Parse html using PHP and loop through table rows and columns?

Question

I'm trying to parse HTML from loadHTML but I'm having trouble, I managed to loop through all <tr>s in the document but I don't know how to loop through the <td> s on each row.

This is what I did so far:

$DOM->loadHTML($url);
$rows= $DOM->getElementsByTagName('tr');

for ($i = 0; $i < $rows->length; $i++) { // loop through rows
    // loop through columns
    ...
}

How can I get loop through the columns in each row?

Easier-to-use [wrappers](http://stackoverflow.com/questions/3577641/how-to-parse-and-process-html-xml-with-php) around the DOM methods exist, specifically for looping over element collections. — mario, Jan 09 '13 at 21:10

Salman A · Accepted Answer · 2013-01-09T21:43:12.553

8

DOMElement also supports getElementsByTagName:

$DOM = new DOMDocument();
$DOM->loadHTMLFile("file path or url");
$rows = $DOM->getElementsByTagName("tr");
for ($i = 0; $i < $rows->length; $i++) {
    $cols = $rows->item($i)->getElementsbyTagName("td");
    for ($j = 0; $j < $cols->length; $j++) {
        echo $cols->item($j)->nodeValue, "\t";
        // you can also use DOMElement::textContent
        // echo $cols->item($j)->textContent, "\t";
    }
    echo "\n";
}

edited Jan 09 '13 at 21:43

answered Jan 09 '13 at 21:20

Salman A

262,204
82
430
521

I haven't been able to echo the col content inside the loop. I tried `echo $cols->item($i)->nodeValue;`, could you edit it? I'll take this one if it works as it's easier to implement in my case – lisovaccaro Jan 09 '13 at 21:32
I have made minor changes to the code. See if it works. And see if the column is not empty. – Salman A Jan 09 '13 at 21:41

nickb · Answer 2 · 2013-01-09T21:35:13.490

2

Use DOMXPath to query out the child column nodes with a relative xpath query, like this:

$xpath = new DOMXPath( $DOM);
$rows= $xpath->query('//table/tr');

foreach( $rows as $row) {
    $cols = $xpath->query( 'td', $row); // Get the <td> elements that are children of this <tr>
    foreach( $cols as $col) {
        echo $col->textContent;
    }
}

Edit: To start at specific rows and stop, keep your own index on the row by changing how you're iterating over the DOMNodeList:

$xpath = new DOMXPath( $DOM);
$rows= $xpath->query('//table/tr');

for( $i = 3, $max = $rows->length - 2; $i < $max, $i++) {
    $row = $rows->item( $i);
    $cols = $xpath->query( 'td', $row);
    foreach( $cols as $col) {
        echo $col->textContent;
    }
}

edited Jan 09 '13 at 21:35

answered Jan 09 '13 at 21:16

nickb

59,313
13
108
143

this works, I just have a problem, how can I start from row 3 and end in totalrows - 2? I was using `($i = 3; $i < $rows->length -2; $i++)` before for the loop – lisovaccaro Jan 09 '13 at 21:27
@Liso - You can keep those counts yourself, I'll update my answer – nickb Jan 09 '13 at 21:32
@Liso - All `$xpath->query()` is giving you back is a DOMNodeList, so you can iterate over it just the same as you were before. The benefit is that now, instead of just using `getElementsByTagName()`, you have much more control over what actually gets put in that DOMNodeList. Try my updated solution, it should work for your requirements. – nickb Jan 09 '13 at 21:41

Zak · Answer 3 · 2013-01-09T21:20:34.117

Would re-looping work?

$DOM->loadHTML($url);
$rows= $DOM->getElementsByTagName('tr');
$tds= $DOM->getElementsByTagName('td');

for ($i = 0; $i < $rows->length; $i++) {
// loop through columns
     for ($i = 0; $i < $tds->length; $i++) {
     // loop through rows

     }

}

EDIT You will also have to check the parent node to make sure that the rows parent is the tr you are currently in. Something like

if ($rows == tds->parent_node){
// do whatever
}

May not be syntactically 100% correct, but the concept is sound.

Parse html using PHP and loop through table rows and columns?

3 Answers3

Linked