[Fribidi-discuss] type of fribidi_iso8859_8_to_unicode changed in CVS

[Tzafrir Cohen]

On Tue, 1 Jan 2002, Behdad Esfahbod wrote:

> Hi,
>
> As you may know, we didn't assume that FriBidiChar* strings has been
> terminated with a U+0000, and ask for length in all functions, then I
> found it better to do the same assumption on char* strings, its a bit
> weird, but makes things more clean, as all the fribidi_*_to_unicode
> functions used to have a line 'int len = strlen(s);', then I decided
> to ask for len in input, It has the advantage that you can pass a part
> of string to converter, I used to terminate my strings in middle, and
> then restore the poor charater back and continue....
>
> As we are still below release 1.x.y, I let myself to change the
> interface a little, also added a FRIBIDI_INTERFACE_VERSION macro,
> which will be increased with each interface change.
>
> If you cannot pass a strlen(s) yourself, I can define some macros to
> do this for you, or maybe you're right, I should rename the new
> interface, define the old ones as before, let me know.
>

I know the length at that point. My only problem is writing a code that
will work with both versions. I suppose I can write there a macro that
will ignore ", length" for versions older than 0.10.0 in that call. But it
looks ugly...

>
> Also I encourage you to use
> 'fribidi_charset_to_unicode(FRIBIDI_CHARSET_ISO8859_8, ...) instead.
>
> behdad
>
> On Tue, 1 Jan 2002, Tzafrir Cohen wrote:
>
> > Hi
> >
> > I noticed that fribidi_iso8859_8_to_unicode of 0.10pre has a "length"
> > parameter, unlike all previous versions (up to 0.9.1).
> >
> > Why is that?
> >
> > (It breaks my program)
> >
> >
>
>

-- 
Tzafrir Cohen
mailto:tzafrir at technion.ac.il
http://www.technion.ac.il/~tzafrir