[SOLVED] Convert to Italics

[raku]

Hi,

 

I'm using the pattern '/_(.+)_/Ui' to match words/phrases surrounded with underscores and convert them to html italics. But, I would like to avoid links that may have double underscores in them. Is there a way to make the pattern only match if there is whitespace or nothing on either side of the underscores?

 

Thanks.

 

[aCa]

To only match the values with whitespace on each side you could use this pattern: /\s_(.+)_\s/iU

Would that solve your problem?

 

I have a beta out of my regex tool that can help you test the expression to make sure it is exactly what you wan't, http://regex.larsolavtorvik.com/.

[raku]

Thanks. The only problem there is that it doesn't match near the end of sentences. For example:

 

"This is _cool_." won't italicize "cool"

 

So essentially, I would also like it to match if there is nothing on the right of the word/phrase or if there is punctuation.

[effigy]

Try /\b_(.+)_\b/iU.

[raku]

Thanks, it works for that example. But, for the url http://www.google.com/_testing/blah_.html it converts the underscores into spaces and makes testing/blah italic.

[Orio]

Can you show the whole code... effigy posted a pattern, not a preg_replace statement.

 

Orio.

[raku]

Sure.

 

<?php
$subs = array(
  '/\b_(.+)_\b/iU' => ' <em>$1</em> '
);

$text = "http://www.google.com/_testing/blah_.html";

$formatted_text = preg_replace(array_keys($subs), array_values($subs), $text);
?>

 

Thanks.

[effigy]

I would like to avoid links that may have double underscores in them.

 

I assume by "double underscores" you mean any sets (pairs) within a URL? So, essentially, you want to avoid links altogether, correct?

[Orio]

How do you want the output (exactly) to look like?

 

Orio.

 

EDIT- lol nvm. I'll leave it to the master

[raku]

Effigy: Yeah that's exactly it.

[raku]

Orio: For that string, I don't want anything to happen to it.

[effigy]

You'll need something similar to this, which was used to ignore HTML:

 

<pre>
<?php

$data = <<<DATA
_italic_http://www.phpfreaks.com
http://www.google.com/_do_not_italicize
_abc__123_xyz
DATA;

### Isolate the URLs from the rest of the data.
$pieces = preg_split('%(http://\S+)%', $data, -1, PREG_SPLIT_NO_EMPTY | PREG_SPLIT_DELIM_CAPTURE);

### Analyze each piece.
foreach ($pieces as &$piece) {
	### Ignore the URLs.
	if (substr($piece, 0, 7) == 'http://') {
		continue;
	}
	$piece = preg_replace('/_([^_]+)_/', '<em>$1</em>', $piece);
}

### Reassemble.
echo $data = join('', $pieces);

?>
</pre>

[raku]

Awesome, thanks so much!

[raku]

Quick question: what does the "&" sign do in "&$piece"?

 

Thanks.

[effigy]

It passes by reference, allowing $piece to be modified by the loop.