[RESOLVED] regexp woes..

leatherback

Hi All,

I am trying to retrieve a price from a link on a website. I just cannot seem to make the pattern match...

I am trying to match this:

and retrieve the price.

Am I at least close with this one? What should the expression be?


  $regex = '/<a\s[^>]*class="txtPreis"[^>]*href="#">([^"]+)<\/a[^>]*>/si';

laserlight

I tested your regex pattern with the given sample input and it worked. Perhaps you should provide the smallest and simplest runnable code that demonstrates the problem 🙂

leatherback

hm..

That is odd..
Could it be an charset issue then? I am using an fopen() to retrieve a page, which Iralized just now is in charset=ISO-8859-1. My site is in utf8 encoding, as are the scripts..

I would love to get you a subset of the code, but it is a bit complex;

The general trend (and the contents is loaded just fine):

  $contents = file_get_contents($url)
  $regex = '/<a\s[^>]*class="txtPreis"[^>]*href="#">([^"]+)<\/a[^>]*>/si';
  if(preg_match($regex, $contents, $matches)) 
    {
	print_r($matches);
   }
  else
  {
  echo 'no match';
  }

laserlight

Could it be an charset issue then?

Maybe. I tested with:

<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<title>Regex Test</title>
</head>
<body>
<p><?php
$pattern = '/<a\s[^>]*class="txtPreis"[^>]*href="#">([^"]+)<\/a[^>]*>/si';
$subject = '<a class="txtPreis" href="#">&#8364; 24,95 </a>';

if (preg_match($pattern, $subject, $matches)) {
    echo '"' . $matches[1] . '"';
}
?></p>
</body>
</html>

Does my test script work for you?

leatherback

Yes. That works beautifully.

The page I am tryin to parse is (e.g.)

http://www.wmf.de/produkt_detail.asp?nv_id=2390&se_id=96&va_id=34962&sp_id=1

laserlight

Oh, take a look at the source:

<td><a href="#" class="txtPreis">&euro;&nbsp;24,95 </a> </td>

You are trying to match the href="#" attribute after the class="txtPreis" attribute, but in the source the href="#" attribute comes before the class="txtPreis" attribute.

leatherback

😐 Why do these guys have to have different templates for diferent sections of their website.. Pfff...

Laser, you are always good at regexps. You think you can help me whip something up?

I think both parts should be there (href="#" and class="txtPreis") but possibly not in that order. Hm.. THere is the or | statement. Patterns are delineated by [] right?

$pattern = '/<a\s[^{>][class="txtPreis"[^>]} | href="#">([^{"]+])<\/a[^>]*>/si';}

or something?

2) Do you know a good place to start reading on how to get & convert the charsets?

laserlight

The tough part is that they are inconsistent, and even have not well formed HTML (a search shows that they have anchor tags with class="txtPreis" that do not have matching closing tags).

Since you will be second guessing their HTML code anyway, you might as well simplify it and then manually adapt if you find that it has changed. For example:

<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<title>Regex Test</title>
</head>
<body>
<p><?php
$pattern = '/<td>\s*<a\s.*class="txtPreis".*>&euro;&nbsp;\s*(\d+,?\d*)\s*<\/a>\s*<\/td>/i';
$subject = file_get_contents('http://www.wmf.de/produkt_detail.asp?nv_id=2390&se_id=96&va_id=34962&sp_id=1&sid={DCE0E564-BEC9-49C3-A7D8-B36C4E3F3F30}');

if (preg_match($pattern, $subject, $matches)) {
    echo $matches[1];
}
?></p>
</body>
</html>

leatherback

Thx Laser,

I realize the horrible website. Ideally I'd parse their URL and just store a basic url & product ID< but even the specific variable used to define which product to show varies from page to page. It almost looks like the files are hand-written 😃

I'll run with your code, untill II notice that it breaks, and adjust it, I suppose.