Converting MS "Smart Quotes" how the heck?

[StuManUSA]

Some Microsoft "Smart Quotes" got put into my database. I'm trying to filter them out as they come out of the database and the text is filtered and displayed on my webpage.

I am using a series of filters and have been unsuccessful thus far with the following:

             <<Note>>: assume that any occurances of &amp; are actually just a &.  

The forum is messing with my attempt to show you examples of my exact code, darnit.

          // not even this searching works
	if (strstr($text, "“")) echo "found one";
	if (strstr($text, "&amp;#8220;")) echo "found one";
	if (strstr($text, "&amp;#8221;")) echo "found one";
	// my failed attempt at finding and replacing
	$text = str_replace("&amp;#8221;", "&quot;", $text); //smart quote left
	$text = str_replace("&amp;#8221;", "&quot;", $text); //smart quote right
	$text = str_replace("&amp;#8216;", "'", $text); //smart apostrophe 1
	$text = str_replace("&amp;#8217;", "'", $text); //smart apostrophe 2
	$text = str_replace("&amp;#8230;", "...", $text); // smart ...
	$text = str_replace("&amp;#8212;", "–", $text); // smart dash
	// more standard approaches don't work either
	$text = strip_tags($text);		
	$text = stripslashes($text);
	$text = htmlspecialchars($text);
	return $text;

You may be curious what the data actually looks like in the MySQL database so here is a select....
+----------------------------------------------------------------------------+
| primary_name |
+----------------------------------------------------------------------------+
| Michael Stipe?s "In the Sun? Brings Hurricane Recovery Challenges to Light |
+----------------------------------------------------------------------------+
1 row in set (0.02 sec)

mysql>

The "?" are evidently the SmartQuotes which MySQL doesn't know how to display.

However once these characters get to the browser they are displayed as: " which sucks because it's breaking my Javascript menu.

Arrgh.

Any suggestions for how to catch this crappy SmartQuote stuff and filter it out?
Thanks,
Stuart

[cbj4074]

I've had this issue come up at least once before, and this may do the job without getting too fancy.

I also noticed that you could use str_replace much more effectively, assuming you're using a version of PHP that is new enough to support arrays for parameters 1 and 2 (see the below example). Basically, the $search array has all of the messed-up characters you're searching for and the $replace array has all of the characters you'd like to use to replace those from the $search array - in the same order.

$search = array('“', '”', '“', '”', '’', '–', '’', '“', '”', '&'); 
$replace = array('"', '"', '"', '"', '\'', '—', '\'', '"', '"', '&');
$result = str_replace($search, $replace, $subject);

The only stupid Microsoft one I didn't include there that is pretty common was the elipsis (...), but I'm sure you can tackle that.

I also noticed that there is a similar discussion taking place here:

http://phpbuilder.com/board/showthread.php?t=10317719

That's somewhat of a more industrial-strength solution, but worth trying if what I suggested doesn't do what you need.

[StuManUSA]

Yours didn't quite work, but this seemed to. I don't know why it does, but I found it somewhere. I thought SmartQuotes were not ASCII characters so I don't see how the chr() function works but it somehow does. Weird.

   		$chrs = array (chr(150), chr(147), chr(148), chr(146));
		$repl = array ("-", "\"", "\"", "'");
		$text = str_replace($chrs, $repl, $text);

Thanks.

[cbj4074]

I suspect that mine didn't work because when I pasted my code into the browser and submitted the post, those characters were translated into something that doesn't look like a question mark. Make sense? 🙂 I should have mentioned that... sorry. I pasted those in from Word->Notepad to get them in the first place, by the way.

Note also that the complement to the chr() function, which is ord(), will return the ASCII value of the input string (the converse of what you're doing), so that could be immensely useful in helping you sort out these messy character situations in the future. You copy the character from Word, paste it into the ord() function, and voila - you have the ASCII integer value.

Anyway, I'm glad that with your modifications it works, and I'm also glad to know the ASCII char numbers now, thanks to your research.

Now, can you just mark the thread resolved?

Cheers!