Rick Ellis | 4c938ae | 2008-09-10 22:58:38 +0000 | [diff] [blame] | 1 | <?php if ( ! defined('BASEPATH')) exit('No direct script access allowed'); |
| 2 | /** |
| 3 | * CodeIgniter |
| 4 | * |
| 5 | * An open source application development framework for PHP 4.3.2 or newer |
| 6 | * |
| 7 | * @package CodeIgniter |
| 8 | * @author ExpressionEngine Dev Team |
| 9 | * @copyright Copyright (c) 2006, EllisLab, Inc. |
| 10 | * @license http://codeigniter.com/user_guide/license.html |
| 11 | * @link http://codeigniter.com |
| 12 | * @since Version 1.0 |
| 13 | * @filesource |
| 14 | */ |
| 15 | |
| 16 | // ------------------------------------------------------------------------ |
| 17 | |
| 18 | /** |
| 19 | * Typography Class |
| 20 | * |
| 21 | * |
| 22 | * @access private |
| 23 | * @category Helpers |
| 24 | * @author ExpressionEngine Dev Team |
| 25 | * @link http://codeigniter.com/user_guide/helpers/ |
| 26 | */ |
| 27 | class CI_Typography { |
| 28 | |
| 29 | // Block level elements that should not be wrapped inside <p> tags |
| 30 | var $block_elements = 'div|blockquote|pre|code|h\d|script|ol|ul'; |
| 31 | |
| 32 | // Elements that should not have <p> and <br /> tags within them. |
| 33 | var $skip_elements = 'pre|ol|ul'; |
| 34 | |
| 35 | // Tags we want the parser to completely ignore when splitting the string. |
| 36 | var $ignore_elements = 'a|b|i|em|strong|span|img|li'; |
| 37 | |
Rick Ellis | 18bd8b5 | 2008-09-10 23:40:35 +0000 | [diff] [blame] | 38 | // Whether to allow Javascript event handlers to be sumitted inside tags |
| 39 | var $allow_js_event_handlers = FALSE; |
Rick Ellis | 4c938ae | 2008-09-10 22:58:38 +0000 | [diff] [blame] | 40 | |
| 41 | /** |
| 42 | * Main Processing Function |
| 43 | * |
| 44 | */ |
| 45 | function convert($str) |
| 46 | { |
| 47 | if ($str == '') |
| 48 | { |
| 49 | return ''; |
| 50 | } |
| 51 | |
| 52 | $str = ' '.$str.' '; |
| 53 | |
| 54 | // Standardize Newlines to make matching easier |
| 55 | if (strpos($str, "\r") !== FALSE) |
| 56 | { |
| 57 | $str = str_replace(array("\r\n", "\r"), "\n", $str); |
| 58 | } |
| 59 | |
| 60 | /* |
| 61 | * Reduce line breaks |
| 62 | * |
| 63 | * If there are more than two consecutive line |
| 64 | * breaks we'll compress them down to a maximum |
| 65 | * of two since there's no benefit to more. |
| 66 | * |
| 67 | */ |
| 68 | $str = preg_replace("/\n\n+/", "\n\n", $str); |
| 69 | |
| 70 | /* |
Rick Ellis | 18bd8b5 | 2008-09-10 23:40:35 +0000 | [diff] [blame] | 71 | * Do we allow JavaScript event handlers? |
| 72 | * |
| 73 | * If not, we strip them from within all tags |
| 74 | */ |
| 75 | if ($this->allow_js_event_handlers == FALSE) |
| 76 | { |
Rick Ellis | e320bc8 | 2008-09-11 18:11:31 +0000 | [diff] [blame] | 77 | $str = preg_replace("#<([^><]+?)([^a-z_\-]on\w*|xmlns)(\s*=\s*[^><]*)([><]*)#i", "<\\1\\4", $str); |
Rick Ellis | 18bd8b5 | 2008-09-10 23:40:35 +0000 | [diff] [blame] | 78 | } |
| 79 | |
| 80 | /* |
Rick Ellis | 4c938ae | 2008-09-10 22:58:38 +0000 | [diff] [blame] | 81 | * Convert quotes within tags to temporary marker |
