I want to parse different web pages so that I can form an inverted index. I want to read only the text, not the a tag elements,menu, etc. Is it possible to do this? Here is what I have so far:
<?php
$ch = curl_init("http://en.wikipedia.org/wiki/Agile_software_development");
curl_setopt($ch,CURLOPT_RETURNTRANSFER,true);
$c1 = curl_exec($ch);
$dom = new DOMDocument();
@$dom->loadHTML($c1);
$links = $dom->getElementsByTagName("body");
echo "<br>";
foreach($links as $links) {
$title = $links->getElementsBytagName("a");
$l= $title->length;
echo $link->nodeValue;
echo"<br>";
} ?>
I would do it like this:
<?php
$html = <<<HTML
<html>
<head>
<title>TITLE</title>
</head>
<body>
<p>PARA 1</p>
<p>PARA <span>2</span></p>
</body>
</html>
HTML;
$dom = new DOMDocument();
@$dom->loadHtml($html);
var_dump($dom->getElementsByTagName("body")[0]->textContent);
?>
The textContent field gives you the contents of the Node itself and of its descendants, in document order. The output of the above is:
string(25) "
PARA 1
PARA 2
"
If you want to normalize the spaces (replace all sequences of 2 or more spaces with just one space and remove the leading and trailing spaces), then you can do this:
var_dump(preg_replace('/\s{2,}/', ' ', trim(
$dom->getElementsByTagName("body")[0]->textContent)));
You can use XPath to extract it.
$html = <<<'HTML'
<html>
<head>
<title>TEST</title>
</head>
<body>
<h1>HEADER</h1>
<p>SOME CONTENT</p>
</body>
</html>
HTML;
$dom = new DOMDocument();
@$dom->loadHtml($html);
$xpath = new DOMXPath($dom);
var_dump($xpath->evaluate('normalize-space(//body)'));
Output:
"HEADER SOME CONTENT"