1 <?php if ( ! defined('BASEPATH')) exit('No direct script access allowed');
5 * An open source application development framework for PHP 5.1.6 or newer
8 * @author ExpressionEngine Dev Team
9 * @copyright Copyright (c) 2008 - 2011, EllisLab, Inc.
10 * @license http://codeigniter.com/user_guide/license.html
11 * @link http://codeigniter.com
16 // ------------------------------------------------------------------------
24 * @author ExpressionEngine Dev Team
25 * @link http://codeigniter.com/user_guide/helpers/
29 // Block level elements that should not be wrapped inside <p> tags
30 var $block_elements = 'address|blockquote|div|dl|fieldset|form|h\d|hr|noscript|object|ol|p|pre|script|table|ul';
32 // Elements that should not have <p> and <br /> tags within them.
33 var $skip_elements = 'p|pre|ol|ul|dl|object|table|h\d';
35 // Tags we want the parser to completely ignore when splitting the string.
36 var $inline_elements = 'a|abbr|acronym|b|bdo|big|br|button|cite|code|del|dfn|em|i|img|ins|input|label|map|kbd|q|samp|select|small|span|strong|sub|sup|textarea|tt|var';
38 // array of block level elements that require inner content to be within another block level element
39 var $inner_block_required = array('blockquote');
41 // the last block element parsed
42 var $last_block_element = '';
44 // whether or not to protect quotes within { curly braces }
45 var $protect_braced_quotes = FALSE;
50 * This function converts text, making it typographically correct:
51 * - Converts double spaces into paragraphs.
52 * - Converts single line breaks into <br /> tags
53 * - Converts single and double quotes into correctly facing curly quote entities.
54 * - Converts three dots into ellipsis.
55 * - Converts double dashes into em-dashes.
56 * - Converts two spaces into entities
60 * @param bool whether to reduce more then two consecutive newlines to two
63 function auto_typography($str, $reduce_linebreaks = FALSE)
70 // Standardize Newlines to make matching easier
71 if (strpos($str, "\r") !== FALSE)
73 $str = str_replace(array("\r\n", "\r"), "\n", $str);
76 // Reduce line breaks. If there are more than two consecutive linebreaks
77 // we'll compress them down to a maximum of two since there's no benefit to more.
78 if ($reduce_linebreaks === TRUE)
80 $str = preg_replace("/\n\n+/", "\n\n", $str);
83 // HTML comment tags don't conform to patterns of normal tags, so pull them out separately, only if needed
84 $html_comments = array();
85 if (strpos($str, '<!--') !== FALSE)
87 if (preg_match_all("#(<!\-\-.*?\-\->)#s", $str, $matches))
89 for ($i = 0, $total = count($matches[0]); $i < $total; $i++)
91 $html_comments[] = $matches[0][$i];
92 $str = str_replace($matches[0][$i], '{@HC'.$i.'}', $str);
97 // match and yank <pre> tags if they exist. It's cheaper to do this separately since most content will
98 // not contain <pre> tags, and it keeps the PCRE patterns below simpler and faster
99 if (strpos($str, '<pre') !== FALSE)
101 $str = preg_replace_callback("#<pre.*?>.*?</pre>#si", array($this, '_protect_characters'), $str);
104 // Convert quotes within tags to temporary markers.
