[SWIG] add size_t manipulation functions

[shuttie]

I'm a maintainer of the lightgbm4j library, and would like to expose the LGBM_DatasetGetFeatureNames function in a more java-friendly way.

The problem with the LGBM_DatasetGetFeatureNames function is that it requires a size_t type param, which is not possible to instantiate at all, as SWIG definition file only exposes int/long/double/float/int64_t/int32_t types.

This one-liner PR also exposes a size_t function family in SWIG wrapper, so lightgbmlibJNI.java will also include these new ones:

>   public final static native long new_size_tp();
>   public final static native long copy_size_tp(long jarg1);
>   public final static native void delete_size_tp(long jarg1);
>   public final static native void size_tp_assign(long jarg1, long jarg2);
>   public final static native long size_tp_value(long jarg1);

[jameslamb]

Thanks for contributing!

@AlbertoEAF or @imatiach-msft , could you comment on this?

[AlbertoEAF]

Hello @shuttie :)

Thanks for contributing.There are already several LightGBM Java providers out there which are using such functions in Java through a int64_t when size_t pointers are needed. However, I agree that is more correct to use a proper size_t pointer.

LGTM ;)

[shuttie]

Using int64_t as a size_t can break code in a case if someone will try to build this library on 32-bit systems (like an Android phone not on Aarch64): AFAIK size_t is defined as unsigned int on 32-bit ARM.

So I see two options here:

• force SWIG to mess with datatypes in a way that int64_t will be casted to size_t, which may break on 32-bit
• change the LGBM_DatasetGetFeatureNames to use platform-independent types for their arguments, which will break API compatibility
• allow SWIG to correctly use the size_t type, so it will both work on 32bit and preserve API compat at the same time.

[AlbertoEAF]

Sure shuttie, I'm not defending using int64_t, that's why I said:

I agree that is more correct to use a proper size_t pointer.

LGTM ;)

[imatiach-msft]

LGTM!

[StrikerRUS]

Merging based on @AlbertoEAF's and @imatiach-msft's approvals. Thank you both guys!
@shuttie Thanks a lot got your contribution!

[github-actions]

This pull request has been automatically locked since there has not been any recent activity since it was closed. To start a new related discussion, open a new issue at https://github.com/microsoft/LightGBM/issues including a reference to this.