBoundary, $tagPos - $textBoundary);

						// Overwrite the whole block
						$this->overwrite($codeTag->getPos(), $tagPos + $tagLen - $codeTag->getPos());
						$codeTag = null;
						$codeFence = null;
					}
					elseif (!isset($codeTag))
					{
						// Create code block
						$codeTag   = $this->parser->addStartTag('CODE', $tagPos, $tagLen);
						$codeFence = substr($m[5][0], 0, strspn($m[5][0], '`~'));
						$codeTag->setAttribute('quoteDepth', $quoteDepth);

						// Ignore the next character, which should be a newline
						$this->parser->addIgnoreTag($tagPos + $tagLen, 1);

						// Add the language if present, e.g. ```php
						$lang = trim(trim($m[5][0], '`~'));
						if ($lang !== '')
						{
							$codeTag->setAttribute('lang', $lang);
						}
					}
				}
			}
			elseif (!empty($m[3][0]) && !$listsCnt && $this->text[$matchPos + $matchLen] !== "\x17")
			{
				// Horizontal rule
				$this->parser->addSelfClosingTag('HR', $matchPos + $ignoreLen, $matchLen - $ignoreLen);
				$breakParagraph = true;

				// Mark the end of the line as a boundary
				$this->markBoundary($lfPos);
			}
			elseif (isset($this->setextLines[$lfPos]) && $this->setextLines[$lfPos]['quoteDepth'] === $quoteDepth && !$lineIsEmpty && !$listsCnt && !isset($codeTag))
class MyClass { }

$x = new MyClass();
$x->foo = true;
			{
				// Setext-style header
				$this->parser->addTagPair(
					$this->setextLines[$lfPos]['tagName'],
					$matchPos + $ignoreLen,
					0,
					$this->setextLines[$lfPos]['endTagPos'],
					$this->setextLines[$lfPos]['endTagLen']
				);

				// Mark the end of the Setext line
				$this->markBoundary($this->setextLines[$lfPos]['endTagPos'] + $this->setextLines[$lfPos]['endTagLen']);
			}

			if ($breakParagraph)
			{
				$this->parser->addParagraphBreak($textBoundary);
				$this->markBoundary($textBoundary);
			}

			if (!$lineIsEmpty)
			{
				$textBoundary = $lfPos;
			}

			if ($ignoreLen)
			{
				$this->parser->addIgnoreTag($matchPos, $ignoreLen, 1000);
			}
		}
	}

	/**
	* Return the length of the markup at the end of an ATX header
	*
	* @param  integer $startPos Start of the header's text
	* @param  integer $endPos   End of the header's text
	* @return integer
	*/
	protected function getAtxHeaderEndTagLen($startPos, $endPos)
	{
		$content = substr($this->text, $startPos, $endPos - $startPos);
		preg_match('/[ \\t]*#*[ \\t]*$/', $content, $m);

		return strlen($m[0]);
	}

	/**
	* Mark the boundary of a block in the original text
	*
	* @param  integer $pos
	* @return void
	*/
	protected function markBoundary($pos)
	{
		$this->text[$pos] = "\x17";
	}

	/**
	* Capture and store lines that contain a Setext-tyle header
	*
	* @return void
	*/
	protected function matchSetextLines()
	{
		$this->setextLines = [];
		if (strpos($this->text, '-') === false && strpos($this->text, '=') === false)
		{
			return;
		}

		// Capture the any series of - or = alone on a line, optionally preceded with the
		// angle brackets notation used in blockquotes
		$regexp = '/^(?=[-=>])(?:> ?)*(?=[-=])(?:-+|=+) *$/m';
		if (!preg_match_all($regexp, $this->text, $matches, PREG_OFFSET_CAPTURE))
		{
			return;
		}

		foreach ($matches[0] as list($match, $matchPos))
		{
			// Compute the position of the end tag. We start on the LF character before the
			// match and keep rewinding until we find a non-space character
			$endTagPos = $matchPos - 1;
			while ($endTagPos > 0 && $this->text[$endTagPos - 1] === ' ')
			{
				--$endTagPos;
			}

			// Store at the offset of the LF character
			$this->setextLines[$matchPos - 1] = [
				'endTagLen'  => $matchPos + strlen($match) - $endTagPos,
				'endTagPos'  => $endTagPos,
				'quoteDepth' => substr_count($match, '>'),
				'tagName'    => ($match[0] === '=') ? 'H1' : 'H2'
			];
		}
	}
}

