depot/third_party/nixpkgs/pkgs/development/python-modules/pkuseg/default.nix
Default email 14910f5943 Project import generated by Copybara.
GitOrigin-RevId: 5aaed40d22f0d9376330b6fa413223435ad6fee5
2022-01-13 15:06:32 -05:00

36 lines
755 B
Nix

{ lib
, buildPythonPackage
, fetchPypi
, isPy3k
, pythonAtLeast
, cython
, numpy
}:
buildPythonPackage rec {
pname = "pkuseg";
version = "0.0.25";
disabled = !isPy3k || pythonAtLeast "3.9";
src = fetchPypi {
inherit pname version;
sha256 = "148yp0l7h8cflxag62pc1iwj5b5liyljnaxwfjaiqwl96vwjn0fx";
};
# Does not seem to have actual tests, but unittest discover
# recognizes some non-tests as tests and fails.
doCheck = false;
nativeBuildInputs = [ cython ];
propagatedBuildInputs = [ numpy ];
pythonImportsCheck = [ "pkuseg" ];
meta = with lib; {
description = "Toolkit for multi-domain Chinese word segmentation";
homepage = "https://github.com/lancopku/pkuseg-python";
license = licenses.unfree;
};
}