Searches Through XML with XPath

Directly accessing and traversing through an XML document, is handy when you know the exact format and ordering of the document in question. Often you may be presented with a document that has varying specifications. The data that you want to access might exist under multiple tags. Or you simply may have a large XML document and know that you want only one small piece of data from it.

In these cases, traversing the document tree may not be the best use of your time. Instead, you want to search through the document for the exact item (or items) that you need. This capability is available also through the SimpleXML extension. Within it, you can perform XPath searches. XPath is a standard for performing these kinds of searches through XML.

You perform a search via xpath('searchstring'). The basic format of an XPath search is of the form: "/x/y/z", where x, y, and z are nested XML tags, such as of the form: <x><y><z/></y></x>. Starting the XPath with a "/" as we did means that "x" had to be at the top level. Had that been left off, such as in "x/y/z", it would match that combination of elements at any depth in the XML structure.

Some common XPath queries are explained in the following list; these different queries can be combined in infinite combinations to generate powerful searches:

x Matches any tag named x

x/y Matches any tag named y, directly contained in a tag named x

x/y/.. Similar to the preceding item but actually matches and returns tag x, instead of tag y

x//y Matches any tag named y that is a descendant of a tag named x (any number of levels deep)

x[5] Matches the fifth tag named x

x[last()] Matches the last tag named x

x[@att] Matches any tag named x, with an attribute named att

x[@att="val"] Matches any tag named x, with an attribute named att that has the value "val"

Many more options exist; these are just the most common. To discover more about XPath and all its various options, read the W3C's specification for XPath, found at http://www.w3.org/TR/xpath.

Example Searches of an XML Document

<?php
// Using SimpleXML, read the file into memory:
$xml = simplexml_load_file('contacts.xml');

// Xpath search to find all 'meta' tags no matter what the depth:
$meta = $xml->xpath('//meta');

// Loop over them and output their IDs
foreach ($meta as $m) {
    echo "Meta - {$m['id']}<br />\n";
}

// Find all email tags within a contact tag from the root of the XML document:
$email = $xml->xpath('/contacts/contact/email');

// Loop over them and output the email addreses:
foreach ($email as $e) {
    echo "Email - {$e}<br />\n";
}

// Finally, find any contact who has a cellphone number
$cell = $xml->xpath('contact/phone[@type="cell"]/..');

// Loop over them and output their names:
foreach ($cell as $c) {
    echo "Cell Contact - {$c->name}<br />\n";
}
?>

Hope it helps.