1
0
mirror of https://github.com/erusev/parsedown.git synced 2023-08-10 21:13:06 +03:00
parsedown/Parsedown.php

1948 lines
49 KiB
PHP
Raw Normal View History

2013-11-09 01:40:00 +04:00
<?php
2013-07-11 00:22:16 +04:00
2013-11-09 01:40:00 +04:00
#
#
2013-07-11 00:22:16 +04:00
# Parsedown
# http://parsedown.org
2013-11-09 01:40:00 +04:00
#
# (c) Emanuil Rusev
2013-07-11 00:22:16 +04:00
# http://erusev.com
2013-11-09 01:40:00 +04:00
#
2014-01-24 03:28:03 +04:00
# For the full license information, view the LICENSE file that was distributed
# with this source code.
2013-11-09 01:40:00 +04:00
#
#
2013-07-11 00:22:16 +04:00
class Parsedown
{
2014-04-17 11:59:35 +04:00
# ~
2013-11-09 01:40:00 +04:00
2018-04-08 04:46:30 +03:00
const version = '1.8.0-beta-1';
2015-01-19 18:11:13 +03:00
# ~
2014-04-17 11:59:35 +04:00
function text($text)
{
$Elements = $this->textElements($text);
# convert to markup
$markup = $this->elements($Elements);
# trim line breaks
$markup = trim($markup, "\n");
return $markup;
}
protected function textElements($text)
{
2014-05-11 23:31:02 +04:00
# make sure no definitions are set
$this->DefinitionData = array();
2014-05-11 23:31:02 +04:00
2014-01-24 02:41:45 +04:00
# standardize line breaks
2015-01-15 03:37:20 +03:00
$text = str_replace(array("\r\n", "\r"), "\n", $text);
2013-11-09 01:40:00 +04:00
2014-01-24 02:41:45 +04:00
# remove surrounding line breaks
$text = trim($text, "\n");
2013-11-09 01:40:00 +04:00
2014-01-24 02:41:45 +04:00
# split text into lines
$lines = explode("\n", $text);
2013-11-09 01:40:00 +04:00
# iterate through lines to identify blocks
return $this->linesElements($lines);
}
2013-11-09 01:40:00 +04:00
#
2014-04-17 11:59:35 +04:00
# Setters
#
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
function setBreaksEnabled($breaksEnabled)
{
2014-04-17 11:59:35 +04:00
$this->breaksEnabled = $breaksEnabled;
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
return $this;
}
2013-11-09 01:40:00 +04:00
2015-01-18 20:38:57 +03:00
protected $breaksEnabled;
2014-09-22 03:36:42 +04:00
function setMarkupEscaped($markupEscaped)
{
2014-09-22 03:36:42 +04:00
$this->markupEscaped = $markupEscaped;
return $this;
}
2015-01-18 20:38:57 +03:00
protected $markupEscaped;
2015-01-08 17:13:55 +03:00
function setUrlsLinked($urlsLinked)
{
$this->urlsLinked = $urlsLinked;
return $this;
}
2015-01-18 20:38:57 +03:00
protected $urlsLinked = true;
function setSafeMode($safeMode)
2015-01-21 05:50:36 +03:00
{
$this->safeMode = (bool) $safeMode;
2015-01-21 05:50:36 +03:00
return $this;
}
protected $safeMode;
2015-01-21 05:50:36 +03:00
function setStrictMode($strictMode)
2018-03-29 19:44:47 +03:00
{
$this->strictMode = (bool) $strictMode;
2018-03-29 19:44:47 +03:00
return $this;
}
protected $strictMode;
2018-03-29 19:44:47 +03:00
protected $safeLinksWhitelist = array(
'http://',
'https://',
'ftp://',
'ftps://',
'mailto:',
'data:image/png;base64,',
'data:image/gif;base64,',
2017-05-02 21:55:03 +03:00
'data:image/jpeg;base64,',
2017-05-05 23:32:27 +03:00
'irc:',
'ircs:',
'git:',
'ssh:',
'news:',
'steam:',
);
2014-04-17 11:59:35 +04:00
#
2014-05-05 15:39:40 +04:00
# Lines
2014-04-17 11:59:35 +04:00
#
2014-05-10 17:28:00 +04:00
protected $BlockTypes = array(
2015-01-05 16:05:18 +03:00
'#' => array('Header'),
2014-04-17 11:59:35 +04:00
'*' => array('Rule', 'List'),
'+' => array('List'),
2015-01-10 14:54:55 +03:00
'-' => array('SetextHeader', 'Table', 'Rule', 'List'),
2014-04-17 11:59:35 +04:00
'0' => array('List'),
'1' => array('List'),
'2' => array('List'),
'3' => array('List'),
'4' => array('List'),
'5' => array('List'),
'6' => array('List'),
'7' => array('List'),
'8' => array('List'),
'9' => array('List'),
2014-04-18 01:19:22 +04:00
':' => array('Table'),
2014-05-14 14:14:49 +04:00
'<' => array('Comment', 'Markup'),
2015-01-10 14:54:55 +03:00
'=' => array('SetextHeader'),
2014-04-17 11:59:35 +04:00
'>' => array('Quote'),
'[' => array('Reference'),
2014-04-17 11:59:35 +04:00
'_' => array('Rule'),
'`' => array('FencedCode'),
'|' => array('Table'),
'~' => array('FencedCode'),
);
2013-11-09 01:40:00 +04:00
2014-05-05 15:39:40 +04:00
# ~
2014-04-17 11:59:35 +04:00
protected $unmarkedBlockTypes = array(
2015-01-05 00:34:19 +03:00
'Code',
2014-04-17 11:59:35 +04:00
);
2014-05-05 15:39:40 +04:00
#
# Blocks
#
protected function lines(array $lines)
2018-03-21 05:18:34 +03:00
{
return $this->elements($this->linesElements($lines));
}
protected function linesElements(array $lines)
2014-04-17 11:59:35 +04:00
{
$Elements = array();
2014-04-17 11:59:35 +04:00
$CurrentBlock = null;
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
foreach ($lines as $line)
{
2018-04-09 17:11:45 +03:00
if (chop($line) === '')
2014-04-17 11:59:35 +04:00
{
2014-04-28 03:27:05 +04:00
if (isset($CurrentBlock))
2014-04-17 11:59:35 +04:00
{
2018-04-09 16:13:10 +03:00
$CurrentBlock['interrupted'] = (isset($CurrentBlock['interrupted'])
? $CurrentBlock['interrupted'] + 1 : 1
);
2014-04-17 11:59:35 +04:00
}
2014-04-28 03:27:05 +04:00
continue;
}
2018-04-09 02:46:26 +03:00
for (
$beforeTab = strstr($line, "\t", true);
$beforeTab !== false;
$beforeTab = strstr($line, "\t", true)
) {
$shortage = 4 - mb_strlen($beforeTab, 'utf-8') % 4;
$line = $beforeTab
. str_repeat(' ', $shortage)
. substr($line, strlen($beforeTab) + 1)
;
2015-01-10 03:45:51 +03:00
}
2018-04-09 02:05:12 +03:00
$indent = strspn($line, ' ');
2014-04-17 11:59:35 +04:00
$text = $indent > 0 ? substr($line, $indent) : $line;
2014-04-17 11:59:35 +04:00
# ~
2014-04-17 11:59:35 +04:00
$Line = array('body' => $line, 'indent' => $indent, 'text' => $text);
2014-05-05 14:46:26 +04:00
# ~
2015-07-02 01:01:14 +03:00
if (isset($CurrentBlock['continuable']))
2014-04-17 11:59:35 +04:00
{
$Block = $this->{"block{$CurrentBlock['type']}Continue"}($Line, $CurrentBlock);
2014-04-17 11:59:35 +04:00
if (isset($Block))
{
$CurrentBlock = $Block;
2014-04-17 11:59:35 +04:00
continue;
}
else
{
2015-12-19 06:45:14 +03:00
if ($this->isBlockCompletable($CurrentBlock['type']))
{
$CurrentBlock = $this->{"block{$CurrentBlock['type']}Complete"}($CurrentBlock);
}
2014-04-17 11:59:35 +04:00
}
}
2014-04-17 11:59:35 +04:00
# ~
2018-04-09 17:11:45 +03:00
$marker = $text[0];
2013-11-09 01:40:00 +04:00
2014-04-27 02:54:52 +04:00
