JQueryXML 错误“请求的资源上没有‘ Access-Control-allow-Origin’标头。”

我工作的这个个人项目,我只是为了好玩,我想读取一个位于 http://www.ecb.europa.eu/stats/eurofxref/eurofxref-daily.xml的 xml 文件,并解析这个 xml 文件,并使用它来转换货币之间的价值。

到目前为止,为了读取 XML,我已经编写了下面的代码,这些代码非常基本,但是我得到了以下错误。

XMLHttpRequest 无法加载 * * * * 。没有“访问控制-允许-起源” 标头出现在请求的资源上 因此,「 http://run.jsbin.com 」不能进入。

$(document).ready(
function() {
$.ajax({
type:  'GET',
url:   'http://www.ecb.europa.eu/stats/eurofxref/eurofxref-daily.xml',
dataType: 'xml',
success: function(xml){
alert('aaa');
}
});
}
);

我没有看到我的代码有什么问题,所以我希望有人能够指出我的代码做错了什么,以及我可以如何修复它。

281435 次浏览

You won't be able to make an ajax call to http://www.ecb.europa.eu/stats/eurofxref/eurofxref-daily.xml from a file deployed at http://run.jsbin.com due to the same-origin policy.


As the source (aka origin) page and the target URL are at different domains (run.jsbin.com and www.ecb.europa.eu), your code is actually attempting to make a Cross-domain (CORS) request, not an ordinary GET.

In a few words, the same-origin policy says that browsers should only allow ajax calls to services at the same domain of the HTML page.


Example:

A page at http://www.example.com/myPage.html can only directly request services that are at http://www.example.com, like http://www.example.com/api/myService. If the service is hosted at another domain (say http://www.ok.com/api/myService), the browser won't make the call directly (as you'd expect). Instead, it will try to make a CORS request.

To put it shortly, to perform a (CORS) request* across different domains, your browser:

  • Will include an Origin header in the original request (with the page's domain as value) and perform it as usual; and then
  • Only if the server response to that request contains the adequate headers (Access-Control-Allow-Origin is one of them) allowing the CORS request, the browse will complete the call (almost** exactly the way it would if the HTML page was at the same domain).
    • If the expected headers don't come, the browser simply gives up (like it did to you).


* The above depicts the steps in a simple request, such as a regular GET with no fancy headers. If the request is not simple (like a POST with application/json as content type), the browser will hold it a moment, and, before fulfilling it, will first send an OPTIONS request to the target URL. Like above, it only will continue if the response to this OPTIONS request contains the CORS headers. This OPTIONS call is known as preflight request.
** I'm saying almost because there are other differences between regular calls and CORS calls. An important one is that some headers, even if present in the response, will not be picked up by the browser if they aren't included in the Access-Control-Expose-Headers header.


How to fix it?

Was it just a typo? Sometimes the JavaScript code has just a typo in the target domain. Have you checked? If the page is at www.example.com it will only make regular calls to www.example.com! Other URLs, such as api.example.com or even example.com or www.example.com:8080 are considered different domains by the browser! Yes, if the port is different, then it is a different domain!

Add the headers. The simplest way to enable CORS is by adding the necessary headers (as Access-Control-Allow-Origin) to the server's responses. (Each server/language has a way to do that - check some solutions here.)

Last resort: If you don't have server-side access to the service, you can also mirror it (through tools such as reverse proxies), and include all the necessary headers there.

There's a kind of hack-tastic way to do it if you have php enabled on your server. Change this line:

url:   'http://www.ecb.europa.eu/stats/eurofxref/eurofxref-daily.xml',

to this line:

url: '/path/to/phpscript.php',

and then in the php script (if you have permission to use the file_get_contents() function):

<?php


header('Content-type: application/xml');
echo file_get_contents("http://www.ecb.europa.eu/stats/eurofxref/eurofxref-daily.xml");


?>

Php doesn't seem to mind if that url is from a different origin. Like I said, this is a hacky answer, and I'm sure there's something wrong with it, but it works for me.

Edit: If you want to cache the result in php, here's the php file you would use:

<?php


$cacheName = 'somefile.xml.cache';
// generate the cache version if it doesn't exist or it's too old!
$ageInSeconds = 3600; // one hour
if(!file_exists($cacheName) || filemtime($cacheName) > time() + $ageInSeconds) {
$contents = file_get_contents('http://www.ecb.europa.eu/stats/eurofxref/eurofxref-daily.xml');
file_put_contents($cacheName, $contents);
}


$xml = simplexml_load_file($cacheName);


header('Content-type: application/xml');
echo $xml;


?>

Caching code take from here.