parquet-cpp: init at 1.4.0

Also fix RPATH issues in pyarrow

3 files changed, 49 insertions, 3 deletions

diff --git a/pkgs/development/libraries/parquet-cpp/default.nix b/pkgs/development/libraries/parquet-cpp/default.nix
new file mode 100644
index 000000000000..e281e604380b
--- /dev/null
+++ b/pkgs/development/libraries/parquet-cpp/default.nix
@@ -0,0 +1,37 @@
+{ stdenv, symlinkJoin, fetchurl, arrow-cpp, boost, cmake, gtest, snappy, thrift, zlib }:
+
+stdenv.mkDerivation rec {
+  name = "parquet-cpp-${version}";
+  version = "1.4.0";
+
+  src = fetchurl {
+    url = "https://github.com/apache/parquet-cpp/archive/apache-${name}.tar.gz";
+    sha256 = "1kn7pjzi5san5f05qbl8l8znqsa3f9cq9bflfr4s2jfwr7k9p2aj";
+  };
+
+  nativeBuildInputs = [ cmake ];
+  buildInputs = [ boost ];
+
+  preConfigure = ''
+    substituteInPlace cmake_modules/FindThrift.cmake --replace CMAKE_STATIC_LIBRARY CMAKE_SHARED_LIBRARY
+    substituteInPlace cmake_modules/FindSnappy.cmake --replace CMAKE_STATIC_LIBRARY CMAKE_SHARED_LIBRARY
+  '';
+
+  ARROW_HOME = arrow-cpp;
+  THRIFT_HOME = thrift;
+  GTEST_HOME = gtest;
+  SNAPPY_HOME = symlinkJoin { name="snappy-wrap"; paths = [ snappy snappy.dev ]; };
+  ZLIB_HOME = symlinkJoin { name="zlib-wrap"; paths = [ zlib.dev zlib.static ]; };
+
+  cmakeFlags = [
+    "-DPARQUET_BUILD_BENCHMARKS=OFF"
+  ];
+
+  meta = {
+    description = "A C++ library to read and write the Apache Parquet columnar data format";
+    homepage = http://parquet.apache.org;
+    license = stdenv.lib.licenses.asl20;
+    platforms = stdenv.lib.platforms.unix;
+    maintainers = with stdenv.lib.maintainers; [ veprbl ];
+  };
+}
diff --git a/pkgs/development/python-modules/pyarrow/default.nix b/pkgs/development/python-modules/pyarrow/default.nix
index 8804953045e4..11c2868c3370 100644
--- a/pkgs/development/python-modules/pyarrow/default.nix
+++ b/pkgs/development/python-modules/pyarrow/default.nix
@@ -1,4 +1,4 @@
-{ lib, buildPythonPackage, fetchurl, arrow-cpp, cmake, cython, futures, numpy, pandas, pytest, pkgconfig, setuptools_scm, six }:
+{ lib, buildPythonPackage, fetchurl, arrow-cpp, cmake, cython, futures, numpy, pandas, pytest, parquet-cpp, pkgconfig, setuptools_scm, six }:
 
 buildPythonPackage rec {
   pname = "pyarrow";
@@ -16,10 +16,14 @@ buildPythonPackage rec {
   checkInputs = [ pandas pytest ];
 
   PYARROW_BUILD_TYPE = "release";
-  PYARROW_BUNDLE_ARROW_CPP = 1; # sets RPATH on darwin
+  PYARROW_CMAKE_OPTIONS = "-DCMAKE_INSTALL_RPATH=${ARROW_HOME}/lib;${PARQUET_HOME}/lib";
 
   preBuild = ''
-    substituteInPlace CMakeLists.txt --replace "''${ARROW_SO_VERSION}" '"0"'
+    substituteInPlace CMakeLists.txt --replace "\''${ARROW_ABI_VERSION}" '"0.0.0"'
+    substituteInPlace CMakeLists.txt --replace "\''${ARROW_SO_VERSION}" '"0"'
+
+    # fix the hardcoded value
+    substituteInPlace cmake_modules/FindParquet.cmake --replace 'set(PARQUET_ABI_VERSION "1.0.0")' 'set(PARQUET_ABI_VERSION "${parquet-cpp.version}")'
   '';
 
   preCheck = ''
@@ -36,6 +40,9 @@ buildPythonPackage rec {
   '';
 
   ARROW_HOME = arrow-cpp;
+  PARQUET_HOME = parquet-cpp;
+
+  setupPyBuildFlags = ["--with-parquet" ];
 
   meta = with lib; {
     description = "A cross-language development platform for in-memory data";
diff --git a/pkgs/top-level/all-packages.nix b/pkgs/top-level/all-packages.nix
index 9ffabd197861..a81fe88b8a85 100644
--- a/pkgs/top-level/all-packages.nix
+++ b/pkgs/top-level/all-packages.nix
@@ -10872,6 +10872,8 @@ with pkgs;
 
   paperkey = callPackage ../tools/security/paperkey { };
 
+  parquet-cpp = callPackage ../development/libraries/parquet-cpp {};
+
   pangoxsl = callPackage ../development/libraries/pangoxsl { };
 
   pcaudiolib = callPackage ../development/libraries/pcaudiolib {