Getting binary data. Lost of 128+ chars

[Escain]

Hello!

I am trying to download an image by ajax (for some reasons i will not explain here).
But i am getting all 127-255 bytes seted to 65533.
Example:
From the image

66 77 54 16 0 0 0 0 0 0 54 0 0 0 40 0 0 0 32 0 0 0 32 0 0 0 1 0 32 0 0 0 0 0 0 0 0 0 19 11 0 0 19 11 0 0 0 0 0 0 0 0 0 0 255 255 255 0 255 255 255 0 0 0 0 28 0 0 0 65 0 0 0 37 135 135 135 0 145 145 144 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.....

i get

66 77 54 16 0 0 0 0 0 0 54 0 0 0 40 0 0 0 32 0 0 0 32 0 0 0 1 0 32 0 0 0 0 0 0 0 0 0 19 11 0 0 19 11 0 0 0 0 0 0 0 0 0 0 65533 65533 65533 0 65533 65533 65533 0 0 0 0 28 0 0 0 65 0 0 0 37 65533 65533 65533 0 65533 65533 65533 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0....

I don't understand why that occur and i cannot use server-side base64 encoding.

Here the code i am using:

		http = new XMLHttpRequest();
		http.onreadystatechange = new Function("",
		"callbacksend('"+url+"');");
		http.open("post", "http://www.myserver.com/Icon.bmp", true);
		http.setRequestHeader("Accept-Charset", "utf-8");
		http.send(null);

function callbacksend(url){
	if(http.readyState == 4){
		if(http.status == 200){
			// todo ok
			for( var i=0; i < 100; i++){
				document.body.appendChild(document.createTextNode(" " + http.responseText.charCodeAt(i)));
			}			
		}
	}
};

Thx for any help and sorry for my poor english

Escain

[Jeff Mott]

It's probably UTF-8 that's doing it. Try ISO-8859-1 instead.

[Escain]

thanks for reply, i tried it but no changes :S

I also uploaded all the test to my server if anyone wants to try it with complete source code:
http://www.leyendaepica.es/include2.html

With this example, i get the following result:

Size: 306
66 77 50 1 0 0 0 0 0 0 54 0 0 0 40 
0 0 0 9 0 0 0 9 0 0 0 1 0 24 0 0 0 
0 0 65533 0 0 0 65533 14 0 0 
65533 14 0 0 0 0 0 0 0 0 0 0 
36 36 36 82 82 82 76 76 76 100 
88 100 84 80 84 76 76 76 76 76 
76 82 82 82 36 36 36 0 82 82 82 
65533 65533 65533 
65533 65533 65533
......

[Jeff Mott]

65533 (or 0xFFFD) is a replacement character, used to replace a character whose value is unknown or unrepresentable in Unicode.

But this explains how to get around that.

There are two pieces of magic. The first is:

http.overrideMimeType('text/plain; charset=x-user-defined');

Which tells the browser: Don't treat it as characters. Don't treat it as anything. Just pass it through.

Now you should be seeing different, but still large, numbers. It seems that an 0xF7 byte will be in the high-order position. I couldn't find any explanation why this happens, but nonetheless we can remove it with a second piece of magic:

http.responseText.charCodeAt(i) & 0xff

And that'll do it. :-)

[Jeff Mott]

I just discovered that this solution doesn't work in IE. I'm checking into it.

[Jeff Mott]

Looks like you need to use VBScript if it's ever going to work in IE. It gets pretty ugly. But fortunately someone wrapped the ugliness in a library.

You can read the source if you're curious what's going on under the hood. Otherwise, just go ahead and load the library into your page and use its API.

<script type="text/javascript" src="http://www.heypage.com/nagoon97/BinFileReader/BinFileReader.js"></script>
<script type="text/javascript">
var bmpFile = new BinFileReader("Icon.bmp");

document.body.appendChild(document.createTextNode("Size: " + bmpFile.getFileSize()));
document.body.appendChild(document.createElement("br"));
for (var i = 0; i < bmpFile.getFileSize(); i++) {
    document.body.appendChild(document.createTextNode(" " + bmpFile.readString(1, i).charCodeAt(0)));
}
</script>

[Escain]

Thanks a lot its really a nice job

I have tried and it run well. I also discovered that it's possible to read directly in the correct charset without the conversion byte to byte ( & 0xff ) changing only that line:

http.overrideMimeType('text/plain; charset=iso-8859-1');

thanks again