Hi!

I'm using XMLHTTP activeX object to access pages in remote servers.
It's working fine but when the contents is text with chars like "à" or "è" it returns unknown information. Actually it melts this char with the following byte and returns only one!

Maybe this has something to do with Unicode or ASCII defaults. do you know what it can be?

thx!!

Usually indicates that the server dishing out the page is not setting the encoding accurately. For example it says the page is in utf-8 when it is not, do you have any control over the source?

--

Joe

Hi,

I have all the control over the source.. it's just a remote server but I have access to the files.. The enconding I need is:

iso 8859-1 (Latin-1)

I must force the server to send it has this.. Ill try to check the HEADER msg containing the enconding it is using!

see ya later!
(do you have any ideis how to force it to send it in a determined iso?)

THX

Show the code used to create the document, a brief bit that is. What server-side technique are you using, ASP, PHP etc?

Joe (MVP - xml)

Ok.. I'm using JSCRIPT.
This is a HTA that gets info from a file and them writes it in a file. The file is a .JS (TEXT) and its enconded in iso 8859-1.

Here is the code I use to retrieve and write the file:
The commented code is the part that writes. It works but not with this enconding. I think the server is not returning the correct enconding like you said:

function GetNewData(address) {
   var objFSO, resposta;

   try {
    objFSO = new ActiveXObject("Scripting.FileSystemObject");
    resposta = new ActiveXObject("Msxml2.XMLHTTP");
    resposta.open("GET",address,false);
    resposta.send();

    document.write(resposta.getAllResponseHeaders());
    //returned contents need to be [iso 8859-1 Latin-1 / ANSI] in order to write to file

    //TextFile = objFSO.CreateTextFile(".\\files\\db\\series.js");
    //TextFile.write(resposta.responseText);
    //TextFile.Close();
   }
   catch(e){
    alert("Failure while updating the DataBase!\nDescription: "+e.description);
    return;
   }
}

And here the headers from the server:

Server: Microsoft-IIS/5.0
X-Powered-By: ASP.NET
Date: Mon, 10 Nov 2003 22:36:34 GMT
Content-Type: application/octet-stream
Accept-Ranges: bytes
Last-Modified: Mon, 10 Nov 2003 11:01:25 GMT
ETag: "10d622fc79a7c31:992"
Content-Length: 141861

Yes, but what is the code on the server that creates the response, that's where the problem lies.
Is it an internet accessible url?
--

Joe

This is supposed to be an Internet accessible address although I'm just testing it in a Intranet.

The problem is that there is no server-side code!! This is a HTTP Request just like your browser makes to the server.. IIS then returns what he is supposed to return acording to the request I made..

Here I am just requesting a page.. The server returns it the way it wants! I can't control it.. What I can't understand is why even the raw answer returned by the property responseBody(Byte-array) has the wrong bytes! If at least this property was correct I could make VBScript code to process the byte array. But even the binary answer is wrong!

if you have any clue please let me know and THX!

Well the header says "Content-Type: application/octet-stream" which is not consistent with a page encoded as iso 8859-1. Is an htm page or a binary file?


--

Joe

Hum.. Good news..

I think there is no way to control what the server sends. I mean.. This issue cannot be treated server-side.
I've been reading some stuff and it looks like microsoft like the octet-like content description for all the contents it sends as long as the software can take into consideration the file extention.

In my case I have to do the same.. I know the file is text in Latin-1 iso. It is sent has an array of bytes that is CORRECT. I thought it was a wrong byte array but it is a correct byte array.

The problem is in the responseText property that assumes the file is unicode and for chars over &H7F (the case of 'à') it messes everything up. The solutions are:

1)
Find a way to force resposeText of the XHMHTTPRequest to treat the data as Latin-1.
(Don't think it's possible)

2)
Handle the byte array. (ResponseBody property)
(it will work but cannot be done with JScript. Have to use VBScript)

3)
Process the return Stream Object. (ResponseStream property)
This can be done in JScript. I think it will work because there is a property of the Stream Object where you can select the charset iso)



I will try the Stream because will be easier to do than process binary data in VBScript. The inly problem is that they say ResponseStream property returns a Stream Object but I still can't use it.. the code returns an Object that's for sure. But I have to find out how to say that the object is a Stream Object. Do you know how to do this?

Thx!

Yes. I am sure that the file is text. This code works just fine if the file has only Standart ASCII chars (0-127).

:)