| 82 | * |
| 83 | * We don't want quotes converted within |
| 84 | * tags so we'll temporarily convert them to |
| 85 | * {@DQ} and {@SQ} |
| 86 | * |
| 87 | */ |
| 88 | if (preg_match_all("#\<.+?>#si", $str, $matches)) |
| 89 | { |
| 90 | for ($i = 0; $i < count($matches['0']); $i++) |
| 91 | { |
| 92 | $str = str_replace($matches['0'][$i], |
| 93 | str_replace(array("'",'"'), array('{@SQ}', '{@DQ}'), $matches['0'][$i]), |
| 94 | $str); |
| 95 | } |
| 96 | } |
Rick Ellis | 4c938ae | 2008-09-10 22:58:38 +0000 | [diff] [blame] | 97 | |
Rick Ellis | 18bd8b5 | 2008-09-10 23:40:35 +0000 | [diff] [blame] | 98 | /* |
Rick Ellis | 4c938ae | 2008-09-10 22:58:38 +0000 | [diff] [blame] | 99 | * Add closing/opening paragraph tags before/after "block" elements |
| 100 | * |
| 101 | * Since block elements (like <blockquotes>, <pre>, etc.) do not get |
| 102 | * wrapped in paragraph tags we will add a closing </p> tag just before |
| 103 | * each block element starts and an opening <p> tag right after the block element |
| 104 | * ends. Later on we'll do some further clean up. |
| 105 | * |
| 106 | */ |
| 107 | $str = preg_replace("#(<)(".$this->block_elements.")(.*?>)#", "</p>\\1\\2\\3", $str); |
| 108 | $str = preg_replace("#(</)(".$this->block_elements.")(.*?>)#", "\\1\\2\\3<p>", $str); |
| 109 | |
| 110 | /* |
| 111 | * Convert "ignore" tags to temporary marker |
| 112 | * |
| 113 | * The parser splits out the string at every tag |
| 114 | * it encounters. Certain inline tags, like image |
| 115 | * tags, links, span tags, etc. will be adversely |
| 116 | * affected if they are split out so we'll convert |
| 117 | * the opening < temporarily to: {@TAG} |
| 118 | * |
| 119 | */ |
| 120 | $str = preg_replace("#<(/*)(".$this->ignore_elements.")#i", "{@TAG}\\1\\2", $str); |
| 121 | |
| 122 | /* |
| 123 | * Split the string at every tag |
| 124 | * |
| 125 | * This creates an array with this prototype: |
| 126 | * |
| 127 | * [array] |
| 128 | * { |
| 129 | * [0] = <opening tag> |
| 130 | * [1] = Content contained between the tags |
| 131 | * [2] = <closing tag> |
| 132 | * Etc... |
| 133 | * } |
| 134 | * |
| 135 | */ |
| 136 | $chunks = preg_split('/(<(?:[^<>]+(?:"[^"]*"|\'[^\']*\')?)+>)/', $str, -1, PREG_SPLIT_DELIM_CAPTURE|PREG_SPLIT_NO_EMPTY); |
| 137 | |
| 138 | /* |
| 139 | * Build our finalized string |
| 140 | * |
| 141 | * We'll cycle through the array, skipping tags, |
| 142 | * and processing the contained text |
| 143 | * |
| 144 | */ |
| 145 | $str = ''; |
| 146 | $process = TRUE; |
| 147 | foreach ($chunks as $chunk) |
| 148 | { |
| 149 | /* |
| 150 | * Are we dealing with a tag? |
| 151 | * |
| 152 | * If so, we'll skip the processing for this cycle. |
| 153 | * Well also set the "process" flag which allows us |
| 154 | * to skip <pre> tags and a few other things. |
| 155 | * |
| 156 | */ |
| 157 | if (preg_match("#<(/*)(".$this->block_elements.").*?\>#", $chunk, $match)) |
| 158 | { |
| 159 | if (preg_match("#".$this->skip_elements."#", $match['2'])) |
