Re: [PATCH] make_unicode: Update unicode collation to version 10.0.0.

[Alexandre Julliard]

Jens Reyer <jre.winesim(a)gmail.com> writes:
> @@ -757,13 +757,12 @@ sub READ_SORTKEYS_FILE()
>      }
>      close $INPUT;
>
> -    # compress the keys to 32 bit:
> -    # key 1 to 16 bits, key 2 to 8 bits, key 3 to 4 bits, key 4 to 1 bit
> +    # compress the keys to 16 bit:
> +    # key 1 to 8 bits, key 2 to 4 bits, key 3 to 2 bits

There's no matching code change, why are you changing this comment?

-- 
Alexandre Julliard
julliard(a)winehq.org

[PATCH] make_unicode: Update unicode collation to version 10.0.0.

[Jens Reyer]

On 11/08/2017 09:02 PM, Alexandre Julliard wrote:
> Jens Reyer <jre.winesim(a)gmail.com> writes:
> > @@ -757,13 +757,12 @@ sub READ_SORTKEYS_FILE()
> >      }
> >      close $INPUT;
> >
> > -    # compress the keys to 32 bit:
> > -    # key 1 to 16 bits, key 2 to 8 bits, key 3 to 4 bits, key 4 to 1 bit
> > +    # compress the keys to 16 bit:
> > +    # key 1 to 8 bits, key 2 to 4 bits, key 3 to 2 bits
>
> There's no matching code change, why are you changing this comment?

Thanks for looking into this!

Because I assumed/hoped that reducing the input from 4 to 3 values
automatically results in 16-bit.  But given your comment it seems I was
wrong.

In the UCA documentation I found "Briefly stated, the Unicode Collation
Algorithm takes an input Unicode string and a Collation Element Table,
containing mapping data for characters. It produces a sort key, which is
an array of unsigned 16-bit integers."  So I guess in the end we also
want to have 16-bit and should change the code.

I might give it another try, but I just lack the basic knowledge here so
it's really just "Let's try to remove some lines/change some values and
check if it still works".  Therefore help or someone else doing it,
would be very welcome, sorry.

Or should we just go as it is? At least make_unicode didn't complain.

Greets
jre