Question

I am using DOMDocument to parse an html document and get some data out of it. Following is the structure of sub-tree of DOM

<div id="tab1">
    <div class="some class name"></div>
    <div class="some other class name">arbitrary data and nodes</div>
    <p> lot of paragraphs to follow </p>
    <p> paragraphs </p>
    <p> paragraphs </p>
    <p> paragraphs </p>
    <p> paragraphs </p>
    <br />
    <br />
    <br />
    <br />
    <br />
    <table />
    <table />
    <table />
    <table />
</div>

I do not want first two children of tab1. I am using following PHP Code

<?php 
$urlArray = file('sitemap.txt');
$dataSet = array();
foreach($urlArray as $url){
    $scrapedData = file_get_contents('./scraped-site/'.trim($url));
    $doc = new DOMDocument();
    @$doc->loadHTML($scrapedData);
    $domXpathDoc = new DOMXPath($doc);
    $results = '';
    $xpathArray = array(
        'info'=>'//*[@id="tabs1"]',
    );
    $set = array();
    foreach($xpathArray as $field => $xpath){
        $domNodeList = $domXpathDoc->query($xpath);
        foreach($domNodeList as $node){
            foreach ($node->childNodes as $child) {
                $set[] = $child->ownerDocument->saveXML( $child );
            }
        }
    }
    $dataSet[] = $set;
}

The code given gives me all children how can I selectively avoid any node?

Answer 1

[EDIT2: I tried the answer below (I learned 🙂 ). This is working for me:

"//*[@id='tabs1'][name() != 'div']"

Basically it tells the xpath to ignore all elements with name ‘div’. You can ignore more than one element like this:

"//*[@id='tabs1'][name() != 'div' and name() != 'foo']"

Only showing elements after the first two would work like this:

"//*[@id='tabs1'][position()>2]"