# ~
$blockTypes = $this->unmarkedBlockTypes;
2014-05-10 17:28:00 +04:00
if (isset($this->BlockTypes[$marker]))
2014-04-17 11:59:35 +04:00
{
2014-05-10 17:28:00 +04:00
foreach ($this->BlockTypes[$marker] as $blockType)
2014-04-17 11:59:35 +04:00
{
$blockTypes []= $blockType;
2014-04-17 11:59:35 +04:00
}
}
2014-04-17 11:59:35 +04:00
#
# ~
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
foreach ($blockTypes as $blockType)
{
$Block = $this->{"block$blockType"}($Line, $CurrentBlock);
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
if (isset($Block))
{
$Block['type'] = $blockType;
2013-11-09 01:40:00 +04:00
2014-05-05 14:46:26 +04:00
if ( ! isset($Block['identified']))
{
if (isset($CurrentBlock))
{
$Elements[] = $CurrentBlock['element'];
}
2014-04-17 11:59:35 +04:00
$Block['identified'] = true;
}
2013-11-10 00:23:56 +04:00
2015-12-19 06:45:14 +03:00
if ($this->isBlockContinuable($blockType))
{
2015-07-02 01:01:14 +03:00
$Block['continuable'] = true;
}
2014-04-17 11:59:35 +04:00
$CurrentBlock = $Block;
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
continue 2;
}
}
2014-04-17 11:59:35 +04:00
# ~
if (isset($CurrentBlock) and $CurrentBlock['type'] === 'Paragraph')
{
$Block = $this->paragraphContinue($Line, $CurrentBlock);
}
if (isset($Block))
{
$CurrentBlock = $Block;
2014-04-17 11:59:35 +04:00
}
else
{
if (isset($CurrentBlock))
{
$Elements[] = $CurrentBlock['element'];
}
2014-04-17 11:59:35 +04:00
2015-01-05 00:34:19 +03:00
$CurrentBlock = $this->paragraph($Line);
2014-05-11 23:31:02 +04:00
2014-05-12 01:34:47 +04:00
$CurrentBlock['identified'] = true;
2014-04-17 11:59:35 +04:00
}
}
# ~
2015-12-19 06:45:14 +03:00
if (isset($CurrentBlock['continuable']) and $this->isBlockCompletable($CurrentBlock['type']))
{
$CurrentBlock = $this->{"block{$CurrentBlock['type']}Complete"}($CurrentBlock);
}
# ~
if (isset($CurrentBlock))
2015-01-11 15:35:09 +03:00
{
$Elements[] = $CurrentBlock['element'];
2015-01-11 15:35:09 +03:00
}
2014-04-17 11:59:35 +04:00
# ~
2018-03-21 05:18:34 +03:00
return $Elements;
2014-04-17 11:59:35 +04:00
}
2015-12-19 06:45:14 +03:00
protected function isBlockContinuable($Type)
{
return method_exists($this, "block${Type}Continue");
}
2015-12-19 06:45:14 +03:00
protected function isBlockCompletable($Type)
{
return method_exists($this, "block${Type}Complete");
}
2014-04-17 11:59:35 +04:00
#
2014-05-05 15:39:40 +04:00
# Code
2014-02-23 20:55:34 +04:00
2015-01-05 00:34:19 +03:00
protected function blockCode($Line, $Block = null)
2014-04-17 11:59:35 +04:00
{
if (isset($Block) and $Block['type'] === 'Paragraph' and ! isset($Block['interrupted']))
2014-12-15 01:52:03 +03:00
{
return;
}
2014-05-05 15:39:40 +04:00
if ($Line['indent'] >= 4)
2014-04-17 11:59:35 +04:00
{
2014-05-05 15:39:40 +04:00
$text = substr($Line['body'], 4);
2014-04-17 11:59:35 +04:00
$Block = array(
'element' => array(
2014-05-05 15:39:40 +04:00
'name' => 'pre',
'element' => array(
2014-05-05 15:39:40 +04:00
'name' => 'code',
'text' => $text,
),
2014-04-17 11:59:35 +04:00
),
);
2014-02-23 20:55:34 +04:00
2014-04-17 11:59:35 +04:00
return $Block;
}
}
2014-02-23 20:55:34 +04:00
2015-01-05 00:34:19 +03:00
protected function blockCodeContinue($Line, $Block)
2014-04-17 11:59:35 +04:00
{
2014-05-05 15:39:40 +04:00
if ($Line['indent'] >= 4)
2014-04-17 11:59:35 +04:00
{
2014-05-05 15:39:40 +04:00
if (isset($Block['interrupted']))
2014-04-17 11:59:35 +04:00
{
2018-04-09 16:13:10 +03:00
$Block['element']['element']['text'] .= str_repeat("\n", $Block['interrupted']);
2014-05-05 15:39:40 +04:00
unset($Block['interrupted']);
2014-04-17 11:59:35 +04:00
}
2013-11-09 01:40:00 +04:00
$Block['element']['element']['text'] .= "\n";
2014-05-05 15:39:40 +04:00
$text = substr($Line['body'], 4);
$Block['element']['element']['text'] .= $text;
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
return $Block;
}
}
2014-01-17 02:36:11 +04:00
2014-05-14 14:14:49 +04:00
#
# Comment
2015-01-05 00:34:19 +03:00
protected function blockComment($Line)
2014-05-14 14:14:49 +04:00
{
if ($this->markupEscaped or $this->safeMode)
{
return;
}
if (strpos($Line['text'], '<!--') === 0)
2014-05-14 14:14:49 +04:00
{
$Block = array(
2018-03-21 05:18:34 +03:00
'element' => array(
'rawHtml' => $Line['body'],
'autobreak' => true,
),
2014-05-14 14:14:49 +04:00
);
if (strpos($Line['text'], '-->') !== false)
2014-05-14 14:14:49 +04:00
{
$Block['closed'] = true;
}
return $Block;
}
}
2015-01-05 00:34:19 +03:00
protected function blockCommentContinue($Line, array $Block)
2014-05-14 14:14:49 +04:00
{
if (isset($Block['closed']))
{
return;
}
$Block['element']['rawHtml'] .= "\n{$Line['body']}";
2014-05-14 14:14:49 +04:00
if (strpos($Line['text'], '-->') !== false)
2014-05-14 14:14:49 +04:00
{
$Block['closed'] = true;
}
return $Block;
}
2014-04-17 11:59:35 +04:00
#
# Fenced Code
2014-02-23 20:55:34 +04:00
2015-01-05 00:34:19 +03:00
protected function blockFencedCode($Line)
2014-04-17 11:59:35 +04:00
{
$marker = $Line['text'][0];
$openerLength = strspn($Line['text'], $marker);
if ($openerLength < 3)
2014-04-17 11:59:35 +04:00
{
return;
}
2014-02-23 20:55:34 +04:00
$infostring = trim(substr($Line['text'], $openerLength), "\t ");
2014-02-23 20:55:34 +04:00
if (strpos($infostring, '`') !== false)
{
return;
}
2014-02-23 20:55:34 +04:00
$Element = array(
'name' => 'code',
'text' => '',
);
2014-02-23 20:55:34 +04:00
if ($infostring !== '')
{
$Element['attributes'] = array('class' => "language-$infostring");
2014-04-17 11:59:35 +04:00
}
$Block = array(
'char' => $marker,
'openerLength' => $openerLength,
'element' => array(
'name' => 'pre',
'element' => $Element,
),
);
return $Block;
2014-04-17 11:59:35 +04:00
}
2014-02-23 20:55:34 +04:00
protected function blockFencedCodeContinue($Line, $Block)
2014-04-17 11:59:35 +04:00
{
if (isset($Block['complete']))
{
return;
}
2014-02-23 20:55:34 +04:00
2014-04-17 11:59:35 +04:00
if (isset($Block['interrupted']))
{
2018-04-09 16:13:10 +03:00
$Block['element']['element']['text'] .= str_repeat("\n", $Block['interrupted']);
2014-02-23 20:55:34 +04:00
2014-04-17 11:59:35 +04:00
unset($Block['interrupted']);
}
2014-02-23 20:55:34 +04:00
if (($len = strspn($Line['text'], $Block['char'])) >= $Block['openerLength']
and chop(substr($Line['text'], $len), ' ') === ''
) {
