1 files changed, 71 insertions, 0 deletions
diff --git a/pkgs/development/python-modules/type-infer/default.nix b/pkgs/development/python-modules/type-infer/default.nix
new file mode 100644
index 00000000000..7adbd97059c
--- /dev/null
+++ b/pkgs/development/python-modules/type-infer/default.nix
@@ -0,0 +1,71 @@
+{ lib
+, buildPythonPackage
+, pythonOlder
+, fetchPypi
+, poetry-core
+, colorlog
+, dataclasses-json
+, langid
+, nltk
+, numpy
+, pandas
+, psutil
+, python-dateutil
+, scipy
+, toml
+, nltk-data
+, symlinkJoin
+}:
+let
+  testNltkData = symlinkJoin {
+    name = "nltk-test-data";
+    paths = [ nltk-data.punkt nltk-data.stopwords ];
+  };
+in
+buildPythonPackage rec {
+  pname = "type-infer";
+  version = "0.0.15";
+  format = "pyproject";
+
+  disable = pythonOlder "3.8";
+
+  # using PyPI because the repo does not have tags or release branches
+  src = fetchPypi {
+    pname = "type_infer";
+    inherit version;
+    hash = "sha256-AnThYE6hHc3Pwu8fl0VBiQJfGVjeEKo4RrCsOl2pfCA=";
+  };
+
+  nativeBuildInputs = [
+    poetry-core
+  ];
+
+  propagatedBuildInputs = [
+    colorlog
+    dataclasses-json
+    langid
+    nltk
+    numpy
+    pandas
+    psutil
+    python-dateutil
+    scipy
+    toml
+  ];
+
+  # PyPI package does not include tests
+  doCheck = false;
+
+  # Package import requires NLTK data to be downloaded
+  # It is the only way to set NLTK_DATA environment variable,
+  # so that it is available in pythonImportsCheck
+  env.NLTK_DATA = testNltkData;
+  pythonImportsCheck = [ "type_infer" ];
+
+  meta = with lib; {
+    description = "Automated type inference for Machine Learning pipelines";
+    homepage = "https://pypi.org/project/type-infer/";
+    license = licenses.gpl3Only;
+    maintainers = with maintainers; [ mbalatsko ];
+  };
+}