| 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 |
|
| 8 |
namespace cebe\markdownparser\inline; |
| 9 |
|
| 10 |
// work around https://github.com/facebook/hhvm/issues/1120 |
| 11 |
defined('ENT_HTML401') || define('ENT_HTML401', 0); |
| 12 |
|
| 13 |
/** |
| 14 |
* Addes links and images as well as url markers. |
| 15 |
* |
| 16 |
* This trait conflicts with the HtmlTrait. If both are used together, |
| 17 |
* you have to define a resolution, by defining the HtmlTrait::parseInlineHtml |
| 18 |
* as private so it is not used directly: |
| 19 |
* |
| 20 |
* ```php |
| 21 |
* use block\HtmlTrait { |
| 22 |
* parseInlineHtml as private parseInlineHtml; |
| 23 |
* } |
| 24 |
* ``` |
| 25 |
* |
| 26 |
* If the method exists it is called internally by this trait. |
| 27 |
* |
| 28 |
* Also make sure to reset references on prepare(): |
| 29 |
* |
| 30 |
* ```php |
| 31 |
* protected function prepare() |
| 32 |
* { |
| 33 |
* // reset references |
| 34 |
* $this->references = []; |
| 35 |
* } |
| 36 |
* ``` |
| 37 |
*/ |
| 38 |
trait LinkTrait |
| 39 |
{ |
| 40 |
/** |
| 41 |
* @var array a list of defined references in this document. |
| 42 |
*/ |
| 43 |
protected $references = []; |
| 44 |
|
| 45 |
/** |
| 46 |
* Remove backslash from escaped characters |
| 47 |
* @param $text |
| 48 |
* @return string |
| 49 |
*/ |
| 50 |
protected function replaceEscape($text) |
| 51 |
{ |
| 52 |
$strtr = []; |
| 53 |
foreach($this->escapeCharacters as $char) { |
| 54 |
$strtr["\\$char"] = $char; |
| 55 |
} |
| 56 |
return strtr($text, $strtr); |
| 57 |
} |
| 58 |
|
| 59 |
/** |
| 60 |
* Parses a link indicated by `[`. |
| 61 |
* @marker [ |
| 62 |
*/ |
| 63 |
protected function parseLink($markdown) |
| 64 |
{ |
| 65 |
if (!in_array('parseLink', array_slice($this->context, 1)) && ($parts = $this->parseLinkOrImage($markdown)) !== false) { |
| 66 |
list($text, $url, $title, $offset, $key) = $parts; |
| 67 |
return [ |
| 68 |
[ |
| 69 |
'link', |
| 70 |
'text' => $this->parseInline($text), |
| 71 |
'url' => $url, |
| 72 |
'title' => $title, |
| 73 |
'refkey' => $key, |
| 74 |
'orig' => substr($markdown, 0, $offset), |
| 75 |
], |
| 76 |
$offset |
| 77 |
]; |
| 78 |
} else { |
| 79 |
// remove all starting [ markers to avoid next one to be parsed as link |
| 80 |
$result = '['; |
| 81 |
$i = 1; |
| 82 |
while (isset($markdown[$i]) && $markdown[$i] === '[') { |
| 83 |
$result .= '['; |
| 84 |
$i++; |
| 85 |
} |
| 86 |
return [['text', $result], $i]; |
| 87 |
} |
| 88 |
} |
| 89 |
|
| 90 |
/** |
| 91 |
* Parses an image indicated by `![`. |
| 92 |
* @marker ![ |
| 93 |
*/ |
| 94 |
protected function parseImage($markdown) |
| 95 |
{ |
| 96 |
if (($parts = $this->parseLinkOrImage(substr($markdown, 1))) !== false) { |
| 97 |
list($text, $url, $title, $offset, $key) = $parts; |
| 98 |
|
| 99 |
return [ |
| 100 |
[ |
| 101 |
'image', |
| 102 |
'text' => $text, |
| 103 |
'url' => $url, |
| 104 |
'title' => $title, |
| 105 |
'refkey' => $key, |
| 106 |
'orig' => substr($markdown, 0, $offset + 1), |
| 107 |
], |
| 108 |
$offset + 1 |
| 109 |
]; |
| 110 |
} else { |
| 111 |
// remove all starting [ markers to avoid next one to be parsed as link |
| 112 |
$result = '!'; |
| 113 |
$i = 1; |
| 114 |
while (isset($markdown[$i]) && $markdown[$i] === '[') { |
| 115 |
$result .= '['; |
