CODEWITHSUNDEEP
phphtmlxmlxpath
Yogie
Yogie

Reputation: 1026

Xpath query for HTML table within XML in PHP DOMDocument

I have an XML file with following tree structure.

<rss xmlns:dc="http://purl.org/dc/elements/1.1/"  xmlns:media="http://search.yahoo.com/mrss/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0">
<channel>
    <title>Videos</title>
    <link>https://www.example.com/r/videos/</link>
    <description>A long description of the video.</description>
    <image>...</image>
    <atom:link rel="self" href="http://www.example.com/videos/.xml" type="application/rss+xml"/>
    <item>
        <title>The most used Jazz lick in history.</title>
        <link>
        http://www.example.com/
        </link>
        <guid isPermaLink="true">
         http://www.example.com/
        </guid>
    <pubDate>Mon, 07 Sep 2015 14:43:34 +0000</pubDate>
    <description>
    <table>
        <tr>
            <td>
                <a href="http://www.example.com/">
                    <img src="http://www.example.com/.jpg" alt="The most used Jazz lick in history." title="The most used Jazz lick in history." />
                </a>
            </td>
            <td> submitted by 
                <a href="http://www.example.com/"> jcepiano </a>
                <br/>
                <a href="http://www.youtube.com/">[link]</a>
                <a href="http://www.example.com/">
                    [508 comments]
                </a>
            </td>
        </tr>
    </table>
    </description>
    <media:title>The most used Jazz lick in history.</media:title>
    <media:thumbnail url="http://example.jpg"/>
    </item>
</channel>
</rss>

Here, the html table element is embedded inside XML and that's confusing me.

Now I want to pick the text node values for //channel/item/title and href value for //channel/item/description/table/tr/td[1]/a[1] (with a text node value = "[link]")

Above in 2nd case, I am looking for the value of 2nd a (with a text node value = "[link]"), inside 2nd td inside tr, table, description, item, channel.

I am using PHP DOMDocument();

I have been looking for a perfect solution for this for 2 days now, can you please let me know how would this happen?

Also I need to count the total number of items in the feed, right now I am doing like this:

...
$queryResult = $xpathvar->query('//item/title');
$total = 1;
foreach($queryResult as $result){
           $total++;
}
echo $title;

And I also need a reference link for XPath query selectors' rules.

Thanks in advance! :)

Upvotes: 3

Views: 1223

Answers (2)

Yogie
Yogie

Reputation: 1026

I finally could make it work with the code below

$url = "https://www.example.com/r/videos/.xml";
$feed_dom = new domDocument; 
$feed_dom->load($url); 
$feed_dom->preserveWhiteSpace = false;
$items = $feed_dom->getElementsByTagName('item');

foreach($items as $item){
    $title = $item->getElementsByTagName('title')->item(0)->nodeValue;
    $desc_table = $item->getElementsByTagName('description')->item(0)->nodeValue;
    echo $title . "<br>";

    $table_dom = new domDocument;
    $table_dom->loadHTML($desc_table);
    $xpath = new DOMXpath($table_dom);
    $table_dom->preserveWhiteSpace = false;
    $yt_link_node = $xpath->query("//table/tr/td[2]/a[2]");

    foreach($yt_link_node as $yt_link){

        $yt = $yt_link->getAttribute('href');
        echo $yt . "<br>";
        echo "<br>";
    }
}

I thank Abel, your help was greatly useful to achieve the tasks. :)

Upvotes: 0

Abel
Abel

Reputation: 57169

You wrote that you wanted the length of the result set of the following query:

$queryResult = $xpathvar->query('//item/title');

I assume that $xpathvar here is of type DOMXPath. If so, it has a length property as described here. Instead of using foreach, simply use:

$length = $xpathvar->query('//item/title')->length;

Now I want to pick the text node values for //channel/item/title

Which you can get with the expression //channel/item/title/text().

and href value for //channel/item/description/table/tr/td[1]/a[1] (with a text node value = "[link]")

Your expression here selects any tr, the first td under that, then the first a. But the first a does not have a value of "[link]" in your source. If you want that, though, you can use:

//channel/item/description/table/tr/td[1]/a[1]/@href

but it looks like you rather want:

//channel/item/description/table/tr/td/a[. = "[link]"][1]/@href

which finds the first a element in the tree that has the value (text node) that is "[link]".

Above in 2nd case, I am looking for the value of 2nd a (with a text node value = "[link]"), inside 2nd td inside tr, table, description, item, channel.

Not sure if this was a separate question or meant to explain the previous one. Regardless, the answer the same as in the previous one, unless you explicitly want to search for 2nd a etc (i.e., search by position), in which case you can use numeric predicates.

Note: you start most of your expressions with //expr, which essentially means: search the whole tree at any depth for the expression expr. This is potentially expensive and if all you need is a (relative) root node for which you know the starting point or expression, it is better, and far more performant, to use a direct path. In your case, you can replace //channel for /*/channel (because it is the first under the root element).

Upvotes: 1

Related Questions