tonyair/silverstripe-framework
1
0
Fork 0
mirror of https://github.com/silverstripe/silverstripe-framework synced 2024-10-22 14:05:37 +02:00
Code Issues Projects Releases Wiki Activity

BUGFIX: Replace HTMLText#Summary with one that works.

Browse Source 
API CHANGE: Added two arguments to HTMLText#Summary. Minimal impact since previously any usage of this function threw an error. 
ENHANCEMENT: Add HTMLText#FirstSentence based on new HTMLText#Summary

git-svn-id: svn://svn.silverstripe.com/silverstripe/open/modules/sapphire/branches/2.3@78618 467b73ca-7a2a-4603-9d3b-597d59a354a9
This commit is contained in:
Hamish Friedlander 2009-06-08 05:38:54 +00:00 committed by Sam Minnee
parent b38315bdba
commit f74069f060
1 changed files with 71 additions and 50 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

121
core/model/fieldtypes/HTMLText.php
View File

@ -23,62 +23,83 @@ class HTMLText extends Text {
}
/**
* Create a summary of the content. This will either be the first paragraph, or the first $maxWords
* words, whichever is shorter
* Create a summary of the content. This will be some section of the first paragraph, limited by
* $maxWords. All internal tags are stripped out - the return value is a string
*
* This is sort of the HTML aware equivilent to Text#Summary, although the logic for summarising is not exactly the same
*
* @param int $maxWords Maximum number of words to return - may return less, but never more. Pass -1 for no limit
* @param int $flex Number of words to search through when looking for a nice cut point
* @param string $add What to add to the end of the summary if we cut at a less-than-ideal cut point
* @return string A nice(ish) summary with no html tags (but possibly still some html entities)
*
* @see sapphire/core/model/fieldtypes/Text#Summary($maxWords)
*/
public function Summary( $maxWords = 50 ) {
// split the string into tags and words
$parts = Convert::xml2array( $this->value );
// store any unmatched tags
$tagStack = array();
$pIndex = 0;
// find the first paragraph tag
for( $i = 0; $i < count( $parts ); $i++ )
if( strpos( $parts[$i], '<p' ) === 0 ) {
$pIndex = $i;
break;
}
$summary = '';
$words = 0;
// create the summary, keeping track of opening and closing tags
while( $words <= $maxWords && $pIndex < count( $parts ) ) {
if( $parts[$pIndex] == '</p>' ) {
$summary .= $parts[$pIndex];
break;
}
elseif( preg_match( '/<\/(\w+)>/', $parts[$pIndex], $endTag ) && $endTag[1] == substr( $tagStack[count($tagStack) - 1], 1, strlen( $endTag[1] ) ) ) {
array_pop( $tagStack );
$words++;
$summary .= $parts[$pIndex++];
} elseif( preg_match( '/^<\w+/', $parts[$pIndex] ) ) {
array_push( $tagStack, $parts[$pIndex] );
$words++;
$summary .= $parts[$pIndex++];
} else
$summary .= $parts[$pIndex++] . ' ';
}
// Tags that shouldn't be closed
$noClose = array("br", "img");
// make sure that the summary is well formed XHTML by closing tags
while( $openTag = array_pop( $tagStack ) ) {
preg_match( '/^<(\w+)\s+/', $openTag, $tagName );
if(sizeof($tagName) > 0) {
if(!in_array($tagName[1], $noClose)) {
$summary .= "</{$tagName[1]}>";
}
public function Summary($maxWords = 50, $flex = 15, $add = '...') {
$str = false;
/* First we need the text of the first paragraph, without tags. Try using SimpleXML first */
if (class_exists('SimpleXMLElement')) {
$doc = new DOMDocument();
$doc->strictErrorChecking = FALSE;
if ($doc->loadHTML('<meta content="text/html; charset=utf-8" http-equiv="Content-type"/>' . $this->value)) {
$xml = simplexml_import_dom($doc);
$res = $xml->xpath('//p');
if (!empty($res)) $str = strip_tags($res[0]->asXML());
}
}
return $summary;
if (!$str) {
/* If that failed, use a simple regex + a strip_tags. We look for the first paragraph with some words in it, not just the first paragraph.
* Not as good on broken HTML, and doesn't understand escaping or cdata blocks, but will probably work on even very malformed HTML */
if (preg_match('{<p[^>]*>(.*[A-Za-z]+.*)</p>}', $this->value, $matches)) {
$str = strip_tags($matches[1]);
}
/* If _that_ failed, just use the whole text with strip_tags */
else {
$str = strip_tags($this->value);
}
}
/* Now split into words. If we are under the maxWords limit, just return the whole string */
$words = preg_split('/\s+/', $str);
if ($maxWords == -1 || count($words) <= $maxWords) return $str;
/* Otherwise work backwards for a looking for a sentence ending (we try to avoid abbreviations, but aren't very good at it) */
for ($i = $maxWords; $i > $maxWords - $flex; $i--) {
if (preg_match('/\.$/', $words[$i]) && !preg_match('/(Dr|Mr|Mrs|Ms|Miss|Sr|Jr|No)\.$/i', $words[$i])) {
return implode(' ', array_slice($words, 0, $i+1));
}
}
/* If we didn't find a sentence ending quickly enough, just cut at the maxWords point and add '...' to the end */
return implode(' ', array_slice($words, 0, $maxWords)) . $add;
}
/**
* Returns the first sentence from the first paragraph. If it can't figure out what the first paragraph is (or there isn't one)
* it returns the same as Summary()
*
* This is the HTML aware equivilent to Text#FirstSentence
*
* @see sapphire/core/model/fieldtypes/Text#FirstSentence()
*/
function FirstSentence() {
/* Use summary's html processing logic to get the first paragraph */
$paragraph = $this->Summary(-1);
/* Then look for the first sentence ending. We could probably use a nice regex, but for now this will do */
$words = preg_split('/\s+/', $paragraph);
foreach ($words as $i => $word) {
if (preg_match('/\.$/', $word) && !preg_match('/(Dr|Mr|Mrs|Ms|Miss|Sr|Jr|No)\.$/i', $word)) {
return implode(' ', array_slice($words, 0, $i+1));
}
}
/* If we didn't find a sentence ending, use the summary. We re-call rather than using paragraph so that Summary will limit the result this time */
return $this->Summary();
}
public function scaffoldFormField($title = null, $params = null) {
return new HtmlEditorField($this->name, $title);
}