Question

I have a quick question about a preg_replace problem I have. I am just a newbie in RegEx.
What I would like to achieve is the following:

• I have a DIV tag <div data-info="sourcefile.ext" class="elm swf">sourcefile</div>
• I would like to extract the (data-info) value and the (class) value
• There might be optional tags but I don’t need te value of these attributes
• This replacement should work multiple times in one string

I have:

$input = '<div data-info="sourcefile.ext" class="elm swf">sourcefile</div>';
$input = preg_replace('/(<div\s(class="(.*?)")\s(data-info="(.*?)")\b[^>]*>)(.*?)<\/div>/i', "$1 class:$2 data-info:$3", $input);

I want to use the values as: <object src="(data-info)" type="(class)">

Is this possible? And can somebody show/explain me how this works?

Thank you very much.

Answer 1

Your regex fails because it lists the attributes in the wrong order. The match pattern does not accomodate for such things (which would be an actual reason to prefer using a DOM parser for such purposes.)

The \b escape is misplaced. And you can wrap the two attributes into (?: .. | .. )+ to allow for a little ambiguity:

 $input = preg_replace('/(?:<div
     (?: \s class="(.*?)"
       | \s data-info="(.*?)"  )+
     [^>]*>)
     (.*?)<\/div>/ix', "$3 class:$1 data-info:$2", $input);

The $1 $2 $3 numbering was off, and maybe you want to use named capture groups here anyway.