International strings/chars

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

Dobieslaw Wroblewski - 31 Dec 2004 05:16 GMT

Hello,

Please help me. I think I found this once but cannot find again :-(

Is there any .NET framework class/method that returns the international
counterpart of a given string or character? For example - for '?' it returns
'o'.

DW.

Reply to this Message

Morten Wennevik - 31 Dec 2004 11:23 GMT

Hi Dobieslaw,

There isn't an easy way to do this, but this hack seem to work for most
characters.

string s = "áàäãâåéèëêíìïîóòöõôøúùüûýÿ";
byte[] b = Encoding.GetEncoding(1251).GetBytes(s);
string t = Encoding.ASCII.GetString(b);

t == aaaaaaeeeeiiiioooooouuuuyy

Signature

Happy Coding!
Morten Wennevik [C# MVP]

Reply to this Message

Dobieslaw Wroblewski - 01 Jan 2005 11:24 GMT

> string s = "??????????????????????????";
> byte[] b = Encoding.GetEncoding(1251).GetBytes(s);
> string t = Encoding.ASCII.GetString(b);

Thanks, it works!

I believe I used Encoding.ASCII for writing to text files, but in such cases
I got question marks instead of the national characters - so this trick is
quite mysterious for me, but - goodie - it works :-).

DW.

Reply to this Message

Jon Skeet [C# MVP] - 01 Jan 2005 17:22 GMT

Dobieslaw Wroblewski
<dobieslaw.wroblewski@reverse_the_next.com.cegedim> wrote:

> > string s = "áàäãâåéèëêíìïîóòöõô=F8úùüûý?";
> > byte[] b = Encoding.GetEncoding(1251).GetBytes(s);
[quoted text clipped - 4 lines]
> I believe I used Encoding.ASCII for writing to text files, but in
> such cases I got question marks instead of the national characters -

Yes, you would - ASCII doesn't contain any accented characters.

> so this trick is
> quite mysterious for me, but - goodie - it works :-).

Unfortunately, the above isn't really guaranteed to work. The Encoding
class doesn't specify any behaviour for when GetString is presented
with bytes which aren't a valid encoded form for that encoding. It may
work now, but there's no guarantee it'll work in the future at all.

I suspect you want something like "normalizatino form D" of the Unicode
string, followed by a removal of all accents etc - but I don't know
enough about normalization to be sure, and I don't know of any
implementations of that for .NET :(

If you know that all the characters you care about are within a certain
range, I'd suggest a hand-crafted mapping.

Signature

Jon Skeet - <skeet@pobox.com>
http://www.pobox.com/~skeet
If replying to the group, please do not mail me too

Reply to this Message

Dobieslaw Wroblewski - 02 Jan 2005 01:36 GMT

> If you know that all the characters you care about are within a certain
> range, I'd suggest a hand-crafted mapping.

Well, that always could be done ;-). I just thought of something smarter.
And I think I saw some function for that, but now I cannot remember for sure
if this was .NET, Win32 API or... maybe even Java ;-).

DW.

Reply to this Message

Jon Skeet [C# MVP] - 02 Jan 2005 07:49 GMT

Dobieslaw Wroblewski
<dobieslaw.wroblewski@reverse_the_next.com.cegedim> wrote:

> > If you know that all the characters you care about are within a certain
> > range, I'd suggest a hand-crafted mapping.
>
> Well, that always could be done ;-). I just thought of something smarter.
> And I think I saw some function for that, but now I cannot remember for sure
> if this was .NET, Win32 API or... maybe even Java ;-).

I wouldn't be surprised if there were a call in Win32, and you may be
able to use P/Invoke to use that call. You could try asking in a
Windows internationalization newsgroup...

Signature

Jon Skeet - <skeet@pobox.com>
http://www.pobox.com/~skeet
If replying to the group, please do not mail me too

Reply to this Message

International strings/chars

Free Magazines