可以将文章内容翻译成中文,广告屏蔽插件可能会导致该功能失效(如失效，请关闭广告屏蔽插件后再试):

问题:

I have an XHTML document being passed to a PHP app via Greasemonkey AJAX. The PHP app uses UTF8. If I output the POST content straight back to a textarea in the AJAX receiving div, everything is still properly encoded in UTF8.

When I try to parse using XPath

$dom = new DOMDocument();
$dom->loadHTML($raw2);
$xpath = new DOMXPath($dom);
$query = '//td/text()';
$nodes = $xpath->query($query);
foreach($nodes as $node) {
  var_dump($node->wholeText);
}

dumped strings are not utf8. How do I force DOM/XPath to use UTF8?

回答1:

If it is a fully fledged valid xhtml document you shouldn't use loadhtml() but load()/loadxml().

Given the example xhtml document

<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
  "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
    <head>
        <title>xhtml test</title>
    </head>
    <body>
        <h1>A Table</h1>
        <table>
            <tr><th>A</th><th>O</th><th>U</th></tr>
            <tr><td>Ä</td><td>Ö</td><td>Ü</td></tr>
            <tr><td>ä</td><td>ö</td><td>ü</td></tr>
        </table>
    </body>
</html>

the script

<?php
$raw2 = 'test.html';

$dom = new DOMDocument();
$dom->load($raw2);
$xpath = new DOMXPath($dom);
var_dump($xpath->registerNamespace('h', 'http://www.w3.org/1999/xhtml'));
$query = '//h:td/text()';
$nodes = $xpath->query($query);
foreach($nodes as $node) {
    foo($node->wholeText);
}


function foo($s) {
    for($i=0; $i<strlen($s); $i++) {
        printf('%02X ', ord($s[$i]));
    }
    echo "\n";
}

prints

bool(true)
C3 84 
C3 96 
C3 9C 
C3 A4 
C3 B6 
C3 BC

i.e. the output/strings are utf-8 encoded

回答2:

I had the same problem and I couldn't use tidy in my webserver. I found this solution and it worked fine:

$html = mb_convert_encoding($html, 'HTML-ENTITIES', "UTF-8");
$dom = new DomDocument();
$dom->loadHTML($html);

回答3:

I have not tried, but the second parameter of DOMDocument::__construct seems to be related to the encoding ; maybe that'll help you :-)

Else, there is an encoding property in DOMDocument, which is writable.

The DOMXpath beeing constructed with the DOMDocument as parameter, maybe it'll work...

回答4:

A bit late in the game, but perhaps it helps someone...

The problem might be in the output, and not in the dom/xpath object itself.

If you would output the nodeValue directly, you would get corrupted characters e.g.:

Ã¬ÂÂÃ¬ÂÂ Ã«Â¹ÂÃ«Â”Â”Ã¬ÂÂ¤
ìì ë¹ë””ì¤ í°ì  íì¤

You have to load your dom object with the second param "utf-8", new \DomDocument('1.0', 'utf-8'), but still when you print the dom node list/element value you get broken characters:

echo $contentItem->item($index)->nodeValue

you have to wrap it up with utf8_decode:

echo utf8_decode($contentItem->item($index)->nodeValue) //output: 者不終朝而會，愚者可浹旬而學

回答5:

Struggled with similar problem (unable to force Xpath to use UTF-8 in combination with loadHTML), in the end this excellent article provided the solution: http://devzone.zend.com/article/8855