1			<?php
2
3			/**
4			* @package s9e\TextFormatter
5			* @copyright Copyright (c) 2010-2017 The s9e Authors
6			* @license http://www.opensource.org/licenses/mit-license.php The MIT License
7			*/
8			namespace s9e\TextFormatter\Plugins\Litedown\Parser;
9
10			use s9e\TextFormatter\Parser as Rules;
11
12			class Blocks extends AbstractParser
13			{
14			/**
15			* Close a list at given offset
16			*
17			* @param array $list
18			* @param integer $textBoundary
19			* @return void
20			*/
21			protected function closeList(array $list, $textBoundary)
22			{
23			$this->parser->addEndTag('LIST', $textBoundary, 0)->pairWith($list['listTag']);
24			$this->parser->addEndTag('LI', $textBoundary, 0)->pairWith($list['itemTag']);
25
26			if ($list['tight'])
27			{
28			foreach ($list['itemTags'] as $itemTag)
29			{
30			$itemTag->removeFlags(Rules::RULE_CREATE_PARAGRAPHS);
31			}
32			}
33			}
34
35			/**
36			* Compute the amount of text to ignore at the start of a quote line
37			*
38			* @param string $str Original quote markup
39			* @param integer $maxQuoteDepth Maximum quote depth
40			* @return integer Number of characters to ignore
41			*/
42			protected function computeQuoteIgnoreLen($str, $maxQuoteDepth)
43			{
44			$remaining = $str;
45			while (--$maxQuoteDepth >= 0)
46			{
47			$remaining = preg_replace('/^ *> ?/', '', $remaining);
48			}
49
50			return strlen($str) - strlen($remaining);
51			}
52
53			/**
54			* {@inheritdoc}
55			*/
56			protected function execute()
57			{
58			$this->matchSetextLines();
59
60			$codeFence = null;
61			$codeIndent = 4;
62			$codeTag = null;
63			$lineIsEmpty = true;
64			$lists = [];
65			$listsCnt = 0;
66			$newContext = false;
67			$quotes = [];
68			$quotesCnt = 0;
69			$textBoundary = 0;
70
71			$regexp = '/^(?:(?=[-+\\d \\t>`~#_])((?: {0,3}> ?)+)?([ \\t]+)?(\\ \\ \\[* ]$\|- - -[- ]$\|_ _ _[_ ]$\|=+$)?((?:[-+]\|\\d+\\.)[ \\t]+(?=\\S))?[ \\t](#{1,6}[ \\t]+\|```+[^`\\n]$\|~~~+[^~\\n]*$)?)?/m';
72			preg_match_all($regexp, $this->text, $matches, PREG_OFFSET_CAPTURE \| PREG_SET_ORDER);
73
74			foreach ($matches as $m)
75			{
76			$matchPos = $m[0][1];
77			$matchLen = strlen($m[0][0]);
78			$ignoreLen = 0;
79			$quoteDepth = 0;
80
81			// If the last line was empty then this is not a continuation, and vice-versa
82			$continuation = !$lineIsEmpty;
83
84			// Capture the position of the end of the line and determine whether the line is empty
85			$lfPos = strpos($this->text, "\n", $matchPos);
86			$lineIsEmpty = ($lfPos === $matchPos + $matchLen && empty($m[3][0]) && empty($m[4][0]) && empty($m[5][0]));
87
88			// If the line is empty and it's the first empty line then we break current paragraph.
89			$breakParagraph = ($lineIsEmpty && $continuation);
90
91			// Count quote marks
92			if (!empty($m[1][0]))
93			{
94			$quoteDepth = substr_count($m[1][0], '>');
95			$ignoreLen = strlen($m[1][0]);
96			if (isset($codeTag) && $codeTag->hasAttribute('quoteDepth'))
97			{
98			$quoteDepth = min($quoteDepth, $codeTag->getAttribute('quoteDepth'));
99			$ignoreLen = $this->computeQuoteIgnoreLen($m[1][0], $quoteDepth);
100			}
101
102			// Overwrite quote markup
103			$this->overwrite($matchPos, $ignoreLen);
104			}
105
106			// Close supernumerary quotes
107			if ($quoteDepth < $quotesCnt && !$continuation)
108			{
109			$newContext = true;
110
111			do
112			{
113			$this->parser->addEndTag('QUOTE', $textBoundary, 0)
114			->pairWith(array_pop($quotes));
115			}
116			while ($quoteDepth < --$quotesCnt);
117			}
118
119			// Open new quotes
120			if ($quoteDepth > $quotesCnt && !$lineIsEmpty)
121			{
122			$newContext = true;
123
124			do
125			{
126			$tag = $this->parser->addStartTag('QUOTE', $matchPos, 0, $quotesCnt - 999);
127			$quotes[] = $tag;
128			}
129			while ($quoteDepth > ++$quotesCnt);
130			}
131
132			// Compute the width of the indentation
133			$indentWidth = 0;
