PHP XML Parsing Basics -- A Tutorial

Parsing XML files in PHP is complex. But, like most things, it's not that hard once you get the hang of it. I don't know about you, but I learn best by seeing actual code. So, here's a real world example.
If you're intereted in some more in-depth information on this subject, take a look at a presentation I gave entitled Parsing, Validating and Saving Data from Complex XML Streams.
#! /usr/local/bin/php -q
<?php

/**
 * A basic example of how to parse XML files.
 *
 * <p>Relies on PHP's {@link http://php.net/ref.xml SAX parser}.</p>
 *
 * <p>This example grabs stock quotes from NASDAQ's website.  You tell
 * the program which stocks to check by adding/removing elements of the
 * <var>$Symbols</var> array. The script outputs a series
 * of fake SQL query strings.</p>
 *
 * <p>Output is placed into an associative array called <var>$Data.</var></p>
 *
 * <p>Element names become the array's keys.  So:
 *    <br /><code><TAG>The info</TAG></code>
 * <br />Would become:
 *    <br /><code>$Data['TAG'] = 'The info';</code>
 * </p>
 *
 * <p>Where elements have attributes, the array's key is the name of
 * the element followed by ":" and the name of the attribute:
 *    <br /><code><TAG ID="55">The info</TAG></code>
 * <br />becomes...
 *    <br /><code>$Data['TAG'] = 'The info';
 *    <br />$Data['TAG:ID'] = 55;</code>
 * </p>
 *
 * <p>Do note, "case folding" is in effect, so lower case tag and
 * attribute names are converted to upper case.</p>
 *
 * <p>Requires PHP version 4.3.0 or later.</p>
 *
 * @author     Daniel Convissor <danielc@AnalysisAndSolutions.com>
 * @copyright  The Analysis and Solutions Company, 2002-2004
 * @version    2004-02-11 11:00:00
 * @link       http://www.AnalysisAndSolutions.com/code/phpxml.htm
 * @link       http://php.net/ref.xml
 */


//  ?  //  Which ticker symbols do you want to evaluate?

$Symbols = array('ET', 'EK');



//  Set the base URI.
$URI = 'http://quotes.nasdaq.com/quote.dll?page=xml&mode=stock&symbol=';

//  Make sure there's no other data with these names.
$ParserProbs = array();
$DataProbs   = array();

//  Array to convert XML entities back to plain text.
$XmlEntities = array(
    '&amp;'  => '&',
    '&lt;'   => '<',
    '&gt;'   => '>',
    '&apos;' => '\'',
    '&quot;' => '"',
);


/**
 * Runs each time an XML element starts.
 */
function StartHandler(&$Parser, &$Elem, &$Attr) {
    global $Data, $CData, $XmlEntities;

    // Start with empty CData array.
    $CData = array();

    // Put each attribute into the Data array.
    foreach ($Attr as $Key => $Value) {
        $Data["$Elem:$Key"] = strtr(trim($Value), $XmlEntities);
        // debug //  echo "$Elem:$Key = {$Data["$Elem:$Key"]}\n";
    }
}


/**
 * Runs each time XML character data is encountered.
 */
function CharacterHandler(&$Parser, &$Line) {
    global $CData;

    /*
     * Place lines into an array because elements
     * can contain more than one line of data.
     */
    $CData[] = $Line;
}


/**
 * Runs each time an XML element ends.
 */
function EndHandler(&$Parser, &$Elem) {
    global $Data, $CData, $DataProbs, $Sym, $XmlEntities;

    /*
     * Mush all of the CData lines into a string
     * and put it into the $Data array.
     */
    $Data[$Elem] = strtr( trim( implode('', $CData) ), $XmlEntities);

