Best way to save user-generated texts

[LostKobrakai]

Hi,

I'm building a small and simple social area in my current project. By now I think I let users only input raw text and then use nl2br and maybe some nl2p function to add simple markup, which gets saved. If this texts gets editied the markup has to be converted back to \n's. 

Now I wonder how much work it woul be to store only the raw text and use a textformatter to add the markup. 

Also does somebody have a reliable nl2p function lying around? The ones I found didn't really consider the "\n", "\r\n" differences.  

[netcarver]

@LostKobrakai

Although it might not be the best solution, have you tried something like Textile before? There's a module available for it and you'll need to use the restricted version if the input is coming from untrusted sources like social users. It should handle the differences in \r, \n, \r\n etc but might be a little more complex than you really need.

Whatever you end up doing I suggest you always pass the finalised field values through htmlspecialchars() or htmlentities() before rendering them on your site.

[LostKobrakai]

Textile is quite interesting, but the inputs are just texts without further formatting, so I think that overkill.

[LostKobrakai]

Ok, should've looks earlier in the modules directory. There's already a textformatter for nl to br / p replacement. But I think I'll go with the solution to store the markup. 

[netcarver]

You could add something like the following (untested) function to your global functions or a common init file to do the job of dealing with different line endings and conversion to paragraphs...

function nl2p($str)
{
    // Change \r to \n...
    $out = str_replace("\r", "\n", $str);
    
    // Change runs of \n to just a single \n...
    $out = preg_replace("/\n{2,}/", "\n", $out);

    // htmlenitites...
    $out = htmlentities($out, ENT_QUOTES, 'UTF-8');

    // Change each \n into a paragraph...
    $out = '<p>' . str_replace("\n", "</p>\n<p>", $out) . "</p>\n";
 
    return $out;
}

This also takes care of the htmlentities() call for you.

Hope that helps.

Edited September 17, 2014 by netcarver
Fixed the htmlentities() call

[LostKobrakai]

Thanks. I ended up using another nl2p function, which does also retain linebreaks, but it's nealy the same.

function nl2p($text){
	$text = htmlentities(strip_tags($text), ENT_QUOTES, 'UTF-8');
	$text = '<p>' . $text . '</p>'; 
	$text = str_replace("\r\n\r\n", "</p><p>", $text); 
	$text = str_replace("\n\n", "</p><p>", $text); 
	$text = str_replace("\r\n", "<br />", $text); 
	$text = str_replace("\n", "<br />", $text);      
	return $text;
}

[netcarver]

@LostKobrakai

Thanks for posting your solution. Looks a little more flexible (allowing breaks and paragraphs) than the one I made up.