1<?php
2/**
3 * @copyright Copyright (c) 2014 Carsten Brandt
4 * @license https://github.com/cebe/markdown/blob/master/LICENSE
5 * @link https://github.com/cebe/markdown#readme
6 */
7
8namespace cebe\markdown;
9
10/**
11 * Markdown parser for the [initial markdown spec](http://daringfireball.net/projects/markdown/syntax).
12 *
13 * @author Carsten Brandt <mail@cebe.cc>
14 */
15class Markdown extends Parser
16{
17	// include block element parsing using traits
18	use block\CodeTrait;
19	use block\HeadlineTrait;
20	use block\HtmlTrait {
21		parseInlineHtml as private;
22	}
23	use block\ListTrait {
24		// Check Ul List before headline
25		identifyUl as protected identifyBUl;
26		consumeUl as protected consumeBUl;
27	}
28	use block\QuoteTrait;
29	use block\RuleTrait {
30		// Check Hr before checking lists
31		identifyHr as protected identifyAHr;
32		consumeHr as protected consumeAHr;
33	}
34
35	// include inline element parsing using traits
36	use inline\CodeTrait;
37	use inline\EmphStrongTrait;
38	use inline\LinkTrait;
39
40	/**
41	 * @var boolean whether to format markup according to HTML5 spec.
42	 * Defaults to `false` which means that markup is formatted as HTML4.
43	 */
44	public $html5 = false;
45
46	/**
47	 * @var array these are "escapeable" characters. When using one of these prefixed with a
48	 * backslash, the character will be outputted without the backslash and is not interpreted
49	 * as markdown.
50	 */
51	protected $escapeCharacters = [
52		'\\', // backslash
53		'`', // backtick
54		'*', // asterisk
55		'_', // underscore
56		'{', '}', // curly braces
57		'[', ']', // square brackets
58		'(', ')', // parentheses
59		'#', // hash mark
60		'+', // plus sign
61		'-', // minus sign (hyphen)
62		'.', // dot
63		'!', // exclamation mark
64		'<', '>',
65	];
66
67
68	/**
69	 * @inheritDoc
70	 */
71	protected function prepare()
72	{
73		// reset references
74		$this->references = [];
75	}
76
77	/**
78	 * Consume lines for a paragraph
79	 *
80	 * Allow headlines and code to break paragraphs
81	 */
82	protected function consumeParagraph($lines, $current)
83	{
84		// consume until newline
85		$content = [];
86		for ($i = $current, $count = count($lines); $i < $count; $i++) {
87			$line = $lines[$i];
88
89			// a list may break a paragraph when it is inside of a list
90			if (isset($this->context[1]) && $this->context[1] === 'list' && !ctype_alpha($line[0]) && (
91				$this->identifyUl($line, $lines, $i) || $this->identifyOl($line, $lines, $i))) {
92				break;
93			}
94
95			if ($line === '' || ltrim($line) === '' || $this->identifyHeadline($line, $lines, $i)) {
96				break;
97			} elseif ($line[0] === "\t" || $line[0] === " " && strncmp($line, '    ', 4) === 0) {
98				// possible beginning of a code block
99				// but check for continued inline HTML
100				// e.g. <img src="file.jpg"
101				//           alt="some alt aligned with src attribute" title="some text" />
102				if (preg_match('~<\w+([^>]+)$~s', implode("\n", $content))) {
103					$content[] = $line;
104				} else {
105					break;
106				}
107			} else {
108				$content[] = $line;
109			}
110		}
111		$block = [
112			'paragraph',
113			'content' => $this->parseInline(implode("\n", $content)),
114		];
115		return [$block, --$i];
116	}
117
118
119	/**
120	 * @inheritdocs
121	 *
122	 * Parses a newline indicated by two spaces on the end of a markdown line.
123	 */
124	protected function renderText($text)
125	{
126		return str_replace("  \n", $this->html5 ? "<br>\n" : "<br />\n", $text[1]);
127	}
128}
129