Front page | perl.perl5.porters |
Postings from February 2001
Simon Cozens <simon@netthink.co.uk> writes:
>On Mon, Feb 19, 2001 at 09:53:14PM -0500, Andrew Pimlott wrote:
>> Let me say first that the reason all of my pseudo-code has been "OO
>> crap" is that I'm trying to make it as painfully clear as I can
>> think to (yes, emphasis on pain). Since nobody else has yet
>> proposed any specific interfaces,
>
>What about the one we've got?
>
>> I'm saying you call an explicit function, eg to_utf8(), which gives
>> you back a string such that if you say "substr $str, 0, 1", you get
>> the first byte of the UTF-8 representation, and "length $str" is the
>> length of the UTF-8 representation. Period.
>
>I wonder if this would work:
>
> sub to_utf8 {
> use bytes
> return $_[0]
> }
That only works of $_[0] is already UTF-8 encoded.
use Encode qw(utf8_encode);
sub to_utf8
{
my $copy = $_[0];
utf8_encode($copy); # upgrade then, turn SvUTF8_off
return $copy;
}
Does work. Personally I would rather the functions in Encode returned modified
value rather than munging-in-place.
The documented interface might be
use Encode qw(from_to);
sub to_utf8
{
my $copy = $_[0];
my $length = from_to($copy,'Unicode','UTF-8');
return $copy;
}
It is unclear what value there is to the _perl_ API in returning the length.
--
Nick Ing-Simmons <nik@tiuk.ti.com>
Via, but not speaking for: Texas Instruments Ltd.