| 116 |
$i++; |
| 117 |
} |
| 118 |
return [['text', $result], $i]; |
| 119 |
} |
| 120 |
} |
| 121 |
|
| 122 |
protected function parseLinkOrImage($markdown) |
| 123 |
{ |
| 124 |
if (strpos($markdown, ']') !== false && preg_match('/\[((?>[^\]\[]+|(?R))*)\]/', $markdown, $textMatches)) { // TODO improve bracket regex |
| 125 |
$text = $textMatches[1]; |
| 126 |
$offset = strlen($textMatches[0]); |
| 127 |
$markdown = substr($markdown, $offset); |
| 128 |
|
| 129 |
$pattern = <<<REGEXP |
| 130 |
/(?(R) # in case of recursion match parentheses |
| 131 |
\(((?>[^\s()]+)|(?R))*\) |
| 132 |
| # else match a link with title |
| 133 |
^\(\s*(((?>[^\s()]+)|(?R))*)(\s+"(.*?)")?\s*\) |
| 134 |
)/x |
| 135 |
REGEXP; |
| 136 |
if (preg_match($pattern, $markdown, $refMatches)) { |
| 137 |
// inline link |
| 138 |
return [ |
| 139 |
$text, |
| 140 |
isset($refMatches[2]) ? $this->replaceEscape($refMatches[2]) : '', // url |
| 141 |
empty($refMatches[5]) ? null: $refMatches[5], // title |
| 142 |
$offset + strlen($refMatches[0]), // offset |
| 143 |
null, // reference key |
| 144 |
]; |
| 145 |
} elseif (preg_match('/^([ \n]?\[(.*?)\])?/s', $markdown, $refMatches)) { |
| 146 |
// reference style link |
| 147 |
if (empty($refMatches[2])) { |
| 148 |
$key = strtolower($text); |
| 149 |
} else { |
| 150 |
$key = strtolower($refMatches[2]); |
| 151 |
} |
| 152 |
return [ |
| 153 |
$text, |
| 154 |
null, // url |
| 155 |
null, // title |
| 156 |
$offset + strlen($refMatches[0]), // offset |
| 157 |
$key, |
| 158 |
]; |
| 159 |
} |
| 160 |
} |
| 161 |
return false; |
| 162 |
} |
| 163 |
|
| 164 |
/** |
| 165 |
* Parses inline HTML. |
| 166 |
* @marker < |
| 167 |
*/ |
| 168 |
protected function parseLt($text) |
| 169 |
{ |
| 170 |
if (strpos($text, '>') !== false) { |
| 171 |
if (!in_array('parseLink', $this->context)) { // do not allow links in links |
| 172 |
if (preg_match('/^<([^\s>]*?@[^\s]*?\.\w+?)>/', $text, $matches)) { |
| 173 |
// email address |
| 174 |
return [ |
| 175 |
['email', $this->replaceEscape($matches[1])], |
| 176 |
strlen($matches[0]) |
| 177 |
]; |
| 178 |
} elseif (preg_match('/^<([a-z]{3,}:\/\/[^\s]+?)>/', $text, $matches)) { |
| 179 |
// URL |
| 180 |
return [ |
| 181 |
['url', $this->replaceEscape($matches[1])], |
| 182 |
strlen($matches[0]) |
| 183 |
]; |
| 184 |
} |
| 185 |
} |
| 186 |
// try inline HTML if it was neither a URL nor email if HtmlTrait is included. |
| 187 |
if (method_exists($this, 'parseInlineHtml')) { |
| 188 |
return $this->parseInlineHtml($text); |
| 189 |
} |
| 190 |
} |
| 191 |
return [['text', '<'], 1]; |
| 192 |
} |
| 193 |
|
| 194 |
protected function renderEmail($block) |
| 195 |
{ |
| 196 |
$email = htmlspecialchars($block[1], ENT_NOQUOTES | ENT_SUBSTITUTE, 'UTF-8'); |
| 197 |
return "<a href=\"mailto:$email\">$email</a>"; |
| 198 |
} |
| 199 |
|
| 200 |
protected function renderUrl($block) |
| 201 |
{ |
| 202 |
$url = htmlspecialchars($block[1], ENT_COMPAT | ENT_HTML401, 'UTF-8'); |
| 203 |
$decodedUrl = urldecode($block[1]); |
| 204 |
$secureUrlText = preg_match('//u', $decodedUrl) ? $decodedUrl : $block[1]; |
| 205 |
$text = htmlspecialchars($secureUrlText, ENT_NOQUOTES | ENT_SUBSTITUTE, 'UTF-8'); |
| 206 |
return "<a href=\"$url\">$text</a>"; |
| 207 |
} |
| 208 |
|
| 209 |
protected function lookupReference($key) |