    // debug //    echo "$Elem = {$Data[$Elem]}\n";

    switch ($Elem) {
        case 'LAST-SALE-PRICE':
            // Make sure the data is clean.
            if ( !preg_match('/^\d{1,8}(\.\d{1,2})?$/', $Data[$Elem]) ) {
                // Make note of the error.
                $DataProbs[] = "$Elem has bad format: {$Data[$Elem]}";
            }
            break;

        case 'TRADE-DATETIME':
            /*
             * Ensure data is clean, plus save match parts to $Atom,
             * which will be used to convert date/time to MySQL format.
             */
            if ( !preg_match('/^(\d{4})(\d{2})(\d{2}) (\d{2}:\d{2}:\d{2})$/',
                    $Data[$Elem], $Atom) ) {
                // Make note of the error.
                $DataProbs[] = "$Elem has bad format: {$Data[$Elem]}";
            }

            $Data[$Elem] = "$Atom[1]-$Atom[2]-$Atom[3] $Atom[4]";

            break;

        case 'EQUITY-QUOTE':
            // Final item tag.  Do something with the data.

            // Make sure the data is clean.
            if ( !preg_match('/^\w{1,9}$/', $Data['EQUITY-QUOTE:CUSIP']) ) {
                // Make note of the error.
                $DataProbs[] = "$Elem has bad format: "
                           . $Data['EQUITY-QUOTE:CUSIP'];
            }

            /*
             * Double check that all of the needed data was set.
             * If it's not, we don't want to run the query,
             * so skip the rest of this section.
             */
            if ( !isset($Data['LAST-SALE-PRICE']) ) {
                $DataProbs[] = "$Sym LAST-SALE-PRICE wasn't set";
            }
            if ( !isset($Data['TRADE-DATETIME']) ) {
                $DataProbs[] = "$Sym TRADE-DATETIME wasn't set";
            }

            if ( count($DataProbs) ) {
                echo "\nData for $Sym had problems:\n";
                echo implode("\n", $DataProbs) . "\n\n";
                $DataProbs = array();
            } else {
                // Construct a sample query string.
                $Query = 'UPDATE Quotes SET '
                        . "TradePrice={$Data['LAST-SALE-PRICE']}, "
                        . "TradeTime='{$Data['TRADE-DATETIME']}' "
                        . "WHERE CUSIP='{$Data['EQUITY-QUOTE:CUSIP']}'";

                /*
                 * In the real world, you could run the query now. But,
                 * for the sake of this exercise, let's just look at it.
                 */
                echo "$Query\n";
            }
    }
}




// Loop through each ticker symbol.
foreach ($Symbols as $Sym) {

    /*
     * Grab the file and stick it into an array.
     * Next, check to see that you actually got the raw info.
     * Then, implode the raw info into one long string.
     *
     * If your data is already in string form, you don't need these steps.
     *
     * This one step requires PHP to be at version 4.3.0 or later.
     */
    $Contents = @file_get_contents("$URI$Sym");

    if (!$Contents) {
        $ParserProbs[] = "$URI$Sym\n    Had problem opening file.";
        /*
         * Start the while loop over again, this time with the
         * next item in the $Symbols array.
         */
        continue;
    }

    // debug //    echo "\n\n$URI$Sym\n";
    // debug //    echo "$Contents";


    /*
     * Take care of characters that choke the parser.
     * While I don't think NASDAQ's data poses these problems,
     * it's good to keep them in mind.
     */

    // Escape ampersands that aren't part of entities.
    $Contents = preg_replace('/&(?!\w{2,6};)/', '&amp;', $Contents);

    // Remove all non-visible characters except SP, TAB, LF and CR.
    $Contents = preg_replace('/[^\x20-\x7E\x09\x0A\x0D]/', "\n", $Contents);

    /*
     * Clean out the Data array so it can be reused
     * to hold the data we parse from the file.
     */
    $Data = array();

    // Initialize the parser.
    $Parser = xml_parser_create('ISO-8859-1');
    xml_set_element_handler($Parser, 'StartHandler', 'EndHandler');
    xml_set_character_data_handler($Parser, 'CharacterHandler');

    // Pass the content string to the parser.
    if ( !xml_parse($Parser, $Contents, TRUE) ) {
        $ParserProbs[] = "Had problem parsing data for $Sym:\n    "
                . xml_error_string(xml_get_error_code($Parser));
    }
}


// Problems?
if ( count($ParserProbs) ) {
    echo "\n" . implode("\n", $ParserProbs);
}

?>