Assignment

Oh I so want to suggest filing this as a bug report - with Microsoft!!! Their 'dirty' html causes issues when people want to copy and paste in all sorts of situations

Richard

problem with splitting just the word and word boundary (\w\b) is that it also split contraction words like "can't" and "don't"  which will result in more words than it should be. the count_words function is too simplicistics and possibly only works with western languages--i'm not sure how to improve the implementation for different languages. here's my hack that may work a little better.

/**
 * Count words in a string.
 *
 * Words are defined as things between whitespace.
 *
 * @param string $string The text to be searched for words.
 * @return int The count of words in the specified string
 */
function count_words($string) {
 $string = strip_tags(html_entity_decode($string, ENT_QUOTES, 'UTF-8'));
 $splitted = preg_split('/\s+/', $string, null, PREG_SPLIT_NO_EMPTY);
 $counted = count($splitted);
 foreach($splitted as $i=>$v) {
 $v = trim($v);
 $joined = preg_split('#[,;\.?!\/]#', $v, null, PREG_SPLIT_NO_EMPTY);
 if (count($joined) < 2)
 continue;
 if ( empty($joined[0]) || empty($joined[1]) )
 continue;
 $joined[1] = trim($joined[1]);
 error_log(quoted_printable_encode($joined[1]));
 if ( preg_match('/\w/', $joined[1]) ) {
 $counted++;
 }
 }//endforeach
 return $counted;
}

We too  have been struggling with the mismatch between MS Word and Moodle version of word count.  The discussion and suggested code changes here are very interesting, but we would much prefer if core Moodle code would do the implementation rather than adding another hack to our implementation.  So, getting back to the original question, I would say yes, absolutely, please do submit this as a bug.  Also, I understand and agree about text pasted from MS applications being a mess, but that's the world we live in.  I think we need to do the best we can with it, as getting MS to change their minds is unlikely.

-- mike