| 210 |
{ |
| 211 |
$normalizedKey = preg_replace('/\s+/', ' ', $key); |
| 212 |
if (isset($this->references[$key]) || isset($this->references[$key = $normalizedKey])) { |
| 213 |
return $this->references[$key]; |
| 214 |
} |
| 215 |
return false; |
| 216 |
} |
| 217 |
|
| 218 |
protected function renderLink($block) |
| 219 |
{ |
| 220 |
if (isset($block['refkey'])) { |
| 221 |
if (($ref = $this->lookupReference($block['refkey'])) !== false) { |
| 222 |
$block = array_merge($block, $ref); |
| 223 |
} else { |
| 224 |
if (strncmp($block['orig'], '[', 1) === 0) { |
| 225 |
return '[' . $this->renderAbsy($this->parseInline(substr($block['orig'], 1))); |
| 226 |
} |
| 227 |
return $block['orig']; |
| 228 |
} |
| 229 |
} |
| 230 |
return '<a href="' . htmlspecialchars($block['url'], ENT_COMPAT | ENT_HTML401, 'UTF-8') . '"' |
| 231 |
. (empty($block['title']) ? '' : ' title="' . htmlspecialchars($block['title'], ENT_COMPAT | ENT_HTML401 | ENT_SUBSTITUTE, 'UTF-8') . '"') |
| 232 |
. '>' . $this->renderAbsy($block['text']) . '</a>'; |
| 233 |
} |
| 234 |
|
| 235 |
protected function renderImage($block) |
| 236 |
{ |
| 237 |
if (isset($block['refkey'])) { |
| 238 |
if (($ref = $this->lookupReference($block['refkey'])) !== false) { |
| 239 |
$block = array_merge($block, $ref); |
| 240 |
} else { |
| 241 |
if (strncmp($block['orig'], '![', 2) === 0) { |
| 242 |
return '![' . $this->renderAbsy($this->parseInline(substr($block['orig'], 2))); |
| 243 |
} |
| 244 |
return $block['orig']; |
| 245 |
} |
| 246 |
} |
| 247 |
return '<img src="' . htmlspecialchars($block['url'], ENT_COMPAT | ENT_HTML401, 'UTF-8') . '"' |
| 248 |
. ' alt="' . htmlspecialchars($block['text'], ENT_COMPAT | ENT_HTML401 | ENT_SUBSTITUTE, 'UTF-8') . '"' |
| 249 |
. (empty($block['title']) ? '' : ' title="' . htmlspecialchars($block['title'], ENT_COMPAT | ENT_HTML401 | ENT_SUBSTITUTE, 'UTF-8') . '"') |
| 250 |
. ($this->html5 ? '>' : ' />'); |
| 251 |
} |
| 252 |
|
| 253 |
// references |
| 254 |
|
| 255 |
protected function identifyReference($line) |
| 256 |
{ |
| 257 |
return isset($line[0]) && ($line[0] === ' ' || $line[0] === '[') && preg_match('/^ {0,3}\[[^\[](.*?)\]:\s*([^\s]+?)(?:\s+[\'"](.+?)[\'"])?\s*$/', $line); |
| 258 |
} |
| 259 |
|
| 260 |
/** |
| 261 |
* Consume link references |
| 262 |
*/ |
| 263 |
protected function consumeReference($lines, $current) |
| 264 |
{ |
| 265 |
while (isset($lines[$current]) && preg_match('/^ {0,3}\[(.+?)\]:\s*(.+?)(?:\s+[\(\'"](.+?)[\)\'"])?\s*$/', $lines[$current], $matches)) { |
| 266 |
$label = strtolower($matches[1]); |
| 267 |
|
| 268 |
$this->references[$label] = [ |
| 269 |
'url' => $this->replaceEscape($matches[2]), |
| 270 |
]; |
| 271 |
if (isset($matches[3])) { |
| 272 |
$this->references[$label]['title'] = $matches[3]; |
| 273 |
} else { |
| 274 |
// title may be on the next line |
| 275 |
if (isset($lines[$current + 1]) && preg_match('/^\s+[\(\'"](.+?)[\)\'"]\s*$/', $lines[$current + 1], $matches)) { |
| 276 |
$this->references[$label]['title'] = $matches[1]; |
| 277 |
$current++; |
| 278 |
} |
| 279 |
} |
| 280 |
$current++; |
| 281 |
} |
| 282 |
return [false, --$current]; |
| 283 |
} |
| 284 |
|
| 285 |
abstract protected function parseInline($text); |
| 286 |
abstract protected function renderAbsy($blocks); |
| 287 |
} |
| 288 |
|
