Web Hosting Talk : Web Hosting Main Forums : Programming Discussion : PHP, Tidy, and parsing through trees

[Adam Hallett]

As you may know Tidy is a tool that cleans up html. It has a binding for php. One of the things you can do with Tidy is generate a parse tree. This tree contains arrays, which give way to objects. Here is the output of var_dump().

http://www.harddrivecaddy.com/temp/tidy.node.output.txt

I want to create a class that provides back/forward navigation (among other thigns) through the tree but I'm unsure how to write that functionality. Not only do I need to go through all the arrays and objects, but I also need to go through all the childs.

[sea otter]

From a (quick) look at it, I think tidy is return the information in serialized format, so you need to do this:

$str = get_tidy_data();  // or whatever you do...
$results = unserialize($str);

It seems that the unserialzed object will be a numerically-indexed array with two elements. Once you've unserialized the data, you can take a quick look with

echo '<pre>';
print_r($results);
echo '</pre>';

The PHP manual page for unserialize is here: http://us3.php.net/manual/en/function.unserialize.php

As for traversal, well, at that point it's like traversing any array in PHP.

Alternatively, if you want to provide a fancy interactive Javascript tree browser, you might want to see if tidy can return the data in JSON format, or if there's a php-serialized-to-json converter out there (if not, it's a simple two-step affair to go from the unserialized array to JSON).

The JSON page is here http://json.org/

[Adam Hallett]

$tidy = new tidy;
$tidy->parseString($file, $config, 'utf8');
$tree = tidy_get_root($tidy);

This produces an array. So there is no need to unserialize() the object. I'm looking an algorithim to go through all arrays that navigates through objects to find more arrays.

for example

$tmp = &$tree[0];

if ( $tmp->isChild() ) {
//now we've found an array, possibly a collection of arrays!

}

[sea otter]

Ah...we're not dealing with regular arrays here. There's an example function to do what you want on the PHP documentation page:

http://us3.php.net/tidy_get_root