How to extract a url when the domain name is not included.

[tommytx]

The code below will parse the web page in web.txt file and extract all full domain
named URLS as long as the domain is included: Now I would also like to extract all
the ones that do not have a domain name listed.... see sample below.

************************ begin code to extract domain URL's ****************
$link = "";

$html = file_get_contents("web.txt");

$urls = '(http|file|ftp)';
$ltrs = '\w';
$gunk = '/#~:.?+=&%@!\-';
$punc = '.:?\-';
$any = "$ltrs$gunk$punc";

preg_match_all("{\b$urls:[$any]+?(?=[$punc]*[^$any]|$)}x",$html,$matches);

//the below prints out all the urls that were extracted.

  foreach ($matches[0] as $u)
  {
    $link =  '?url=' . urlencode($u);
    echo "<A HREF='$link'>$u</A><BR>\n";
  }

// the below prints the total number of url's located.

printf("Output of URLs %d URLs<P>\n", sizeof($matches[0]));


***************** End of code to extract domain urls. ****************

Below urls 3 and 4 have the domain name and extract fine.. what can I add to
extract the non domain urls as in 5 and 6. For best results, I would like to
simple duplicate the above and run it again on the web.txt file after changing the
regex to allow pulling out the short urls..

1. <a href="httpcolon//www.idaho.com/vahud/hud_tom.htm>HUD Homes</a>
2. <a href="httpcolon//www.idaho.com/tomchambers">MLS Homes</a>


3. <a href="buy_sell_house_hampton.htm">buy_sell_house_hampton</a>
4. <a href="newport_news_homes_for_sale.htm">newport_news_homes_for_sale</a>

[effigy]

You need to make the domain portion optional, but your pattern is too general to do so. See if [url=http://www.phpfreaks.com/forums/index.php/topic,123121.0.html]this[/url] topic helps you match URLs better. Once you've got this working, a pattern of[tt] (?:[i]domain_portion[/i])? [/tt]will make the domains optional.