105 $str = preg_replace_callback("#<.+?>#si", array($this, '_protect_characters'), $str);
107 // Do the same with braces if necessary
108 if ($this->protect_braced_quotes === TRUE)
110 $str = preg_replace_callback("#\{.+?\}#si", array($this, '_protect_characters'), $str);
113 // Convert "ignore" tags to temporary marker. The parser splits out the string at every tag
114 // it encounters. Certain inline tags, like image tags, links, span tags, etc. will be
115 // adversely affected if they are split out so we'll convert the opening bracket < temporarily to: {@TAG}
116 $str = preg_replace("#<(/*)(".$this->inline_elements.")([ >])#i", "{@TAG}\\1\\2\\3", $str);
118 // Split the string at every tag. This expression creates an array with this prototype:
122 // [0] = <opening tag>
124 // [2] = <closing tag>
127 $chunks = preg_split('/(<(?:[^<>]+(?:"[^"]*"|\'[^\']*\')?)+>)/', $str, -1, PREG_SPLIT_DELIM_CAPTURE|PREG_SPLIT_NO_EMPTY);
129 // Build our finalized string. We cycle through the array, skipping tags, and processing the contained text
134 $total_chunks = count($chunks);
136 foreach ($chunks as $chunk)
140 // Are we dealing with a tag? If so, we'll skip the processing for this cycle.
141 // Well also set the "process" flag which allows us to skip <pre> tags and a few other things.
142 if (preg_match("#<(/*)(".$this->block_elements.").*?>#", $chunk, $match))
144 if (preg_match("#".$this->skip_elements."#", $match[2]))
146 $process = ($match[1] == '/') ? TRUE : FALSE;
151 $this->last_block_element = $match[2];
158 if ($process == FALSE)
164 // Force a newline to make sure end tags get processed by _format_newlines()
165 if ($current_chunk == $total_chunks)
170 // Convert Newlines into <p> and <br /> tags
171 $str .= $this->_format_newlines($chunk);
174 // No opening block level tag? Add it if needed.
175 if ( ! preg_match("/^\s*<(?:".$this->block_elements.")/i", $str))
177 $str = preg_replace("/^(.*?)<(".$this->block_elements.")/i", '<p>$1</p><$2', $str);
180 // Convert quotes, elipsis, em-dashes, non-breaking spaces, and ampersands
181 $str = $this->format_characters($str);
183 // restore HTML comments
184 for ($i = 0, $total = count($html_comments); $i < $total; $i++)
186 // remove surrounding paragraph tags, but only if there's an opening paragraph tag
187 // otherwise HTML comments at the ends of paragraphs will have the closing tag removed
188 // if '<p>{@HC1}' then replace <p>{@HC1}</p> with the comment, else replace only {@HC1} with the comment
189 $str = preg_replace('#(?(?=<p>\{@HC'.$i.'\})<p>\{@HC'.$i.'\}(\s*</p>)|\{@HC'.$i.'\})#s', $html_comments[$i], $str);
195 // If the user submitted their own paragraph tags within the text
196 // we will retain them instead of using our tags.
197 '/(<p[^>*?]>)<p>/' => '$1', // <?php BBEdit syntax coloring bug fix
199 // Reduce multiple instances of opening/closing paragraph tags to a single one
200 '#(</p>)+#' => '</p>',
201 '/(<p>\W*<p>)+/' => '<p>',
203 // Clean up stray paragraph tags that appear before block level elements
204 '#<p></p><('.$this->block_elements.')#' => '<$1',
206 // Clean up stray non-breaking spaces preceeding block elements
207 '#( \s*)+<('.$this->block_elements.')#' => ' <$2',
209 // Replace the temporary markers we added earlier
216 // An unintended consequence of the _format_newlines function is that
217 // some of the newlines get truncated, resulting in <p> tags
218 // starting immediately after <block> tags on the same line.
219 // This forces a newline after such occurrences, which looks much nicer.
220 "/><p>\n/" => ">\n<p>",
222 // Similarly, there might be cases where a closing </block> will follow
223 // a closing </p> tag, so we'll correct it by adding a newline in between
224 "#</p></#" => "</p>\n</"
227 // Do we need to reduce empty lines?