| 160 | { |
| 161 | $process = ($match['1'] == '/') ? TRUE : FALSE; |
| 162 | } |
| 163 | |
| 164 | $str .= $chunk; |
| 165 | continue; |
| 166 | } |
| 167 | |
| 168 | if ($process == FALSE) |
| 169 | { |
| 170 | $str .= $chunk; |
| 171 | continue; |
| 172 | } |
| 173 | |
| 174 | // Convert Newlines into <p> and <br /> tags |
| 175 | $str .= $this->format_newlines($chunk); |
| 176 | } |
| 177 | |
| 178 | // FINAL CLEAN UP |
| 179 | // IMPORTANT: DO NOT ALTER THE ORDER OF THE ITEMS BELOW! |
| 180 | |
| 181 | /* |
| 182 | * Clean up paragraph tags before/after "block" elements |
| 183 | * |
| 184 | * Earlier we added <p></p> tags before/after block level elements. |
| 185 | * Then, we added paragraph tags around double line breaks. This |
| 186 | * potentially created incorrectly formatted paragraphs so we'll |
| 187 | * clean it up here. |
| 188 | * |
| 189 | */ |
| 190 | $str = preg_replace("#<p>({@TAG}.*?)(".$this->block_elements.")(.*?>)#", "\\1\\2\\3", $str); |
| 191 | $str = preg_replace("#({@TAG}/.*?)(".$this->block_elements.")(.*?>)</p>#", "\\1\\2\\3", $str); |
| 192 | |
| 193 | // Convert Quotes and other characters |
| 194 | $str = $this->format_characters($str); |
| 195 | |
| 196 | // Fix an artifact that happens during the paragraph replacement |
| 197 | $str = preg_replace('#(<p>\n*</p>)#', '', $str); |
| 198 | |
| 199 | // If the user submitted their own paragraph tags with class data |
| 200 | // in them we will retain them instead of using our tags. |
Rick Ellis | 18bd8b5 | 2008-09-10 23:40:35 +0000 | [diff] [blame] | 201 | $str = preg_replace('#(<p.*?>)<p>#', "\\1", $str); // <?php BBEdit syntax coloring fix |
Rick Ellis | 4c938ae | 2008-09-10 22:58:38 +0000 | [diff] [blame] | 202 | |
| 203 | // Final clean up |
| 204 | $str = str_replace( |
| 205 | array( |
| 206 | '</p></p>', |
| 207 | '</p><p>', |
| 208 | '<p> ', |
| 209 | ' </p>', |
| 210 | '{@TAG}', |
| 211 | '{@DQ}', |
| 212 | '{@SQ}', |
| 213 | '<p></p>' |
| 214 | ), |
| 215 | array( |
| 216 | '</p>', |
| 217 | '<p>', |
| 218 | '<p>', |
| 219 | '</p>', |
| 220 | '<', |
| 221 | '"', |
| 222 | "'", |
| 223 | '' |
| 224 | ), |
| 225 | $str |
| 226 | ); |
| 227 | |
| 228 | return $str; |
| 229 | } |
| 230 | |
| 231 | // -------------------------------------------------------------------- |
| 232 | |
| 233 | /** |
| 234 | * Format Characters |
| 235 | * |
| 236 | * This function mainly converts double and single quotes |
Derek Jones | ab504b8 | 2008-09-11 17:04:30 +0000 | [diff] [blame] | 237 | * to curly entities, but it also converts em-dashes, |
| 238 | * double spaces, and ampersands |
Rick Ellis | 4c938ae | 2008-09-10 22:58:38 +0000 | [diff] [blame] | 239 | */ |
| 240 | function format_characters($str) |
Derek Jones | ab504b8 | 2008-09-11 17:04:30 +0000 | [diff] [blame] | 241 | { |
| 242 | static $table; |
| 243 | |
| 244 | if ( ! isset($table)) |
| 245 | { |
| 246 | $table = array( |
| 247 | // nested smart quotes, opening and closing |
| 248 | // note that rules for grammar (English) allow only for two levels deep |
