UTF-8 with PHP DOMDocument loadHTML?

2019-05-10 20:42发布

Consider this example, test.php:

<?php
$mystr = "<p>Hello, με काचं  ça øy jeść</p>";
var_dump($mystr);
$domdoc = new DOMDocument('1.0', 'utf-8'); //DOMDocument();
$domdoc->loadHTML($mystr); // already here corrupt UTF-8?
var_dump($domdoc);
?>

If I run this with PHP 5.5.9 (cli), I get in terminal:

$ php test.php 
string(50) "<p>Hello, με काचं  ça øy jeść</p>"
object(DOMDocument)#1 (34) {
  ["doctype"]=>
  string(22) "(object value omitted)"
...
  ["actualEncoding"]=>
  NULL
  ["encoding"]=>
  NULL
  ["xmlEncoding"]=>
  NULL
...
  ["textContent"]=>
  string(70) "Hello, με à¤à¤¾à¤à¤  ça øy jeÅÄ"
}

Clearly, the original string is correct as UTF-8, but the textContent of the DOMDocument is incorrectly encoded.

So, how can I get the content as correct UTF-8 in the DOMDocument?

标签: php utf-8
2条回答
叼着烟拽天下
2楼-- · 2019-05-10 21:28

The DOM extension was built on libxml2 whose HTML parser was made for HTML 4 - the default encoding for which is ISO-8859-1. Unless it encounters an appropriate meta tag or XML declaration stating otherwise loadHTML() will assume the content is ISO-8859-1.

Specifying the encoding when creating the DOMDocument as you have does not influence what the parser does - loading HTML (or XML) replaces both the xml version and encoding that you gave its constructor.


Workarounds:

First use mb_convert_encoding() to translate anything above the ASCII range into its html entity equivalent.

$domdoc->loadHTML(mb_convert_encoding($mystr, 'HTML-ENTITIES', 'UTF-8'));

Or hack in a meta tag or xml declaration specifying UTF-8.

$domdoc->loadHTML('<meta http-equiv="Content-Type" content="charset=utf-8" />' . $mystr);
$domdoc->loadHTML('<?xml encoding="UTF-8">' . $mystr);
查看更多
Summer. ? 凉城
3楼-- · 2019-05-10 21:34

Just wanted to post the OP code with the fixes that work for me:

<?php
$mystr = "<p>Hello, με काचं  ça øy jeść</p>";
var_dump($mystr);
$domdoc = new DOMDocument('1.0', 'UTF-8'); //DOMDocument();
$domdoc->substituteEntities = true; // no effect if hack is done
//~ $domdoc->actualEncoding = 'UTF-8'; // Cannot write property
$domdoc->encoding = 'UTF-8'; // no effect
//~ $domdoc->xmlEncoding = 'UTF-8'; // Cannot write property
//~ $domdoc->loadHTML($mystr); // already here corrupt UTF-8?
//~ $domdoc->loadHTML(utf8_decode($mystr)); // this gets to <p>Hello, ?? ?????  ça øy je??</p>, so not all
//~ $domdoc->loadHTML( mb_convert_encoding($mystr, 'utf-8', mb_detect_encoding($mystr)) ); // no dice
$domdoc->loadHTML('<?xml encoding="UTF-8">' . $mystr); // hack, http://php.net/manual/en/domdocument.loadhtml.php#95251
// dirty fix
foreach ($domdoc->childNodes as $item)
    if ($item->nodeType == XML_PI_NODE)
        $domdoc->removeChild($item); // remove hack
$domdoc->encoding = 'UTF-8'; // insert proper (sets all three)
var_dump($domdoc);
print $domdoc->saveXML(); // without ->encoding = 'UTF-8': Hello, &#x3BC;&#x3B5; &#xFEFF;&#x915;&#x93E;&#x91A;&#x902; else OK
//~ print mb_convert_encoding($domdoc->saveXML(), 'UTF-8', 'HTML-ENTITIES'); // if without ->encoding = 'UTF-8', this is then OK: <p>Hello, με काचं  ça øy jeść</p>
?>

This outputs:

$ php test.php 
string(50) "<p>Hello, με काचं  ça øy jeść</p>"
object(DOMDocument)#1 (34) {
  ["doctype"]=>
  string(22) "(object value omitted)"
...
  ["actualEncoding"]=>
  string(5) "UTF-8"
  ["encoding"]=>
  string(5) "UTF-8"
  ["xmlEncoding"]=>
  string(5) "UTF-8"
...
  ["textContent"]=>
  string(43) "Hello, με काचं  ça øy jeść"
}
<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">
<html><body><p>Hello, με काचं  ça øy jeść</p></body></html>

... which is all good now :)

查看更多
登录 后发表回答