134			$indentPos = 0;
135			if (!empty($m[2][0]) && !$codeFence)
136			{
137			$indentStr = $m[2][0];
138			$indentLen = strlen($indentStr);
139			do
140			{
141			if ($indentStr[$indentPos] === ' ')
142			{
143			++$indentWidth;
144			}
145			else
146			{
147			$indentWidth = ($indentWidth + 4) & ~3;
148			}
149			}
150			while (++$indentPos < $indentLen && $indentWidth < $codeIndent);
151			}
152
153			// Test whether we're out of a code block
154			if (isset($codeTag) && !$codeFence && $indentWidth < $codeIndent && !$lineIsEmpty)
155			{
156			$newContext = true;
157			}
158
159			if ($newContext)
160			{
161			$newContext = false;
162
163			// Close the code block if applicable
164			if (isset($codeTag))
165			{
166			if ($textBoundary > $codeTag->getPos())
167			{
168			// Overwrite the whole block
169			$this->overwrite($codeTag->getPos(), $textBoundary - $codeTag->getPos());
170
171			$endTag = $this->parser->addEndTag('CODE', $textBoundary, 0, -1);
172			$endTag->pairWith($codeTag);
173			}
174			else
175			{
176			// The code block is empty
177			$codeTag->invalidate();
178			}
179
180			$codeTag = null;
181			$codeFence = null;
182			}
183
184			// Close all the lists
185			foreach ($lists as $list)
186			{
187			$this->closeList($list, $textBoundary);
188			}
189			$lists = [];
190			$listsCnt = 0;
191
192			// Mark the block boundary
193			if ($matchPos)
194			{
195			$this->markBoundary($matchPos - 1);
196			}
197			}
198
199			if ($indentWidth >= $codeIndent)
200			{
201			if (isset($codeTag) \|\| !$continuation)
202			{
203			// Adjust the amount of text being ignored
204			$ignoreLen += $indentPos;
205
206			if (!isset($codeTag))
207			{
208			// Create code block
209			$codeTag = $this->parser->addStartTag('CODE', $matchPos + $ignoreLen, 0, -999);
210			}
211
212			// Clear the captures to prevent any further processing
213			$m = [];
214			}
215			}
216			else
217			{
218			$hasListItem = !empty($m[4][0]);
219
220			if (!$indentWidth && !$continuation && !$hasListItem)
221			{
222			// Start of a new context
223			$listIndex = -1;
224			}
225			elseif ($continuation && !$hasListItem)
226			{
227			// Continuation of current list item or paragraph
228			$listIndex = $listsCnt - 1;
229			}
230			elseif (!$listsCnt)
231			{
232			// We're not inside of a list already, we can start one if there's a list item
233			// and it's either not in continuation of a paragraph or immediately after a
234			// block
235			if ($hasListItem && (!$continuation \|\| $this->text[$matchPos - 1] === "\x17"))
236			{
237			// Start of a new list
238			$listIndex = 0;
239			}
240			else
241			{
242			// We're in a normal paragraph
243			$listIndex = -1;
244			}
245			}
246			else
247			{
248			// We're inside of a list but we need to compute the depth
249			$listIndex = 0;
250			while ($listIndex < $listsCnt && $indentWidth > $lists[$listIndex]['maxIndent'])
251			{
252			++$listIndex;
253			}
254			}
255
256			// Close deeper lists
257			while ($listIndex < $listsCnt - 1)
258			{
259			$this->closeList(array_pop($lists), $textBoundary);
260			--$listsCnt;
261			}
262
263			// If there's no list item at current index, we'll need to either create one or
264			// drop down to previous index, in which case we have to adjust maxIndent
265			if ($listIndex === $listsCnt && !$hasListItem)
266			{
267			--$listIndex;
268			}
269
270			if ($hasListItem && $listIndex >= 0)
271			{
272			$breakParagraph = true;
273
274			// Compute the position and amount of text consumed by the item tag
275			$tagPos = $matchPos + $ignoreLen + $indentPos;
276			$tagLen = strlen($m[4][0]);
277
278			// Create a LI tag that consumes its markup
279			$itemTag = $this->parser->addStartTag('LI', $tagPos, $tagLen);
280
281			// Overwrite the markup
282			$this->overwrite($tagPos, $tagLen);
283
284			// If the list index is within current lists count it means this is not a new
