Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

WIN32OLE doesn't seem to support UTF-8.

10 views
Skip to first unread message

Bret Pettichord

unread,
Sep 8, 2005, 1:03:06 AM9/8/05
to
Has anyone been able to use WIN32OLE with international characters?

I would like to access international text from Excel and Explorer using
WIN32OLE. But whenever i do this in Ruby, i get "????????". (This code
works fine for English).

data = $sheet.cells(1, 'b').value
$sheet.cells(2, 'b').value = data

I'm pretty sure this is a Ruby problem, because the equivalent script in
VBA works fine.

Set From = Worksheets("Sheet1").Cells(1, "b")
Set Target = Worksheets("Sheet1").Cells(2, "b")
Target.Value = From.Value

I've been testing this with UTF-8 encoded Japanese text, although i am
hoping to find a general solution (not specific to any code-page).

When i looked through the Ruby code (win32ole.c) and it seems that string
return values are being passed through this function:

static LPWSTR
ole_mb2wc(pm, len)
char *pm;
int len;
{
int size;
LPWSTR pw;
size = MultiByteToWideChar(CP_ACP, 0, pm, len, NULL, 0);
pw = SysAllocStringLen(NULL, size - 1);
MultiByteToWideChar(CP_ACP, 0, pm, len, pw, size);
return pw;
}

"The MultiByteToWideChar function maps a character string to a
wide-character (Unicode) string. The character string mapped by this
function is not necessarily from a multibyte character."

I don't really understand what is happening here. It seems to me that both
the input and output strings have to have their encoding specified here.
Here ANSI is hard-coded (CP_ACP) but i don't know if this is for input or
output.

http://msdn.microsoft.com/library/default.asp?url=/library/en-us/intl/unicode_17si.asp

I'd appreciate any insight into this matter that you might have. I've spent
a couple days reading up on UTF-8 and Unicode, but don't really have any
experience with these things.

I am using 1.8.2-15

Bret


_____________________
Bret Pettichord
www.pettichord.com

nobu....@softhome.net

unread,
Sep 8, 2005, 2:09:13 AM9/8/05
to
Hi,

At Thu, 8 Sep 2005 14:03:06 +0900,
Bret Pettichord wrote in [ruby-talk:155298]:


> I'd appreciate any insight into this matter that you might have. I've spent
> a couple days reading up on UTF-8 and Unicode, but don't really have any
> experience with these things.

There is WIN32OLE.codepage attribute in CVS, so it will be
available in 1.8.3.

--
Nobu Nakada


Bret Pettichord

unread,
Sep 8, 2005, 12:29:15 PM9/8/05
to
At 01:09 AM 9/8/2005, nobu....@softhome.net wrote:
>There is WIN32OLE.codepage attribute in CVS, so it will be
>available in 1.8.3.

Nobu,

AWESOME! I just downloaded 1.8.3, ran my test file with the new code_page
attribute, and verified that it works as expected. Very good news.

You should know that the Watir users have been really happy with the
WIN32OLE library.

On a separate note, I have a developer who's created a DLL for me that
converts a window handle to a pointer to an IHTMLDocument2 object. I've
wrapped this in WIN32API, but that isn't quite good enough. What i want in
Ruby is the ability to create a wrapper that returns a WIN32OLE object,
rather than just a string with a pointer address in it. Do you know of any
way to convert the pointer to a WIN32OLE object?

0 new messages