Failed to execute 'btoa' on 'Window

Question

The error in the title is thrown only in Google Chrome, according to my tests. I'm base64 encoding a big XML file so that it can be downloaded:

this.loader.src = "data:application/x-forcedownload;base64,"+
                  btoa("<?xml version=\"1.0\" encoding=\"utf-8\"?>"
                  +"<"+this.gamesave.tagName+">"
                  +this.xml.firstChild.innerHTML
                  +"</"+this.gamesave.tagName+">");

this.loader is hidden iframe.

This error is actually quite a change because normally, Google Chrome would crash upon btoa call. Mozilla Firefox has no problems here, so the issue is browser related. I'm not aware of any strange characters in file. Actually I do believe there are no non-ascii characters.

Q: How do I find the problematic characters and replace them so that Chrome stops complaining?

_{I have tried to use Downloadify to initiate the download, but it does not work. It's unreliable and throws no errors to allow debug.}

Answer 1

As an complement to Stefan Steiger answer: (as it doesn't look nice as a comment)

Extending String prototype:

String.prototype.b64encode = function() { 
    return btoa(unescape(encodeURIComponent(this))); 
};
String.prototype.b64decode = function() { 
    return decodeURIComponent(escape(atob(this))); 
};

Usage:

var str = "äöüÄÖÜçéèñ";
var encoded = str.b64encode();
console.log( encoded.b64decode() );

NOTE:

As stated in the comments, using unescape is not recommended as it may be removed in the future:

Warning: Although unescape() is not strictly deprecated (as in "removed from the Web standards"), it is defined in Annex B of the ECMA-262 standard, whose introduction states: … All of the language features and behaviours specified in this annex have one or more undesirable characteristics and in the absence of legacy usage would be removed from this specification.

Note: Do not use unescape to decode URIs, use decodeURI or decodeURIComponent instead.