Hi guys, I have a problem regarding a `cinterop` ...

# kotlin-native

[Vlad Balan]

Hi guys,

I have a problem regarding a cinterop library and the architectures: arm32 and x64 .

I am using the Cairo library as cinterop library for both targets. After the cinterop process the following struct:

typedef struct {
    unsigned long        index;
    double               x;
    double               y;
} cairo_glyph_t;

is mapped for arm32 to:

@kotlinx.cinterop.internal.CStruct public final class cairo_glyph_t public constructor(rawPtr: kotlinx.cinterop.NativePtr /* = kotlin.native.internal.NativePtr */) : kotlinx.cinterop.CStructVar {
    public companion object : kotlinx.cinterop.CStructVar.Type {
    }

    public final var index: kotlin.UInt /* compiled code */

    public final var x: kotlin.Double /* compiled code */

    public final var y: kotlin.Double /* compiled code */
}

and for x64 to:

@kotlinx.cinterop.internal.CStruct public final class cairo_glyph_t public constructor(rawPtr: kotlinx.cinterop.NativePtr /* = kotlin.native.internal.NativePtr */) : kotlinx.cinterop.CStructVar {
    public companion object : kotlinx.cinterop.CStructVar.Type {
    }

    public final var index: kotlin.ULong /* compiled code */

    public final var x: kotlin.Double /* compiled code */

    public final var y: kotlin.Double /* compiled code */
}

As you can see the index field has different types.
I am trying to write the code once and compile it for both platforms, but the following example does not compile:

val glyph = cValue<cairo_glyph_t> {
  index =  0
  x = 0.0
  y = 0.0
}

Does anybody have an idea how I can overcome this issue?

Thanks in advance!

[Dominaezzz]

This can be overcome by using index = 0.convert() . It basically gives you implicit casting.

[Vlad Balan]

Great! Thanks!