Question

我使用TDD（SimpleTest）创建了我的第一堂课。它工作得很好。此类解析XML配置文件并将其作为数组返回。我怎么能改进它（性能，任何提示）？班级责任怎么样？也许XMLtoArray应该被移到另一个类，我不知道......

<?php
class Configuration
{
    private $domdocument_object;
    private $domxpath_object;

    public function __construct($filename) {
        $this->loadXML($filename);
        $this->domxpath_object = new DOMXPath($this->domdocument_object);
    }

    private function loadXML($filename)
    {
        if (!file_exists($filename))
        {
            throw new ConfigurationException('Configuration file not found');
        }

        $this->domdocument_object = $domdocument_object = new DOMDocument();
        $this->domdocument_object->preserveWhiteSpace = false;

        if (!$this->domdocument_object->load($filename))
        {
            throw new ConfigurationException('Malformed configuration file');
        }
    }

    public function get($path = '/*') {
        $configuration = array();

        $domnodelist_object = $this->domxpath_object->query($path);
        $configuration = $this->XMLToArray($domnodelist_object);

        /**
         * Get a configuration entry as string or array
         *
         * For example:
         * $xml = '<foo><bar>baz</bar></foo>'
         * $path = '/foo/bar/'
         * return just baz, as string instead of an array('baz');
         *
         * Another example:
         * $xml = '<foo><bar>baz</bar><lorem>ipsum</lorem></foo>';
         * $path = '/foo'
         * return just array('bar' => 'baz', 'lorem' => ipsum);
         * instead of array('foo' => array('bar' => 'baz', 'lorem' => ipsum));
         */
        while (!is_string($configuration) && count($configuration) == 1)
        {
            $configuration_values = array_values($configuration);
            $configuration = $configuration_values[0];
        }

        if (empty($configuration))
        {
            $configuration = null;
        }

        return $configuration;
    }

    public function XMLToArray(DOMNodeList $domnodelist_object) {
        $configuration = array();

        foreach ($domnodelist_object as $element)
        {
            if ($element->nodeType == XML_DOCUMENT_NODE)
            {
                if ($element->hasChildNodes())
                {
                    $configuration = $this->XMLToArray($element->childNodes);
                }
            }
            else if ($element->nodeType == XML_ELEMENT_NODE)
            {
                if (!$element->hasChildNodes())
                {
                    $configuration[$element->nodeName] = null;
                }
                else if (
                    $element->firstChild->nodeType == XML_TEXT_NODE ||
                    $element->firstChild->nodeType == XML_CDATA_SECTION_NODE
                )
                {
                    $configuration[$element->nodeName] = $element->nodeValue;
                }
                else if ($element->firstChild->nodeType == XML_ELEMENT_NODE)
                {
                    $configuration[$element->nodeName] = $this->XMLToArray($element->childNodes);
                }
            }
        }

        return $configuration;
    }
}
?>

此类忽略XML属性。谢谢。

Answer 1

对我来说很突出的是你每次执行一个对象时创建一个新对象，你应该在本地存储对象（在对象中），然后它只使用你内存的一部分。

以下是我要做的更改：

class Configuration
{
    private $domdocument_object;
    private $domxpath_object; //+

    public function __construct($filename)
    {
        $this->loadXML($filename);
        $this->domxpath_object = new DOMXPath($this->domdocument_object); //+
    }

    public function get($path = '/*')
    {
        //Remove the following
        $domxpath_object = new DOMXPath($this->domdocument_object);
    }
}

然后将$domxpath_object更改为$this->domxpath_object所需的位置。

但是这应该转移到CoderReview，因为这不是主题。

Answer 2

这对性能不利：

$xml = preg_replace("/>\s+</", "><", $xml);

另外，它不能保证可靠（这可能会以不希望的方式改变评论和CDATA部分）。虽然找到更好的解决方案并不容易。迭代所有文本节点修剪它们将更可靠，但不会更快。

如果你只是想让这个数组成为php的SAX解析器或SimpleXML可能更合适。两种选择都可能更快（我没有测试）。

PHP类改进

2 个答案: