how-to strip empty HTML tags like <b> </b>

[russmann]

how-to strip empty HTML tags like <b> </b>

Originally posted as a Categorized Question.

[tachyon]

A more complete option is not Perl. Use HTML Tidy which is by w3c's Dave Raggett. I just tested it and it will strip empty tags a well as a doing many other useful things. Get your free copy here http://www.w3.org/People/Raggett/tidy/ or get a Win32 GUI version here http://perso.wanadoo.fr/ablavier/TidyGUI/

[tachyon]

The normal answer would be use HTML::TokeParser if you want a reliable solution to parse HTML. For this *particular* task a well constructed regex should suffice. This will strip all the <b> </b> tags that are empty. I believe it covers all bases.

$text = join '', <DATA>;
print $text;
$text =~ s#<\s*b\s*>(?:[\s\n]|&nbsp;)*<\s*/\s*b\s*>##ig;
print $text;

__DATA__
<p>test
<p>test<b></b>
<p>test<b
>&nbsp;  
</b>
<p>test<b>  </b>
<p>test<b 
></b >
<p>test<B></B>
<p>test<B>

</B>
<p>test<B ></B>
<p>test<B
>  </B >
<p>test<B>  &nbsp;  &nbsp; </B>
<p><b>I am not empty!</b>

[Hofmator]

I'd not replace the non-breaking spaces, maybe you could remove the bold tags, but removing the nbsp can mess up the layout of a table.

Just my euro 0.02

-- Hofmator

[Hofmator]

I'd also go with tachyon's suggestion of HTML Tidy, but if you are trying to do this quick and dirty somewhere in the middle of a script, I'd use this regex $text =~ s#<\s*([^>]*)\s*>[\s\n]*<\s*/\s*\1\s*>##ig; It should remove any empty tags which don't contain any attributes (not just bold tags), so it works on

__DATA__
<i> </ I>
< B ></b>
< em>  < / eM  >

[Anonymous Monk]

You might find this interesting: We made a javascript function that trims leading/trailing spaces from field values, but found it did not include the nonbreaking space character (nbsp;) when it appeared.
Before fixing it, our function looked like this:
-------------------------------------
String.prototype.Trim = function() {

return this.replace(/(^\s*)|(\s*$)/g, "");

}
-------------------------------------
After discovering that values that included nbsp; were not getting trimmed, I found the following to be true:

nbsp; = chr(160) = xA0

So have now modified the Trim function to read:

-------------------------------------
String.prototype.Trim = function() {
return this.replace(/(^{\s\xA0#}*)|({\s\xA0#}*$)/g, "");
}
-------------------------------------

(Please note that I used CURLY BRACKETS in the example above, but if you use this, use SQUARE BRACKETS -- they just wouldn't display on this page when I typed them in...)

And it works!! FYI, to clean field values, our javascript code calls it this way:
document.forms[0].myField.value.Trim();
Hope this helps someone,
Susan Henesy

Originally posted as a Categorized Answer.