$Block['element']['element']['text'] = substr($Block['element']['element']['text'], 1);
2014-02-23 20:55:34 +04:00
2014-04-17 11:59:35 +04:00
$Block['complete'] = true;
2014-02-23 20:55:34 +04:00
2014-04-17 11:59:35 +04:00
return $Block;
}
2014-02-23 20:55:34 +04:00
$Block['element']['element']['text'] .= "\n{$Line['body']}";
2014-02-23 20:55:34 +04:00
2014-05-01 02:05:31 +04:00
return $Block;
}
2015-01-05 16:05:18 +03:00
#
# Header
protected function blockHeader($Line)
{
2018-04-09 02:55:36 +03:00
$level = strspn($Line['text'], '#');
2015-01-05 16:05:18 +03:00
if ($level > 6)
{
return;
}
2015-01-10 03:45:51 +03:00
$text = trim($Line['text'], '#');
if ($this->strictMode and isset($text[0]) and $text[0] !== ' ')
{
return;
}
$text = trim($text, ' ');
2015-01-05 16:05:18 +03:00
$Block = array(
'element' => array(
'name' => 'h' . min(6, $level),
'handler' => array(
'function' => 'lineElements',
'argument' => $text,
'destination' => 'elements',
)
),
);
2015-01-05 16:05:18 +03:00
return $Block;
2015-01-05 16:05:18 +03:00
}
2014-04-17 11:59:35 +04:00
#
# List
2014-02-23 20:55:34 +04:00
protected function blockList($Line, array $CurrentBlock = null)
2014-04-17 11:59:35 +04:00
{
2018-04-08 17:41:18 +03:00
list($name, $pattern) = $Line['text'][0] <= '-' ? array('ul', '[*+-]') : array('ol', '[0-9]{1,9}+[.\)]');
2018-04-08 17:41:18 +03:00
if (preg_match('/^('.$pattern.'([ ]++|$))(.*+)/', $Line['text'], $matches))
2014-04-17 11:59:35 +04:00
{
$contentIndent = strlen($matches[2]);
2016-10-13 20:52:38 +03:00
if ($contentIndent >= 5)
{
$contentIndent -= 1;
$matches[1] = substr($matches[1], 0, -$contentIndent);
$matches[3] = str_repeat(' ', $contentIndent) . $matches[3];
2016-10-13 20:52:38 +03:00
}
elseif ($contentIndent === 0)
{
$matches[1] .= ' ';
}
2018-04-09 00:32:42 +03:00
$markerWithoutWhitespace = strstr($matches[1], ' ', true);
2014-04-17 11:59:35 +04:00
$Block = array(
'indent' => $Line['indent'],
'pattern' => $pattern,
'data' => array(
'type' => $name,
'marker' => $matches[1],
2018-04-09 00:32:42 +03:00
'markerType' => ($name === 'ul' ? $markerWithoutWhitespace : substr($markerWithoutWhitespace, -1)),
),
2014-04-17 11:59:35 +04:00
'element' => array(
'name' => $name,
'elements' => array(),
2014-04-17 11:59:35 +04:00
),
);
2018-04-08 17:41:18 +03:00
$Block['data']['markerTypeRegex'] = preg_quote($Block['data']['markerType'], '/');
if ($name === 'ol')
{
$listStart = ltrim(strstr($matches[1], $Block['data']['markerType'], true), '0') ?: '0';
2018-03-27 13:23:04 +03:00
if ($listStart !== '1')
{
if (
isset($CurrentBlock)
and $CurrentBlock['type'] === 'Paragraph'
and ! isset($CurrentBlock['interrupted'])
) {
return;
}
$Block['element']['attributes'] = array('start' => $listStart);
}
}
2014-04-17 11:59:35 +04:00
$Block['li'] = array(
'name' => 'li',
'handler' => array(
'function' => 'li',
'argument' => !empty($matches[3]) ? array($matches[3]) : array(),
'destination' => 'elements'
)
2014-04-17 11:59:35 +04:00
);
$Block['element']['elements'] []= & $Block['li'];
2014-04-17 11:59:35 +04:00
return $Block;
}
}
2014-02-23 20:55:34 +04:00
2015-01-05 00:34:19 +03:00
protected function blockListContinue($Line, array $Block)
2014-04-17 11:59:35 +04:00
{
if (isset($Block['interrupted']) and empty($Block['li']['handler']['argument']))
2016-10-13 20:52:38 +03:00
{
return null;
}
$requiredIndent = ($Block['indent'] + strlen($Block['data']['marker']));
2018-03-27 14:11:00 +03:00
if ($Line['indent'] < $requiredIndent
and (
(
$Block['data']['type'] === 'ol'
2018-04-08 17:41:18 +03:00
and preg_match('/^[0-9]++'.$Block['data']['markerTypeRegex'].'(?:[ ]++(.*)|$)/', $Line['text'], $matches)
2018-03-27 14:11:00 +03:00
) or (
$Block['data']['type'] === 'ul'
2018-04-08 17:41:18 +03:00
and preg_match('/^'.$Block['data']['markerTypeRegex'].'(?:[ ]++(.*)|$)/', $Line['text'], $matches)
)
)
2018-03-27 14:11:00 +03:00
) {
2014-04-17 11:59:35 +04:00
if (isset($Block['interrupted']))
{
$Block['li']['handler']['argument'] []= '';
2014-02-23 20:55:34 +04:00
2017-02-19 19:12:04 +03:00
$Block['loose'] = true;
2014-04-17 11:59:35 +04:00
unset($Block['interrupted']);
}
2014-02-23 20:55:34 +04:00
2014-04-17 11:59:35 +04:00
unset($Block['li']);
2014-02-23 20:55:34 +04:00
2015-01-15 23:31:31 +03:00
$text = isset($matches[1]) ? $matches[1] : '';
$Block['indent'] = $Line['indent'];
2014-04-17 11:59:35 +04:00
$Block['li'] = array(
'name' => 'li',
'handler' => array(
'function' => 'li',
'argument' => array($text),
'destination' => 'elements'
)
2014-04-17 11:59:35 +04:00
);
2014-02-23 20:55:34 +04:00
$Block['element']['elements'] []= & $Block['li'];
2014-02-23 20:55:34 +04:00
2014-04-17 11:59:35 +04:00
return $Block;
}
elseif ($Line['indent'] < $requiredIndent and $this->blockList($Line))
{
return null;
}
2014-02-23 20:55:34 +04:00
2015-01-15 23:04:02 +03:00
if ($Line['text'][0] === '[' and $this->blockReference($Line))
{
return $Block;
}
if ($Line['indent'] >= $requiredIndent)
2014-04-17 11:59:35 +04:00
{
if (isset($Block['interrupted']))
{
$Block['li']['handler']['argument'] []= '';
$Block['loose'] = true;
unset($Block['interrupted']);
}
$text = substr($Line['body'], $requiredIndent);
2014-02-23 20:55:34 +04:00
$Block['li']['handler']['argument'] []= $text;
2014-02-23 20:55:34 +04:00
2014-04-17 11:59:35 +04:00
return $Block;
}
2014-02-23 20:55:34 +04:00
if ( ! isset($Block['interrupted']))
2014-04-17 11:59:35 +04:00
{
2018-04-08 17:41:18 +03:00
$text = preg_replace('/^[ ]{0,'.$requiredIndent.'}+/', '', $Line['body']);
$Block['li']['handler']['argument'] []= $text;
2014-04-17 11:59:35 +04:00
return $Block;
}
}
2017-02-19 19:12:04 +03:00
protected function blockListComplete(array $Block)
{
if (isset($Block['loose']))
{
foreach ($Block['element']['elements'] as &$li)
2017-02-19 19:12:04 +03:00
{
if (end($li['handler']['argument']) !== '')
2017-02-19 19:12:04 +03:00
{
$li['handler']['argument'] []= '';
2017-02-19 19:12:04 +03:00
}
}
}
return $Block;
}
2014-04-17 11:59:35 +04:00
#
# Quote
2013-11-09 01:40:00 +04:00
2015-01-05 00:34:19 +03:00
protected function blockQuote($Line)
2014-04-17 11:59:35 +04:00
{
2018-04-08 17:41:18 +03:00
if (preg_match('/^>[ ]?+(.*+)/', $Line['text'], $matches))
2014-04-17 11:59:35 +04:00
{
$Block = array(
'element' => array(
'name' => 'blockquote',
'handler' => array(
'function' => 'linesElements',
