prematch for replace html tags

Answer 1

Using regular expressions to extract data from HTML sources is frowned at at stackoverflow. Please consider using a html parser for this task (e.g. SimpleHTMLDom).

If you want to do this once, and quick and very dirty, maybe you can get away with something like

"<span class=bld>([^<]*)</span>"

This assumes that all and only all the currency values you are interested in are contained in span tags, with class bld and no other attributes.

Answer 2

Never use regex, always use a parser:

$htmlfragment = "<table align='center' id='tbl_currency'> <tr> <td><span class=bld>631.0075 USD</span></td></tr></table>";

$domdoc = new DomDocument();
$domdoc->loadHTML($htmlfragment);

$xpath = new DOMXPath($domdoc);
$result = $xpath->query("//table[@id='tbl_currency']//span[@class='bld']");

if ($result->length > 0) {
  $currency_span = $result->item(0);
  print $currency_span->nodeValue;
} else {
  print "nothing found";
}

prints

631.0075 USD

Wrap that in a function and you are good to go.

You might want to skim through an XPath tutorial if you've never use XPath before.