228 if ($reduce_linebreaks === TRUE)
230 $table['#<p>\n*</p>#'] = '';
234 // If we have empty paragraph tags we add a non-breaking space
235 // otherwise most browsers won't treat them as true paragraphs
236 $table['#<p></p>#'] = '<p> </p>';
239 return preg_replace(array_keys($table), $table, $str);
243 // --------------------------------------------------------------------
248 * This function mainly converts double and single quotes
249 * to curly entities, but it also converts em-dashes,
250 * double spaces, and ampersands
256 function format_characters($str)
260 if ( ! isset($table))
263 // nested smart quotes, opening and closing
264 // note that rules for grammar (English) allow only for two levels deep
265 // and that single quotes are _supposed_ to always be on the outside
266 // but we'll accommodate both
267 // Note that in all cases, whitespace is the primary determining factor
268 // on which direction to curl, with non-word characters like punctuation
269 // being a secondary factor only after whitespace is addressed.
270 '/\'"(\s|$)/' => '’”$1',
271 '/(^|\s|<p>)\'"/' => '$1‘“',
272 '/\'"(\W)/' => '’”$1',
273 '/(\W)\'"/' => '$1‘“',
274 '/"\'(\s|$)/' => '”’$1',
275 '/(^|\s|<p>)"\'/' => '$1“‘',
276 '/"\'(\W)/' => '”’$1',
277 '/(\W)"\'/' => '$1“‘',
279 // single quote smart quotes
280 '/\'(\s|$)/' => '’$1',
281 '/(^|\s|<p>)\'/' => '$1‘',
282 '/\'(\W)/' => '’$1',
283 '/(\W)\'/' => '$1‘',
285 // double quote smart quotes
286 '/"(\s|$)/' => '”$1',
287 '/(^|\s|<p>)"/' => '$1“',
288 '/"(\W)/' => '”$1',
289 '/(\W)"/' => '$1“',
292 "/(\w)'(\w)/" => '$1’$2',
294 // Em dash and ellipses dots
295 '/\s?\-\-\s?/' => '—',
296 '/(\w)\.{3}/' => '$1…',
298 // double space after sentences
299 '/(\W) /' => '$1 ',
301 // ampersands, if not a character entity
302 '/&(?!#?[a-zA-Z0-9]{2,};)/' => '&'
306 return preg_replace(array_keys($table), $table, $str);
309 // --------------------------------------------------------------------
314 * Converts newline characters into either <p> tags or <br />
320 function _format_newlines($str)
327 if (strpos($str, "\n") === FALSE && ! in_array($this->last_block_element, $this->inner_block_required))
332 // Convert two consecutive newlines to paragraphs
333 $str = str_replace("\n\n", "</p>\n\n<p>", $str);
335 // Convert single spaces to <br /> tags
336 $str = preg_replace("/([^\n])(\n)([^\n])/", "\\1<br />\\2\\3", $str);
338 // Wrap the whole enchilada in enclosing paragraphs
341 // We trim off the right-side new line so that the closing </p> tag
342 // will be positioned immediately following the string, matching
343 // the behavior of the opening <p> tag
344 $str = '<p>'.rtrim($str).'</p>';
347 // Remove empty paragraphs if they are on the first line, as this
348 // is a potential unintended consequence of the previous code
349 $str = preg_replace("/<p><\/p>(.*)/", "\\1", $str, 1);
354 // ------------------------------------------------------------------------
359 * Protects special characters from being formatted later
360 * We don't want quotes converted within tags so we'll temporarily convert them to {@DQ} and {@SQ}
361 * and we don't want double dashes converted to emdash entities, so they are marked with {@DD}
362 * likewise double spaces are converted to {@NBS} to prevent entity conversion
368 function _protect_characters($match)
370 return str_replace(array("'",'"','--',' '), array('{@SQ}', '{@DQ}', '{@DD}', '{@NBS}'), $match[0]);
373 // --------------------------------------------------------------------
376 * Convert newlines to HTML line breaks except within PRE tags
382 function nl2br_except_pre($str)
384 $ex = explode("pre>",$str);
388 for ($i = 0; $i < $ct; $i++)
392 $newstr .= nl2br($ex[$i]);
407 // END Typography Class
409 /* End of file Typography.php */
410 /* Location: ./system/libraries/Typography.php */