'argument' => (array) $matches[1],
'destination' => 'elements',
)
2014-04-17 11:59:35 +04:00
),
);
return $Block;
}
}
2015-01-05 00:34:19 +03:00
protected function blockQuoteContinue($Line, array $Block)
2014-04-17 11:59:35 +04:00
{
2018-03-28 05:27:09 +03:00
if (isset($Block['interrupted']))
2014-04-17 11:59:35 +04:00
{
2018-03-28 05:27:09 +03:00
return;
}
2018-04-08 17:41:18 +03:00
if ($Line['text'][0] === '>' and preg_match('/^>[ ]?+(.*+)/', $Line['text'], $matches))
2018-03-28 05:27:09 +03:00
{
$Block['element']['handler']['argument'] []= $matches[1];
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
return $Block;
}
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
if ( ! isset($Block['interrupted']))
{
$Block['element']['handler']['argument'] []= $Line['text'];
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
return $Block;
}
}
2014-05-05 15:39:40 +04:00
#
# Rule
2015-01-05 00:34:19 +03:00
protected function blockRule($Line)
2014-05-05 15:39:40 +04:00
{
2018-04-09 04:31:36 +03:00
$marker = $Line['text'][0];
if (substr_count($Line['text'], $marker) >= 3 and chop($Line['text'], " $marker") === '')
2014-05-05 15:39:40 +04:00
{
$Block = array(
'element' => array(
'name' => 'hr',
2014-05-05 15:39:40 +04:00
),
);
return $Block;
}
}
#
# Setext
2015-01-10 14:54:55 +03:00
protected function blockSetextHeader($Line, array $Block = null)
2014-05-05 15:39:40 +04:00
{
if ( ! isset($Block) or $Block['type'] !== 'Paragraph' or isset($Block['interrupted']))
2014-05-05 15:39:40 +04:00
{
return;
}
2018-03-26 00:50:42 +03:00
if (
chop(chop($Line['text'], ' '), $Line['text'][0]) === ''
and $Line['indent'] < 4
) {
2014-05-05 15:39:40 +04:00
$Block['element']['name'] = $Line['text'][0] === '=' ? 'h1' : 'h2';
return $Block;
}
}
#
# Markup
2015-01-05 00:34:19 +03:00
protected function blockMarkup($Line)
2014-05-05 15:39:40 +04:00
{
if ($this->markupEscaped or $this->safeMode)
{
2014-09-22 03:36:42 +04:00
return;
}
2018-04-08 17:41:18 +03:00
if (preg_match('/^<[\/]?+(\w*)(?:[ ]*+'.$this->regexHtmlAttribute.')*+[ ]*+(\/)?>/', $Line['text'], $matches))
2014-12-15 02:07:29 +03:00
{
2015-07-31 01:33:21 +03:00
$element = strtolower($matches[1]);
if (in_array($element, $this->textLevelElements))
2014-05-05 15:39:40 +04:00
{
return;
}
$Block = array(
'name' => $matches[1],
2018-03-21 05:18:34 +03:00
'element' => array(
'rawHtml' => $Line['text'],
'autobreak' => true,
),
);
2014-05-05 15:39:40 +04:00
return $Block;
}
2014-05-05 15:39:40 +04:00
}
2015-01-05 00:34:19 +03:00
protected function blockMarkupContinue($Line, array $Block)
2014-05-05 15:39:40 +04:00
{
if (isset($Block['closed']) or isset($Block['interrupted']))
2014-05-05 15:39:40 +04:00
{
return;
}
$Block['element']['rawHtml'] .= "\n{$Line['body']}";
2014-05-05 15:39:40 +04:00
return $Block;
}
#
# Reference
protected function blockReference($Line)
{
2018-04-08 17:41:18 +03:00
if (preg_match('/^\[(.+?)\]:[ ]*+<?(\S+?)>?(?:[ ]+["\'(](.+)["\')])?[ ]*+$/', $Line['text'], $matches))
{
$id = strtolower($matches[1]);
$Data = array(
'url' => $matches[2],
'title' => isset($matches[3]) ? $matches[3] : null,
);
$this->DefinitionData['Reference'][$id] = $Data;
$Block = array(
'element' => array(),
);
return $Block;
}
}
2014-04-17 11:59:35 +04:00
#
# Table
2015-01-05 00:34:19 +03:00
protected function blockTable($Line, array $Block = null)
2014-04-17 11:59:35 +04:00
{
if ( ! isset($Block) or $Block['type'] !== 'Paragraph' or isset($Block['interrupted']))
2014-04-17 11:59:35 +04:00
{
return;
}
if (
strpos($Block['element']['handler']['argument'], '|') === false
and strpos($Line['text'], '|') === false
and strpos($Line['text'], ':') === false
or strpos($Block['element']['handler']['argument'], "\n") !== false
) {
return;
}
2013-11-09 01:40:00 +04:00
if (chop($Line['text'], ' -:|') !== '')
{
return;
}
$alignments = array();
2013-11-10 00:23:56 +04:00
$divider = $Line['text'];
2013-11-09 01:40:00 +04:00
$divider = trim($divider);
$divider = trim($divider, '|');
2013-11-09 01:40:00 +04:00
$dividerCells = explode('|', $divider);
2013-11-09 01:40:00 +04:00
foreach ($dividerCells as $dividerCell)
{
$dividerCell = trim($dividerCell);
2013-11-09 01:40:00 +04:00
if ($dividerCell === '')
{
return;
}
2013-11-09 01:40:00 +04:00
$alignment = null;
2013-11-09 01:40:00 +04:00
if ($dividerCell[0] === ':')
{
$alignment = 'left';
2014-04-17 11:59:35 +04:00
}
if (substr($dividerCell, - 1) === ':')
{
$alignment = $alignment === 'left' ? 'center' : 'right';
}
$alignments []= $alignment;
}
# ~
$HeaderElements = array();
2013-11-09 01:40:00 +04:00
$header = $Block['element']['handler']['argument'];
2014-02-05 16:03:23 +04:00
$header = trim($header);
$header = trim($header, '|');
2013-11-02 23:42:55 +04:00
$headerCells = explode('|', $header);
2014-02-05 16:03:23 +04:00
if (count($headerCells) !== count($alignments))
{
return;
}
2014-02-05 16:03:23 +04:00
foreach ($headerCells as $index => $headerCell)
{
$headerCell = trim($headerCell);
2014-02-05 16:03:23 +04:00
$HeaderElement = array(
'name' => 'th',
'handler' => array(
'function' => 'lineElements',
'argument' => $headerCell,
'destination' => 'elements',
)
);
2014-02-05 16:03:23 +04:00
if (isset($alignments[$index]))
{
$alignment = $alignments[$index];
2013-12-26 23:53:48 +04:00
$HeaderElement['attributes'] = array(
'style' => "text-align: $alignment;",
);
}
2014-04-17 11:59:35 +04:00
$HeaderElements []= $HeaderElement;
}
2014-04-17 11:59:35 +04:00
# ~
2014-04-17 11:59:35 +04:00
$Block = array(
'alignments' => $alignments,
'identified' => true,
'element' => array(
'name' => 'table',
'elements' => array(),
),
);
$Block['element']['elements'] []= array(
'name' => 'thead',
);
$Block['element']['elements'] []= array(
'name' => 'tbody',
'elements' => array(),
);
$Block['element']['elements'][0]['elements'] []= array(
'name' => 'tr',
'elements' => $HeaderElements,
);
2014-04-17 11:59:35 +04:00
return $Block;
2014-04-17 11:59:35 +04:00
}
2014-02-05 16:03:23 +04:00
2015-01-05 00:34:19 +03:00
protected function blockTableContinue($Line, array $Block)
2014-04-17 11:59:35 +04:00
{
2015-01-15 03:45:45 +03:00
if (isset($Block['interrupted']))
{
return;
}
if (count($Block['alignments']) === 1 or $Line['text'][0] === '|' or strpos($Line['text'], '|'))
2014-04-17 11:59:35 +04:00
{
$Elements = array();
