Am 2017-01-17 um 12:09 schrieb rastersoft:
> That's true for the first one, but the second is as fast as str[0] == 0,
> because as soon as one byte doesn't fit, the function ends.
You are right, because str=="" indeed calls g_strcmp0(), but
str[0]=='\0' calls string_get() which seems not faster t
Hi:
El 17/01/17 a las 10:24, Ulink escribió:
> But I think str.length==0 and str=="" are no high performance solutions,
> especially for big/long strings.
That's true for the first one, but the second is as fast as str[0] == 0,
because as soon as one byte doesn't fit, the function ends.
--
Nos l
As Guillaume wrote, you have to use string? str and check string==null
FIRST.
But I think str.length==0 and str=="" are no high performance solutions,
especially for big/long strings.
Maybe this is better?
bool is_null_or_empty (string? str)
{
return str == null || str[0] == '\0';
}
My mistake. Thanks for correcting me.
17.01.2017, 10:55, "Jens Georg" :
>> In Vala you can simply just check for str != "", this is enough.
>
> No, if str is nullable it's semantically different, because str != ""
> will be generated as
>
> if (g_strcmp0 (_tmp1_, "") != 0) {
> ..
> }
>
> This
In Vala you can simply just check for str != "", this is enough.
No, if str is nullable it's semantically different, because str != ""
will be generated as
if (g_strcmp0 (_tmp1_, "") != 0) {
..
}
This will be TRUE for _tmp1_ being NULL because NULL isn't "" ( and
g_strcmp0 will return -1