| 249 | // and that single quotes are _supposed_ to always be on the outside |
| 250 | // but we'll accommodate both |
| 251 | '/(^|\W|\s)\'"/' => '$1‘“', |
| 252 | '/\'"(\s|\W|$)/' => '’”$1', |
| 253 | '/(^|\W|\s)"\'/' => '$1“‘', |
| 254 | '/"\'(\s|\W|$)/' => '”’$1', |
Rick Ellis | 4c938ae | 2008-09-10 22:58:38 +0000 | [diff] [blame] | 255 | |
Derek Jones | ab504b8 | 2008-09-11 17:04:30 +0000 | [diff] [blame] | 256 | // single quote smart quotes |
| 257 | '/\'(\s|\W|$)/' => '’$1', |
| 258 | '/(^|\W|\s)\'/' => '$1‘', |
Rick Ellis | 4c938ae | 2008-09-10 22:58:38 +0000 | [diff] [blame] | 259 | |
Derek Jones | ab504b8 | 2008-09-11 17:04:30 +0000 | [diff] [blame] | 260 | // double quote smart quotes |
| 261 | '/"(\s|\W|$)/' => '”$1', |
| 262 | '/(^|\W|\s)"/' => '$1“', |
Rick Ellis | 4c938ae | 2008-09-10 22:58:38 +0000 | [diff] [blame] | 263 | |
Derek Jones | ab504b8 | 2008-09-11 17:04:30 +0000 | [diff] [blame] | 264 | // apostrophes |
| 265 | "/(\w)'(\w)/" => '$1’$2', |
| 266 | |
| 267 | // Em dash and ellipses dots |
| 268 | '/\s?\-\-\s?/' => '—', |
| 269 | '/\w\.{3}/' => '…', |
| 270 | |
| 271 | // double space after sentences |
Derek Jones | 50500a2 | 2008-09-11 18:21:18 +0000 | [diff] [blame^] | 272 | '/(\W) /' => '$1 ', |
Derek Jones | ab504b8 | 2008-09-11 17:04:30 +0000 | [diff] [blame] | 273 | |
| 274 | // ampersands, if not a character entity |
| 275 | '/&(?!#?[a-zA-Z0-9]{2,};)/' => '&' |
| 276 | ); |
| 277 | } |
| 278 | |
| 279 | return preg_replace(array_keys($table), $table, $str); |
Rick Ellis | 4c938ae | 2008-09-10 22:58:38 +0000 | [diff] [blame] | 280 | } |
| 281 | |
| 282 | // -------------------------------------------------------------------- |
| 283 | |
| 284 | /** |
| 285 | * Format Newlines |
| 286 | * |
| 287 | * Converts newline characters into either <p> tags or <br /> |
| 288 | * |
| 289 | */ |
| 290 | function format_newlines($str) |
| 291 | { |
| 292 | if ($str == '') |
| 293 | { |
| 294 | return $str; |
| 295 | } |
| 296 | |
| 297 | if (strpos($str, "\n") === FALSE) |
| 298 | { |
| 299 | return '<p>'.$str.'</p>'; |
| 300 | } |
| 301 | |
| 302 | $str = str_replace("\n\n", "</p>\n\n<p>", $str); |
| 303 | $str = preg_replace("/([^\n])(\n)([^\n])/", "\\1<br />\\2\\3", $str); |
| 304 | |
| 305 | return '<p>'.$str.'</p>'; |
Rick Ellis | 18bd8b5 | 2008-09-10 23:40:35 +0000 | [diff] [blame] | 306 | } |
| 307 | |
| 308 | // -------------------------------------------------------------------- |
| 309 | |
| 310 | /** |
| 311 | * Allow JavaScript Event Handlers? |
| 312 | * |
| 313 | * For security reasons, by default we disallow JS event handlers |
| 314 | * |
| 315 | */ |
| 316 | function allow_js_event_handlers($val = FALSE) |
| 317 | { |
| 318 | $this->allow_js_event_handlers = ($val === FALSE) ? FALSE : TRUE; |
| 319 | } |
| 320 | |
| 321 | |
Rick Ellis | 4c938ae | 2008-09-10 22:58:38 +0000 | [diff] [blame] | 322 | } |
| 323 | // END Typography Class |
| 324 | |
| 325 | /* End of file Typography.php */ |
| 326 | /* Location: ./system/libraries/Typography.php */ |