2014-02-05 16:03:23 +04:00
2014-04-17 11:59:35 +04:00
$row = $Line['text'];
2014-02-05 16:03:23 +04:00
2014-04-17 11:59:35 +04:00
$row = trim($row);
$row = trim($row, '|');
2014-02-05 16:03:23 +04:00
2018-04-08 17:41:18 +03:00
preg_match_all('/(?:(\\\\[|])|[^|`]|`[^`]++`|`)++/', $row, $matches);
2014-02-06 04:36:22 +04:00
$cells = array_slice($matches[0], 0, count($Block['alignments']));
foreach ($cells as $index => $cell)
2014-04-17 11:59:35 +04:00
{
$cell = trim($cell);
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
$Element = array(
'name' => 'td',
'handler' => array(
'function' => 'lineElements',
'argument' => $cell,
'destination' => 'elements',
)
2014-04-17 11:59:35 +04:00
);
2014-02-05 16:03:23 +04:00
2014-04-17 11:59:35 +04:00
if (isset($Block['alignments'][$index]))
{
$Element['attributes'] = array(
'style' => "text-align: {$Block['alignments'][$index]};",
2014-04-17 11:59:35 +04:00
);
}
2014-04-17 11:59:35 +04:00
$Elements []= $Element;
}
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
$Element = array(
'name' => 'tr',
'elements' => $Elements,
2014-04-17 11:59:35 +04:00
);
$Block['element']['elements'][1]['elements'] []= $Element;
2014-04-17 11:59:35 +04:00
return $Block;
}
}
2014-04-17 11:59:35 +04:00
#
# ~
#
2013-11-09 01:40:00 +04:00
2015-01-05 00:34:19 +03:00
protected function paragraph($Line)
2014-05-11 23:31:02 +04:00
{
return array(
'type' => 'Paragraph',
2014-05-11 23:31:02 +04:00
'element' => array(
'name' => 'p',
'handler' => array(
'function' => 'lineElements',
'argument' => $Line['text'],
'destination' => 'elements',
),
2014-05-11 23:31:02 +04:00
),
);
}
protected function paragraphContinue($Line, array $Block)
{
if (isset($Block['interrupted']))
{
return;
}
$Block['element']['handler']['argument'] .= "\n".$Line['text'];
return $Block;
}
2014-04-17 11:59:35 +04:00
#
# Inline Elements
2014-04-17 11:59:35 +04:00
#
2013-11-09 01:40:00 +04:00
2015-01-05 00:34:19 +03:00
protected $InlineTypes = array(
2014-12-15 02:07:29 +03:00
'!' => array('Image'),
'&' => array('SpecialCharacter'),
2014-04-17 11:59:35 +04:00
'*' => array('Emphasis'),
':' => array('Url'),
'<' => array('UrlTag', 'EmailTag', 'Markup'),
2014-04-17 11:59:35 +04:00
'[' => array('Link'),
'_' => array('Emphasis'),
2015-01-05 00:34:19 +03:00
'`' => array('Code'),
2014-04-17 11:59:35 +04:00
'~' => array('Strikethrough'),
'\\' => array('EscapeSequence'),
);
2014-05-05 15:39:40 +04:00
# ~
protected $inlineMarkerList = '!*_&[:<`~\\';
2014-05-05 15:39:40 +04:00
#
# ~
#
2018-04-08 22:37:36 +03:00
public function line($text, $nonNestables = array())
2014-04-17 11:59:35 +04:00
{
2018-03-21 05:18:34 +03:00
return $this->elements($this->lineElements($text, $nonNestables));
}
protected function lineElements($text, $nonNestables = array())
{
$Elements = array();
$nonNestables = array_combine($nonNestables, $nonNestables);
2015-06-25 01:05:05 +03:00
# $excerpt is based on the first occurrence of a marker
2015-06-25 01:05:05 +03:00
while ($excerpt = strpbrk($text, $this->inlineMarkerList))
2014-04-17 11:59:35 +04:00
{
$marker = $excerpt[0];
2013-11-09 01:40:00 +04:00
$markerPosition = strlen($text) - strlen($excerpt);
$Excerpt = array('text' => $excerpt, 'context' => $text);
2013-11-09 01:40:00 +04:00
foreach ($this->InlineTypes[$marker] as $inlineType)
{
# check to see if the current inline type is nestable in the current context
if (isset($nonNestables[$inlineType]))
{
continue;
}
$Inline = $this->{"inline$inlineType"}($Excerpt);
2014-01-20 11:26:25 +04:00
if ( ! isset($Inline))
2014-04-17 11:59:35 +04:00
{
2014-04-28 03:10:18 +04:00
continue;
}
2014-01-20 11:26:25 +04:00
2015-06-25 01:05:05 +03:00
# makes sure that the inline belongs to "our" marker
if (isset($Inline['position']) and $Inline['position'] > $markerPosition)
{
continue;
}
2015-06-25 01:05:05 +03:00
# sets a default inline position
if ( ! isset($Inline['position']))
{
$Inline['position'] = $markerPosition;
}
# cause the new element to 'inherit' our non nestables
2018-04-08 22:37:36 +03:00
foreach ($nonNestables as $nonNestable)
{
2018-04-08 22:37:36 +03:00
$Inline['element']['nonNestables'][] = $nonNestable;
}
2015-06-25 01:05:05 +03:00
# the text that comes before the inline
$unmarkedText = substr($text, 0, $Inline['position']);
2015-06-25 01:05:05 +03:00
# compile the unmarked text
2018-03-21 05:18:34 +03:00
$InlineText = $this->inlineText($unmarkedText);
$Elements[] = $InlineText['element'];
2014-01-21 00:19:23 +04:00
2015-06-25 01:05:05 +03:00
# compile the inline
2018-03-21 05:18:34 +03:00
$Elements[] = $Inline['element'];
2014-01-21 00:19:23 +04:00
2015-06-25 01:05:05 +03:00
# remove the examined text
$text = substr($text, $Inline['position'] + $Inline['extent']);
2014-04-28 03:10:18 +04:00
continue 2;
2014-04-17 11:59:35 +04:00
}
2015-06-25 01:05:05 +03:00
# the marker does not belong to an inline
$unmarkedText = substr($text, 0, $markerPosition + 1);
2018-03-21 05:18:34 +03:00
$InlineText = $this->inlineText($unmarkedText);
$Elements[] = $InlineText['element'];
2013-07-11 00:22:16 +04:00
2015-06-25 01:05:05 +03:00
$text = substr($text, $markerPosition + 1);
2014-04-17 11:59:35 +04:00
}
2018-03-21 05:18:34 +03:00
$InlineText = $this->inlineText($text);
$Elements[] = $InlineText['element'];
2018-04-08 19:33:01 +03:00
foreach ($Elements as &$Element)
{
2018-04-08 22:37:36 +03:00
if ( ! isset($Element['autobreak']))
{
$Element['autobreak'] = false;
}
2018-04-08 19:33:01 +03:00
}
2018-03-21 05:18:34 +03:00
return $Elements;
2014-04-17 11:59:35 +04:00
}
2014-04-17 11:59:35 +04:00
#
# ~
#
2018-03-19 01:37:40 +03:00
protected function inlineText($text)
{
$Inline = array(
'extent' => strlen($text),
'element' => array(),
2018-03-19 01:37:40 +03:00
);
$safeText = self::escape($text, true);
2018-03-19 01:37:40 +03:00
if ($this->breaksEnabled)
{
$Inline['element']['rawHtml'] = preg_replace('/[ ]*+\n/', "<br />\n", $safeText);
$Inline['element']['allowRawHtmlInSafeMode'] = true;
2018-03-19 01:37:40 +03:00
}
else
{
$Inline['element']['rawHtml'] = preg_replace('/(?:[ ]*+\\\\|[ ]{2,}+)\n/', "<br />\n", $safeText);
$Inline['element']['allowRawHtmlInSafeMode'] = true;
2018-03-19 01:37:40 +03:00
}
return $Inline;
}
protected function inlineCode($Excerpt)
2014-12-15 01:52:03 +03:00
{
$marker = $Excerpt['text'][0];