285			// list and we have to close the last item. Otherwise, it's a new list that we
286			// have to create
287			if ($listIndex < $listsCnt)
288			{
289			$this->parser->addEndTag('LI', $textBoundary, 0)
290			->pairWith($lists[$listIndex]['itemTag']);
291
292			// Record the item in the list
293			$lists[$listIndex]['itemTag'] = $itemTag;
294			$lists[$listIndex]['itemTags'][] = $itemTag;
295			}
296			else
297			{
298			++$listsCnt;
299
300			if ($listIndex)
301			{
302			$minIndent = $lists[$listIndex - 1]['maxIndent'] + 1;
303			$maxIndent = max($minIndent, $listIndex * 4);
304			}
305			else
306			{
307			$minIndent = 0;
308			$maxIndent = $indentWidth;
309			}
310
311			// Create a 0-width LIST tag right before the item tag LI
312			$listTag = $this->parser->addStartTag('LIST', $tagPos, 0);
313
314			// Test whether the list item ends with a dot, as in "1."
315			if (strpos($m[4][0], '.') !== false)
316			{
317			$listTag->setAttribute('type', 'decimal');
318
319			$start = (int) $m[4][0];
320			if ($start !== 1)
321			{
322			$listTag->setAttribute('start', $start);
323			}
324			}
325
326			// Record the new list depth
327			$lists[] = [
328			'listTag' => $listTag,
329			'itemTag' => $itemTag,
330			'itemTags' => [$itemTag],
331			'minIndent' => $minIndent,
332			'maxIndent' => $maxIndent,
333			'tight' => true
334			];
335			}
336			}
337
338			// If we're in a list, on a non-empty line preceded with a blank line...
339			if ($listsCnt && !$continuation && !$lineIsEmpty)
340			{
341			// ...and this is not the first item of the list...
342			if (count($lists[0]['itemTags']) > 1 \|\| !$hasListItem)
343			{
344			// ...every list that is currently open becomes loose
345			foreach ($lists as &$list)
346			{
347			$list['tight'] = false;
348			}
349			unset($list);
350			}
351			}
352
353			$codeIndent = ($listsCnt + 1) * 4;
354			}
355
356			if (isset($m[5]))
357			{
358			// Headers
359			if ($m[5][0][0] === '#')
360			{
361			$startTagLen = strlen($m[5][0]);
362			$startTagPos = $matchPos + $matchLen - $startTagLen;
363			$endTagLen = $this->getAtxHeaderEndTagLen($matchPos + $matchLen, $lfPos);
364			$endTagPos = $lfPos - $endTagLen;
365
366			$this->parser->addTagPair('H' . strspn($m[5][0], '#', 0, 6), $startTagPos, $startTagLen, $endTagPos, $endTagLen);
367
368			// Mark the start and the end of the header as boundaries
369			$this->markBoundary($startTagPos);
370			$this->markBoundary($lfPos);
371
372			if ($continuation)
373			{
374			$breakParagraph = true;
375			}
376			}
377			// Code fence
378			elseif ($m[5][0][0] === '`' \|\| $m[5][0][0] === '~')
379			{
380			$tagPos = $matchPos + $ignoreLen;
381			$tagLen = $lfPos - $tagPos;
382
383			if (isset($codeTag) && $m[5][0] === $codeFence)
384			{
385			$endTag = $this->parser->addEndTag('CODE', $tagPos, $tagLen, -1);
386			$endTag->pairWith($codeTag);
387
388			$this->parser->addIgnoreTag($textBoundary, $tagPos - $textBoundary);
389
390			// Overwrite the whole block
391			$this->overwrite($codeTag->getPos(), $tagPos + $tagLen - $codeTag->getPos());
392			$codeTag = null;
393			$codeFence = null;
394			}
395			elseif (!isset($codeTag))
396			{
397			// Create code block
398			$codeTag = $this->parser->addStartTag('CODE', $tagPos, $tagLen);
399			$codeFence = substr($m[5][0], 0, strspn($m[5][0], '`~'));
400			$codeTag->setAttribute('quoteDepth', $quoteDepth);
401
402			// Ignore the next character, which should be a newline
403			$this->parser->addIgnoreTag($tagPos + $tagLen, 1);
404
405			// Add the language if present, e.g. ```php
406			$lang = trim(trim($m[5][0], '`~'));
407			if ($lang !== '')
408			{
409			$codeTag->setAttribute('lang', $lang);
410			}
411			}
412			}
413			}
414			elseif (!empty($m[3][0]) && !$listsCnt && $this->text[$matchPos + $matchLen] !== "\x17")
415			{
416			// Horizontal rule
