Stripping HTML

dast

I have a client who has a site that he wants stripping to put the info in an Excel sheet because his old developer is playing funny games.

So for an example I need to find this DIV in the HTML

Code:

<div><strong><a href="http://www.test.com/online_store/colour/white.cfm">White</a></strong></span>

</div>

And just get the word 'White' into a variable. There are also other div ID's such as brand, type etc that I need to do this with. Could anyone give me some advice/help?

I was thinking if there is a way to find the line <div id="Colour"> and the end of that DIV Then just stripping everything within tags in that DIV.

Do you know of any way or the best way to do that?

Much appreciated!!

nrg_alpha

Perhaps some DOMDocument class execution?

$classOrIdArray = array('Colour'); // list of div classes / ids
$divContentArray = array();

$dom = new DOMDocument();
@$dom->loadHTMLFile('your_Html_File_Here'); // @ suppresses warnings in the event the html in question is invalid
$div = $dom->getElementsByTagName('div');
foreach($div as $divValue){
    foreach($classOrIdArray as $classId){
        if($divValue->hasAttribute('class') and $divValue->getAttribute('class') == $classId or $divValue->hasAttribute('id') and $divValue->getAttribute('id') == $classId){
            $divContentArray[] = trim($divValue->textContent);
        }
    }
}
echo '<pre>'.print_r($divContentArray, true);

Ouput:

Array
(
    [0] => White
)

EDIT - The output is as such as I dumped the following line into an HTML file:

<div id="Colour"><div><strong><a href="http://www.test.com/online_store/colour/white.cfm">White</a></strong></span></div>

you can add any additional div class or id names to the $classOrIdArray array.
Obviously, substitute your_Html_File_Here with your appropriate html file path and name.

Weedpacket

$xpath = new DOMXPath($dom);
$colour = $xpath->query('//*[@id="Colour"]//a')->item(0)->nodeValue;
echo $colour;

nrg_alpha

Weedpacket;10905874 wrote:

$xpath = new DOMXPath($dom);
$colour = $xpath->query('//*[@id="Colour"]//a')->item(0)->nodeValue;
echo $colour;

Damn slick as usual, WP!
I was looking up this DOMXPath in the php manual. I was intrigued by some of the initial examples..

So I can see that it is basically a path-type navigation to a specific tag.

//$elements = $xpath->query("/html/body/div[@id='yourTagIdHere']");

I get the wild card substitute as an example to shortcut the DOM node path as in /div[@id=...... (which I suppose you could have used too, as when I substituted your initial / with , it worked all the same). I was thrown off by the //a bit after the closing square bracket. Then it donned on me, is this access to the <a> tag nested within the div tag? This is the only thing I can think of given the example HTML example by the OP.. if so, I'm not sure what's going on with the item(0) business... Care to elaborate a bit?

Weedpacket

nrg_alpha wrote:
Then it donned on me, is this access to the <a> tag nested within the div tag?

XPath. "//" is an abbreviation for "/descendant-or-self::node()/" - the current node and all of its descendants - the entire subtree from that point down. Since the path started with '/', the current node is the document root, and therefore starting with '//' nets you every node in the entire document.

"*" represents all element children of the current node(s) - as opposed to children that are attributes, text, comment, and whatnot. It's really only there because I needed something to hang the predicate off of.

"a" is short for "child::a" - children of the current node(s) that are <a> elements.

I'm not sure what's going on with the item(0) business...

DomXPath::query() returns a DomNodeList, and I'm just interested in the item at the head of the list, that's item(0). I don't want the <a> element itself, only its text content.

What gets me is $dom->getElementById('Colour') doesn't work because there's no inherent mapping between the attribute named "id", and the id type. Even if you specify loading the document as HTML.

You can specify which attribute is the ID by using setIdAttribute on the element, but to do that you have to already have the element – so you no longer need to get it, by ID or otherwise.

nrg_alpha

Thanks for the reply, WP. Clears things up somewhat. I think I will have to invest some sizable time in really getting comfortable with DOMDocument and DOMXpath, as I can see it really could make life easier in tracking down things within specific nodes by example.

Admittedly, with the amount of methods (inherited or not) there are, it is going to take some time to really get on the ball with. But seems worthwhile for ease of DOM navigation / manipulation down the road.

Thanks again for your DOMXPath sample and explanations.. really appreciate it.

Cheers