2014-12-15 01:52:03 +03:00
2018-04-08 17:41:18 +03:00
if (preg_match('/^(['.$marker.']++)[ ]*+(.+?)[ ]*+(?<!['.$marker.'])\1(?!'.$marker.')/s', $Excerpt['text'], $matches))
2014-04-17 11:59:35 +04:00
{
2015-01-16 04:18:07 +03:00
$text = $matches[2];
2018-04-08 17:41:18 +03:00
$text = preg_replace('/[ ]*+\n/', ' ', $text);
2014-04-17 11:59:35 +04:00
return array(
'extent' => strlen($matches[0]),
'element' => array(
2015-01-16 04:18:07 +03:00
'name' => 'code',
'text' => $text,
2014-04-17 11:59:35 +04:00
),
);
}
}
protected function inlineEmailTag($Excerpt)
2014-04-17 11:59:35 +04:00
{
$hostnameLabel = '[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?';
$commonMarkEmail = '[a-zA-Z0-9.!#$%&\'*+\/=?^_`{|}~-]++@'
. $hostnameLabel . '(?:\.' . $hostnameLabel . ')*';
if (strpos($Excerpt['text'], '>') !== false
and preg_match("/^<((mailto:)?$commonMarkEmail)>/i", $Excerpt['text'], $matches)
){
2014-12-15 01:52:03 +03:00
$url = $matches[1];
if ( ! isset($matches[2]))
{
$url = "mailto:$url";
2014-12-15 01:52:03 +03:00
}
2014-04-17 11:59:35 +04:00
return array(
'extent' => strlen($matches[0]),
'element' => array(
'name' => 'a',
'text' => $matches[1],
'attributes' => array(
2014-12-15 01:52:03 +03:00
'href' => $url,
2014-04-17 11:59:35 +04:00
),
),
);
}
}
protected function inlineEmphasis($Excerpt)
2014-04-17 11:59:35 +04:00
{
if ( ! isset($Excerpt['text'][1]))
{
2014-09-22 03:36:42 +04:00
return;
}
$marker = $Excerpt['text'][0];
2015-01-16 04:18:07 +03:00
if ($Excerpt['text'][1] === $marker and preg_match($this->StrongRegex[$marker], $Excerpt['text'], $matches))
2015-01-16 02:24:02 +03:00
{
2015-01-16 04:18:07 +03:00
$emphasis = 'strong';
2015-01-16 02:24:02 +03:00
}
elseif (preg_match($this->EmRegex[$marker], $Excerpt['text'], $matches))
2014-04-17 11:59:35 +04:00
{
2015-01-16 04:18:07 +03:00
$emphasis = 'em';
2015-01-16 02:44:35 +03:00
}
2015-01-16 04:18:07 +03:00
else
2015-01-16 02:44:35 +03:00
{
2015-01-16 04:18:07 +03:00
return;
2014-04-17 11:59:35 +04:00
}
2015-01-16 04:18:07 +03:00
return array(
'extent' => strlen($matches[0]),
'element' => array(
'name' => $emphasis,
'handler' => array(
'function' => 'lineElements',
'argument' => $matches[1],
'destination' => 'elements',
)
2015-01-16 04:18:07 +03:00
),
);
2014-04-17 11:59:35 +04:00
}
protected function inlineEscapeSequence($Excerpt)
2014-04-17 11:59:35 +04:00
{
if (isset($Excerpt['text'][1]) and in_array($Excerpt['text'][1], $this->specialCharacters))
2014-04-17 11:59:35 +04:00
{
return array(
'element' => array('rawHtml' => $Excerpt['text'][1]),
2015-01-16 04:18:07 +03:00
'extent' => 2,
2014-04-17 11:59:35 +04:00
);
}
}
protected function inlineImage($Excerpt)
2014-12-15 02:07:29 +03:00
{
if ( ! isset($Excerpt['text'][1]) or $Excerpt['text'][1] !== '[')
2014-12-15 02:07:29 +03:00
{
return;
}
$Excerpt['text']= substr($Excerpt['text'], 1);
2014-12-15 02:07:29 +03:00
$Link = $this->inlineLink($Excerpt);
2014-12-15 02:07:29 +03:00
if ($Link === null)
2014-12-16 14:58:33 +03:00
{
return;
}
$Inline = array(
'extent' => $Link['extent'] + 1,
'element' => array(
'name' => 'img',
'attributes' => array(
'src' => $Link['element']['attributes']['href'],
'alt' => $Link['element']['handler']['argument'],
),
'autobreak' => true,
),
);
2014-12-15 02:07:29 +03:00
$Inline['element']['attributes'] += $Link['element']['attributes'];
unset($Inline['element']['attributes']['href']);
2014-12-15 02:07:29 +03:00
return $Inline;
2014-12-15 02:07:29 +03:00
}
protected function inlineLink($Excerpt)
2014-04-17 11:59:35 +04:00
{
2014-12-15 02:07:29 +03:00
$Element = array(
'name' => 'a',
'handler' => array(
'function' => 'lineElements',
'argument' => null,
'destination' => 'elements',
),
'nonNestables' => array('Url', 'Link'),
2014-12-15 02:07:29 +03:00
'attributes' => array(
'href' => null,
'title' => null,
),
);
$extent = 0;
$remainder = $Excerpt['text'];
2014-12-15 02:07:29 +03:00
if (preg_match('/\[((?:[^][]++|(?R))*+)\]/', $remainder, $matches))
2014-04-17 11:59:35 +04:00
{
$Element['handler']['argument'] = $matches[1];
2014-04-17 11:59:35 +04:00
$extent += strlen($matches[0]);
2014-01-20 00:49:43 +04:00
2014-12-15 02:07:29 +03:00
$remainder = substr($remainder, $extent);
}
else
{
return;
}
2014-01-20 00:49:43 +04:00
2018-04-08 17:41:18 +03:00
if (preg_match('/^[(]\s*+((?:[^ ()]++|[(][^ )]+[)])++)(?:[ ]+("[^"]*+"|\'[^\']*+\'))?\s*+[)]/', $remainder, $matches))
2014-12-15 02:07:29 +03:00
{
$Element['attributes']['href'] = $matches[1];
2014-01-20 00:49:43 +04:00
2014-12-15 02:07:29 +03:00
if (isset($matches[2]))
2014-04-17 11:59:35 +04:00
{
2014-12-15 02:07:29 +03:00
$Element['attributes']['title'] = substr($matches[2], 1, - 1);
2014-04-17 11:59:35 +04:00
}
2014-12-15 02:07:29 +03:00
$extent += strlen($matches[0]);
}
else
{
if (preg_match('/^\s*\[(.*?)\]/', $remainder, $matches))
{
$definition = strlen($matches[1]) ? $matches[1] : $Element['handler']['argument'];
2014-12-15 02:07:29 +03:00
$definition = strtolower($definition);
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
$extent += strlen($matches[0]);
}
else
2014-12-15 02:07:29 +03:00
{
$definition = strtolower($Element['handler']['argument']);
2014-12-15 02:07:29 +03:00
}
if ( ! isset($this->DefinitionData['Reference'][$definition]))
2014-04-17 11:59:35 +04:00
{
return;
}
$Definition = $this->DefinitionData['Reference'][$definition];
2013-11-09 01:40:00 +04:00
2014-12-15 02:07:29 +03:00
$Element['attributes']['href'] = $Definition['url'];
$Element['attributes']['title'] = $Definition['title'];
2014-04-17 11:59:35 +04:00
}
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
return array(
'extent' => $extent,
'element' => $Element,
);
}
2013-11-09 01:40:00 +04:00
protected function inlineMarkup($Excerpt)
2014-04-17 11:59:35 +04:00
{
if ($this->markupEscaped or $this->safeMode or strpos($Excerpt['text'], '>') === false)
2014-04-17 11:59:35 +04:00
{
return;
}
2013-11-09 01:40:00 +04:00
2018-04-08 17:41:18 +03:00
if ($Excerpt['text'][1] === '/' and preg_match('/^<\/\w[\w-]*+[ ]*+>/s', $Excerpt['text'], $matches))
2014-04-17 11:59:35 +04:00
{
2015-01-16 04:18:07 +03:00
return array(
'element' => array('rawHtml' => $matches[0]),
2015-01-16 04:18:07 +03:00
'extent' => strlen($matches[0]),
);
2014-04-17 11:59:35 +04:00
}
2015-01-16 04:18:07 +03:00