417			$this->parser->addSelfClosingTag('HR', $matchPos + $ignoreLen, $matchLen - $ignoreLen);
418			$breakParagraph = true;
419
420			// Mark the end of the line as a boundary
421			$this->markBoundary($lfPos);
422			}
423			elseif (isset($this->setextLines[$lfPos]) && $this->setextLines[$lfPos]['quoteDepth'] === $quoteDepth && !$lineIsEmpty && !$listsCnt && !isset($codeTag))
			0 ignored issues – show Bug introduced 2017-10-16 09:14 UTC by Report Bug Copy Issue Report The property `setextLines` does not exist. Did you maybe forget to declare it? In PHP it is possible to write to properties without declaring them. For example, the following is perfectly valid PHP code: class MyClass { } $x = new MyClass(); $x->foo = true; Generally, it is a good practice to explictly declare properties to avoid accidental typos and provide IDE auto-completion: class MyClass { public $foo; } $x = new MyClass(); $x->foo = true; Loading history...
424			{
425			// Setext-style header
426			$this->parser->addTagPair(
427			$this->setextLines[$lfPos]['tagName'],
428			$matchPos + $ignoreLen,
429			0,
430			$this->setextLines[$lfPos]['endTagPos'],
431			$this->setextLines[$lfPos]['endTagLen']
432			);
433
434			// Mark the end of the Setext line
435			$this->markBoundary($this->setextLines[$lfPos]['endTagPos'] + $this->setextLines[$lfPos]['endTagLen']);
436			}
437
438			if ($breakParagraph)
439			{
440			$this->parser->addParagraphBreak($textBoundary);
441			$this->markBoundary($textBoundary);
442			}
443
444			if (!$lineIsEmpty)
445			{
446			$textBoundary = $lfPos;
447			}
448
449			if ($ignoreLen)
450			{
451			$this->parser->addIgnoreTag($matchPos, $ignoreLen, 1000);
452			}
453			}
454			}
455
456			/**
457			* Return the length of the markup at the end of an ATX header
458			*
459			* @param integer $startPos Start of the header's text
460			* @param integer $endPos End of the header's text
461			* @return integer
462			*/
463			protected function getAtxHeaderEndTagLen($startPos, $endPos)
464			{
465			$content = substr($this->text, $startPos, $endPos - $startPos);
466			preg_match('/[ \\t]#[ \\t]*$/', $content, $m);
467
468			return strlen($m[0]);
469			}
470
471			/**
472			* Mark the boundary of a block in the original text
473			*
474			* @param integer $pos
475			* @return void
476			*/
477			protected function markBoundary($pos)
478			{
479			$this->text[$pos] = "\x17";
480			}
481
482			/**
483			* Capture and store lines that contain a Setext-tyle header
484			*
485			* @return void
486			*/
487			protected function matchSetextLines()
488			{
489			$this->setextLines = [];
490			if (strpos($this->text, '-') === false && strpos($this->text, '=') === false)
491			{
492			return;
493			}
494
495			// Capture the any series of - or = alone on a line, optionally preceded with the
496			// angle brackets notation used in blockquotes
497			$regexp = '/^(?=[-=>])(?:> ?)(?=[-=])(?:-+\|=+) $/m';
498			if (!preg_match_all($regexp, $this->text, $matches, PREG_OFFSET_CAPTURE))
499			{
500			return;
501			}
502
503			foreach ($matches[0] as list($match, $matchPos))
504			{
505			// Compute the position of the end tag. We start on the LF character before the
506			// match and keep rewinding until we find a non-space character
507			$endTagPos = $matchPos - 1;
508			while ($endTagPos > 0 && $this->text[$endTagPos - 1] === ' ')
509			{
510			--$endTagPos;
511			}
512
513			// Store at the offset of the LF character
514			$this->setextLines[$matchPos - 1] = [
515			'endTagLen' => $matchPos + strlen($match) - $endTagPos,
516			'endTagPos' => $endTagPos,
517			'quoteDepth' => substr_count($match, '>'),
518			'tagName' => ($match[0] === '=') ? 'H1' : 'H2'
519			];
520			}
521			}
522			}

s9e / TextFormatter

Branch — wip/litedown (377511)

Blocks::matchSetextLines() C

Complexity

Size

Duplication

Code Coverage

Importance

Duplication Side-by-Side

Filter issues like