apache-arrow: Enable Parquet/Dataset Module, add buildlink3 for use by the python module

[Matthew Danielson <matthewd%fastmail.us@localhost>]

Module Name:	pkgsrc-wip
Committed By:	Matthew Danielson <matthewd%fastmail.us@localhost>
Pushed By:	matthewd
Date:		Wed Apr 27 08:25:14 2022 -0700
Changeset:	9dff7764540849d636c58f703c1c9174085c9203

Modified Files:
	apache-arrow/Makefile
	apache-arrow/options.mk
Added Files:
	apache-arrow/buildlink3.mk

Log Message:
apache-arrow: Enable Parquet/Dataset Module, add buildlink3 for use by
the python module

To see a diff of this commit:
https://wip.pkgsrc.org/cgi-bin/gitweb.cgi?p=pkgsrc-wip.git;a=commitdiff;h=9dff7764540849d636c58f703c1c9174085c9203

Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.

diffstat:
 apache-arrow/Makefile      |  4 +++-
 apache-arrow/buildlink3.mk | 13 +++++++++++++
 apache-arrow/options.mk    |  5 +++++
 3 files changed, 21 insertions(+), 1 deletion(-)

diffs:
diff --git a/apache-arrow/Makefile b/apache-arrow/Makefile
index 7e926abd44..c773943bef 100644
--- a/apache-arrow/Makefile
+++ b/apache-arrow/Makefile
@@ -32,11 +32,13 @@ USE_TOOLS=	cmake gmake
 USE_CMAKE=	yes
 CMAKE_ARGS+=	-DCMAKE_INSTALL_LIBDIR=lib
 CMAKE_ARGS+=	-DARROW_DEPENDENCY_SOURCE=SYSTEM
-# Not yet
+# Not yet for Gandiva
 CMAKE_ARGS+=	-DARROW_GANDIVA=OFF
+# These seem OK
 CMAKE_ARGS+=	-DARROW_FILESYSTEM=ON
 CMAKE_ARGS+=	-DARROW_BUILD_UTILITIES=ON
 CMAKE_ARGS+=	-DARROW_CSV=ON
+CMAKE_ARGS+=	-DARROW_DATASET=ON
 
 # Set environment variable to find the extra source packages
 CONFIGURE_ENV+=	ARROW_JEMALLOC_URL=/${DISTDIR}/${JEMALLOC}
diff --git a/apache-arrow/buildlink3.mk b/apache-arrow/buildlink3.mk
new file mode 100644
index 0000000000..7ea74f162a
--- /dev/null
+++ b/apache-arrow/buildlink3.mk
@@ -0,0 +1,13 @@
+# $NetBSD: buildlink3.mk,v 1.25 2013/12/04 14:14:33 obache Exp $
+
+BUILDLINK_TREE+=	apache-arrow
+
+.if !defined(APACHE_ARROW_BUILDLINK3_MK)
+APACHE_ARROW_BUILDLINK3_MK:=
+
+BUILDLINK_API_DEPENDS.apache-arrow+=	apache-arrow>=7.0.0
+BUILDLINK_ABI_DEPENDS.apache-arrow+=	apache-arrow>=7.0.0
+BUILDLINK_PKGSRCDIR.apache-arrow?=	../../wip/apache-arrow
+.endif # APACHE_ARROW_BUILDLINK3_MK
+
+BUILDLINK_TREE+=	-apache-arrow
diff --git a/apache-arrow/options.mk b/apache-arrow/options.mk
index dc16c8d70e..801d0cae7c 100644
--- a/apache-arrow/options.mk
+++ b/apache-arrow/options.mk
@@ -6,6 +6,9 @@ PKG_SUGGESTED_OPTIONS=	brotli bzip2 google-log json lz4 parquet python snappy zl
 
 .include "../../mk/bsd.options.mk"
 
+# The CMAKE_ARGS are for the C++ libraries
+# The MAKE_ENV are for the python bindings
+
 .if !empty(PKG_OPTIONS:Mbrotli)
 .include "../../archivers/brotli/buildlink3.mk"
 CMAKE_ARGS+=	-DARROW_WITH_BROTLI=ON
@@ -33,6 +36,7 @@ CMAKE_ARGS+=	-DARROW_WITH_LZ4=ON
 
 .if !empty(PKG_OPTIONS:Mparquet)
 CMAKE_ARGS+=	-DARROW_PARQUET=ON
+MAKE_ENV+=	PYARROW_WITH_PARQUET=1
 .endif
 
 .if !empty(PKG_OPTIONS:Mpython)
@@ -40,6 +44,7 @@ CMAKE_ARGS+=	-DARROW_PARQUET=ON
 .include "../../lang/python/pyversion.mk"
 CMAKE_ARGS+=	-DARROW_PYTHON=ON
 .endif
+
 .if !empty(PKG_OPTIONS:Msnappy)
 .include "../../devel/snappy/buildlink3.mk"
 CMAKE_ARGS+=	-DARROW_WITH_SNAPPY=ON