2018-04-08 17:41:18 +03:00
if ($Excerpt['text'][1] === '!' and preg_match('/^<!---?[^>-](?:-?+[^-])*-->/s', $Excerpt['text'], $matches))
2014-04-17 11:59:35 +04:00
{
2015-01-16 04:18:07 +03:00
return array(
'element' => array('rawHtml' => $matches[0]),
2015-01-16 04:18:07 +03:00
'extent' => strlen($matches[0]),
);
2014-04-17 11:59:35 +04:00
}
2015-01-16 04:18:07 +03:00
2018-04-08 17:41:18 +03:00
if ($Excerpt['text'][1] !== ' ' and preg_match('/^<\w[\w-]*+(?:[ ]*+'.$this->regexHtmlAttribute.')*+[ ]*+\/?>/s', $Excerpt['text'], $matches))
2014-04-17 11:59:35 +04:00
{
2015-01-16 04:18:07 +03:00
return array(
'element' => array('rawHtml' => $matches[0]),
2015-01-16 04:18:07 +03:00
'extent' => strlen($matches[0]),
);
2014-04-17 11:59:35 +04:00
}
2015-01-16 04:18:07 +03:00
}
2013-11-22 02:20:45 +04:00
protected function inlineSpecialCharacter($Excerpt)
2015-01-16 04:18:07 +03:00
{
if (preg_match('/^&(#?+[0-9a-zA-Z]++);/', $Excerpt['text'], $matches))
{
return array(
'element' => array('rawHtml' => "&{$matches[1]};"),
'extent' => strlen($matches[0]),
);
}
return;
2014-04-17 11:59:35 +04:00
}
2013-11-22 02:20:45 +04:00
protected function inlineStrikethrough($Excerpt)
2015-01-16 04:18:07 +03:00
{
if ( ! isset($Excerpt['text'][1]))
2015-01-16 04:18:07 +03:00
{
return;
}
if ($Excerpt['text'][1] === '~' and preg_match('/^~~(?=\S)(.+?)(?<=\S)~~/', $Excerpt['text'], $matches))
2015-01-16 04:18:07 +03:00
{
return array(
'extent' => strlen($matches[0]),
'element' => array(
'name' => 'del',
'handler' => array(
'function' => 'lineElements',
'argument' => $matches[1],
'destination' => 'elements',
)
2015-01-16 04:18:07 +03:00
),
);
}
}
protected function inlineUrl($Excerpt)
2015-01-16 04:18:07 +03:00
{
if ($this->urlsLinked !== true or ! isset($Excerpt['text'][2]) or $Excerpt['text'][2] !== '/')
{
return;
}
2018-04-08 17:41:18 +03:00
if (preg_match('/\bhttps?+:[\/]{2}[^\s<]+\b\/*+/ui', $Excerpt['context'], $matches, PREG_OFFSET_CAPTURE))
{
$url = $matches[0][0];
$Inline = array(
'extent' => strlen($matches[0][0]),
'position' => $matches[0][1],
'element' => array(
'name' => 'a',
'text' => $url,
'attributes' => array(
'href' => $url,
),
),
);
return $Inline;
}
}
protected function inlineUrlTag($Excerpt)
{
2018-04-08 17:41:18 +03:00
if (strpos($Excerpt['text'], '>') !== false and preg_match('/^<(\w++:\/{2}[^ >]++)>/i', $Excerpt['text'], $matches))
2015-01-16 04:18:07 +03:00
{
$url = $matches[1];
2015-01-16 04:18:07 +03:00
return array(
'extent' => strlen($matches[0]),
'element' => array(
'name' => 'a',
'text' => $url,
'attributes' => array(
'href' => $url,
),
),
);
}
}
# ~
protected function unmarkedText($text)
{
2018-03-19 02:06:26 +03:00
$Inline = $this->inlineText($text);
return $this->element($Inline['element']);
2014-04-17 11:59:35 +04:00
}
2013-11-09 01:40:00 +04:00
2014-04-17 11:59:35 +04:00
#
2015-01-16 04:18:07 +03:00
# Handlers
2014-04-17 11:59:35 +04:00
#
2013-11-21 15:39:00 +04:00
protected function handle(array $Element)
{
if (isset($Element['handler']))
{
if (!isset($Element['nonNestables']))
{
$Element['nonNestables'] = array();
}
2018-03-21 20:31:40 +03:00
if (is_string($Element['handler']))
{
$function = $Element['handler'];
$argument = $Element['text'];
$destination = 'rawHtml';
}
else
{
$function = $Element['handler']['function'];
$argument = $Element['handler']['argument'];
$destination = $Element['handler']['destination'];
}
$Element[$destination] = $this->{$function}($argument, $Element['nonNestables']);
2018-03-31 14:07:53 +03:00
if ($destination === 'handler')
{
$Element = $this->handle($Element);
}
2018-04-08 19:38:09 +03:00
unset($Element['handler']);
}
return $Element;
}
protected function handleElementRecursive(array $Element)
{
return $this->elementApplyRecursive(array($this, 'handle'), $Element);
}
protected function handleElementsRecursive(array $Elements)
{
return $this->elementsApplyRecursive(array($this, 'handle'), $Elements);
}
protected function elementApplyRecursive($closure, array $Element)
{
$Element = call_user_func($closure, $Element);
if (isset($Element['elements']))
{
$Element['elements'] = $this->elementsApplyRecursive($closure, $Element['elements']);
}
elseif (isset($Element['element']))
{
$Element['element'] = $this->elementApplyRecursive($closure, $Element['element']);
}
return $Element;
}
protected function elementApplyRecursiveDepthFirst($closure, array $Element)
{
if (isset($Element['elements']))
{
2018-04-08 20:40:50 +03:00
$Element['elements'] = $this->elementsApplyRecursiveDepthFirst($closure, $Element['elements']);
}
elseif (isset($Element['element']))
{
2018-04-08 20:40:50 +03:00
$Element['element'] = $this->elementsApplyRecursiveDepthFirst($closure, $Element['element']);
}
$Element = call_user_func($closure, $Element);
return $Element;
}
protected function elementsApplyRecursive($closure, array $Elements)
{
foreach ($Elements as &$Element)
{
$Element = $this->elementApplyRecursive($closure, $Element);
}
return $Elements;
}
protected function elementsApplyRecursiveDepthFirst($closure, array $Elements)
{
foreach ($Elements as &$Element)
{
$Element = $this->elementApplyRecursiveDepthFirst($closure, $Element);
}
return $Elements;
}
2015-01-16 04:18:07 +03:00
protected function element(array $Element)
{
if ($this->safeMode)
{
$Element = $this->sanitiseElement($Element);
}
# identity map if element has no handler
$Element = $this->handle($Element);
2017-06-13 22:28:32 +03:00
$hasName = isset($Element['name']);
2015-01-16 04:18:07 +03:00
2017-06-13 22:28:32 +03:00
$markup = '';
if ($hasName)
2015-01-16 04:18:07 +03:00
{
$markup .= "<{$Element['name']}";
2017-06-13 22:28:32 +03:00
if (isset($Element['attributes']))
2015-01-16 04:18:07 +03:00
{
2017-06-13 22:28:32 +03:00
foreach ($Element['attributes'] as $name => $value)
2015-01-16 04:18:07 +03:00
{
2017-06-13 22:28:32 +03:00
if ($value === null)
{
continue;
}
2015-01-16 04:18:07 +03:00
$markup .= " $name=\"".self::escape($value).'"';
2017-06-13 22:28:32 +03:00
}
2015-01-16 04:18:07 +03:00
}
}
$permitRawHtml = false;
2015-01-16 04:18:07 +03:00
if (isset($Element['text']))
{
$text = $Element['text'];
}
// very strongly consider an alternative if you're writing an
// extension
elseif (isset($Element['rawHtml']))
{
$text = $Element['rawHtml'];
