Hello Mike,
Yeah, I would agree about the impact of having utf everywhere, but
majority of our products end up getting translated. And I would say
about 95% of the strings end up holding text in different languages.
As a result, I would rather eat the performance hit than have to recode
the chunk of the code or have programmers use specialize functions for
something as simple as strings.
So, I assume there is no simple option or tweak that I could do to turn
all strings to utf right?
On 3/8/2010 2:26 AM, Michael Pliskin wrote:
Hi,
There are a bunch of utf8-related functions in std.ndll, so you can deal with
utf8 strings quite easily. Changing strings from raw bytes to utf8 will have
performance implications I think, that's why it is not done..
Mike
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Serguei Kloubkov
Sent: Sunday, March 07, 2010 7:34 PM
To: [email protected]
Subject: [Neko] UTF-8 default support for String
Hello
Just figured out how to embed the VM and do all sorts of nice things, and
everything is working as expected.
I was however shocked to discover that it doesn't support UTF strings by
default. As a result all the string related function do not return expected
values (length etc).
What would it take to modify the VM to use wchar_t or wstrings and what impact
would it have on the functions that work with strings?
Thanks
Sergey
--
Neko : One VM to run them all
(http://nekovm.org)
--
Neko : One VM to run them all
(http://nekovm.org)