$allowRawHtmlInSafeMode = isset($Element['allowRawHtmlInSafeMode']) && $Element['allowRawHtmlInSafeMode'];
$permitRawHtml = !$this->safeMode || $allowRawHtmlInSafeMode;
}
$hasContent = isset($text) || isset($Element['element']) || isset($Element['elements']);
if ($hasContent)
2015-01-16 04:18:07 +03:00
{
2017-06-13 22:28:32 +03:00
$markup .= $hasName ? '>' : '';
2015-01-16 04:18:07 +03:00
if (isset($Element['elements']))
2015-01-16 04:18:07 +03:00
{
$markup .= $this->elements($Element['elements']);
}
elseif (isset($Element['element']))
{
$markup .= $this->element($Element['element']);
2015-01-16 04:18:07 +03:00
}
else
{
if (!$permitRawHtml)
{
$markup .= self::escape($text, true);
}
else
{
$markup .= $text;
}
2015-01-16 04:18:07 +03:00
}
$markup .= $hasName ? "</{$Element['name']}>" : '';
2015-01-16 04:18:07 +03:00
}
2017-06-13 22:28:32 +03:00
elseif ($hasName)
2015-01-16 04:18:07 +03:00
{
$markup .= ' />';
}
return $markup;
}
protected function elements(array $Elements)
{
$markup = '';
$autoBreak = true;
2015-01-16 04:18:07 +03:00
foreach ($Elements as $Element)
{
if (empty($Element))
{
continue;
}
2018-04-08 22:37:36 +03:00
$autoBreakNext = (isset($Element['autobreak'])
? $Element['autobreak'] : isset($Element['name'])
2018-03-21 05:18:34 +03:00
);
// (autobreak === false) covers both sides of an element
2018-03-21 05:18:34 +03:00
$autoBreak = !$autoBreak ? $autoBreak : $autoBreakNext;
$markup .= ($autoBreak ? "\n" : '') . $this->element($Element);
2018-03-21 05:18:34 +03:00
$autoBreak = $autoBreakNext;
2015-01-16 04:18:07 +03:00
}
$markup .= $autoBreak ? "\n" : '';
2015-01-16 04:18:07 +03:00
return $markup;
}
# ~
2014-04-17 11:59:35 +04:00
protected function li($lines)
{
$Elements = $this->linesElements($lines);
if ( ! in_array('', $lines)
and isset($Elements[0]) and isset($Elements[0]['name'])
and $Elements[0]['name'] === 'p'
) {
unset($Elements[0]['name']);
}
2013-11-09 01:40:00 +04:00
return $Elements;
}
2014-01-18 17:10:24 +04:00
2018-03-19 01:37:40 +03:00
#
# AST Convenience
#
/**
* Replace occurrences $regexp with $Elements in $text. Return an array of
* elements representing the replacement.
*/
protected static function pregReplaceElements($regexp, $Elements, $text)
{
$newElements = array();
while (preg_match($regexp, $text, $matches, PREG_OFFSET_CAPTURE))
{
$offset = $matches[0][1];
$before = substr($text, 0, $offset);
$after = substr($text, $offset + strlen($matches[0][0]));
$newElements[] = array('text' => $before);
foreach ($Elements as $Element)
{
$newElements[] = $Element;
}
$text = $after;
}
$newElements[] = array('text' => $text);
return $newElements;
}
2014-02-21 04:22:31 +04:00
#
2015-01-16 04:18:07 +03:00
# Deprecated Methods
#
function parse($text)
{
$markup = $this->text($text);
return $markup;
}
protected function sanitiseElement(array $Element)
{
static $goodAttribute = '/^[a-zA-Z0-9][a-zA-Z0-9-_]*+$/';
static $safeUrlNameToAtt = array(
'a' => 'href',
'img' => 'src',
);
if ( ! isset($Element['name']))
{
unset($Element['attributes']);
return $Element;
}
if (isset($safeUrlNameToAtt[$Element['name']]))
{
$Element = $this->filterUnsafeUrlInAttribute($Element, $safeUrlNameToAtt[$Element['name']]);
}
2017-05-01 17:44:04 +03:00
if ( ! empty($Element['attributes']))
{
foreach ($Element['attributes'] as $att => $val)
{
# filter out badly parsed attribute
if ( ! preg_match($goodAttribute, $att))
{
unset($Element['attributes'][$att]);
}
# dump onevent attribute
elseif (self::striAtStart($att, 'on'))
{
unset($Element['attributes'][$att]);
}
}
2017-05-01 17:44:04 +03:00
}
return $Element;
}
protected function filterUnsafeUrlInAttribute(array $Element, $attribute)
{
foreach ($this->safeLinksWhitelist as $scheme)
{
if (self::striAtStart($Element['attributes'][$attribute], $scheme))
{
return $Element;
}
}
$Element['attributes'][$attribute] = str_replace(':', '%3A', $Element['attributes'][$attribute]);
return $Element;
}
2015-01-16 04:18:07 +03:00
#
# Static Methods
2014-04-17 11:59:35 +04:00
#
2014-02-21 04:22:31 +04:00
protected static function escape($text, $allowQuotes = false)
{
return htmlspecialchars($text, $allowQuotes ? ENT_NOQUOTES : ENT_QUOTES, 'UTF-8');
}
protected static function striAtStart($string, $needle)
{
$len = strlen($needle);
if ($len > strlen($string))
{
return false;
}
else
{
return strtolower(substr($string, 0, $len)) === strtolower($needle);
}
}
2014-02-21 04:22:31 +04:00
static function instance($name = 'default')
{
if (isset(self::$instances[$name]))
{
return self::$instances[$name];
}
$instance = new static();
2014-02-21 04:22:31 +04:00
self::$instances[$name] = $instance;
return $instance;
}
private static $instances = array();
#
2014-01-26 21:53:24 +04:00
# Fields
#
protected $DefinitionData;
2014-01-26 21:53:24 +04:00
#
2015-01-16 04:18:07 +03:00
# Read-Only
2014-01-18 17:10:24 +04:00
2014-04-17 11:59:35 +04:00
protected $specialCharacters = array(
2015-01-15 22:32:18 +03:00
'\\', '`', '*', '_', '{', '}', '[', ']', '(', ')', '>', '#', '+', '-', '.', '!', '|',
2014-04-17 11:59:35 +04:00
);
2014-05-10 17:28:00 +04:00
protected $StrongRegex = array(
2018-04-08 17:41:18 +03:00
'*' => '/^[*]{2}((?:\\\\\*|[^*]|[*][^*]*+[*])+?)[*]{2}(?![*])/s',
'_' => '/^__((?:\\\\_|[^_]|_[^_]*+_)+?)__(?!_)/us',
);
2014-01-18 17:10:24 +04:00
2014-05-10 17:28:00 +04:00
protected $EmRegex = array(
2014-11-28 15:03:12 +03:00
'*' => '/^[*]((?:\\\\\*|[^*]|[*][*][^*]+?[*][*])+?)[*](?![*])/s',
'_' => '/^_((?:\\\\_|[^_]|__[^_]*__)+?)_(?!_)\b/us',
);
2014-01-27 02:58:18 +04:00
2018-04-08 17:41:18 +03:00
protected $regexHtmlAttribute = '[a-zA-Z_:][\w:.-]*+(?:\s*+=\s*+(?:[^"\'=<>`\s]+|"[^"]*+"|\'[^\']*+\'))?+';
2014-10-29 23:29:23 +03:00
protected $voidElements = array(
'area', 'base', 'br', 'col', 'command', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source',
);
2014-04-17 11:59:35 +04:00
protected $textLevelElements = array(
2014-01-27 02:58:18 +04:00
'a', 'br', 'bdo', 'abbr', 'blink', 'nextid', 'acronym', 'basefont',
'b', 'em', 'big', 'cite', 'small', 'spacer', 'listing',
2014-05-01 03:02:14 +04:00
'i', 'rp', 'del', 'code', 'strike', 'marquee',
'q', 'rt', 'ins', 'font', 'strong',
2017-03-29 19:04:15 +03:00
's', 'tt', 'kbd', 'mark',
'u', 'xm', 'sub', 'nobr',
'sup', 'ruby',
'var', 'span',
'wbr', 'time',
2014-01-27 02:58:18 +04:00
);
}