From 1eca6fb65ed0431e3706ebac231361cb24d47888 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Robert=20Sch=C3=BCtz?= <mail@dotlambda.de>
Date: Wed, 11 Dec 2024 16:02:23 -0800
Subject: [PATCH 1/2] python312Packages.type-infer: unpin py3langid

---
 .../development/python-modules/type-infer/default.nix | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/pkgs/development/python-modules/type-infer/default.nix b/pkgs/development/python-modules/type-infer/default.nix
index 2d9022520150..294769e7c3fb 100644
--- a/pkgs/development/python-modules/type-infer/default.nix
+++ b/pkgs/development/python-modules/type-infer/default.nix
@@ -30,7 +30,7 @@ in
 buildPythonPackage rec {
   pname = "type-infer";
   version = "0.0.20";
-  format = "pyproject";
+  pyproject = true;
 
   disabled = pythonOlder "3.8";
 
@@ -41,11 +41,14 @@ buildPythonPackage rec {
     hash = "sha256-F+gfA7ofrbMEE5SrVt9H3s2mZKQLyr6roNUmL4EMJbI=";
   };
 
-  pythonRelaxDeps = [ "psutil" ];
+  pythonRelaxDeps = [
+    "psutil"
+    "py3langid"
+  ];
 
-  nativeBuildInputs = [ poetry-core ];
+  build-system = [ poetry-core ];
 
-  propagatedBuildInputs = [
+  dependencies = [
     colorlog
     dataclasses-json
     langid

From b5abcd6832a951da51e6f8417687fa51072a27e0 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Robert=20Sch=C3=BCtz?= <mail@dotlambda.de>
Date: Wed, 11 Dec 2024 16:26:35 -0800
Subject: [PATCH 2/2] python312Packages.type-infer: run tests

---
 .../python-modules/type-infer/default.nix     | 34 ++++++++++++-------
 1 file changed, 22 insertions(+), 12 deletions(-)

diff --git a/pkgs/development/python-modules/type-infer/default.nix b/pkgs/development/python-modules/type-infer/default.nix
index 294769e7c3fb..b4e544a607ae 100644
--- a/pkgs/development/python-modules/type-infer/default.nix
+++ b/pkgs/development/python-modules/type-infer/default.nix
@@ -2,16 +2,16 @@
   lib,
   buildPythonPackage,
   pythonOlder,
-  fetchPypi,
+  fetchFromGitHub,
   poetry-core,
   colorlog,
   dataclasses-json,
-  langid,
   nltk,
   numpy,
   pandas,
   psutil,
   py3langid,
+  pytestCheckHook,
   python-dateutil,
   scipy,
   toml,
@@ -23,22 +23,26 @@ let
     name = "nltk-test-data";
     paths = [
       nltk-data.punkt
+      nltk-data.punkt_tab
       nltk-data.stopwords
     ];
   };
+
+  version = "0.0.20";
+  tag = "v${version}";
 in
 buildPythonPackage rec {
   pname = "type-infer";
-  version = "0.0.20";
+  inherit version;
   pyproject = true;
 
   disabled = pythonOlder "3.8";
 
-  # using PyPI because the repo does not have tags or release branches
-  src = fetchPypi {
-    pname = "type_infer";
-    inherit version;
-    hash = "sha256-F+gfA7ofrbMEE5SrVt9H3s2mZKQLyr6roNUmL4EMJbI=";
+  src = fetchFromGitHub {
+    owner = "mindsdb";
+    repo = "type_infer";
+    inherit tag;
+    hash = "sha256-2Y+NPwUnQMj0oXoCMfUOG40lqduy9GTcqxfyuFDOkHc=";
   };
 
   pythonRelaxDeps = [
@@ -51,7 +55,6 @@ buildPythonPackage rec {
   dependencies = [
     colorlog
     dataclasses-json
-    langid
     nltk
     numpy
     pandas
@@ -62,8 +65,14 @@ buildPythonPackage rec {
     toml
   ];
 
-  # PyPI package does not include tests
-  doCheck = false;
+  nativeCheckInputs = [
+    pytestCheckHook
+  ];
+
+  disabledTests = [
+    # test hangs
+    "test_1_stack_overflow_survey"
+  ];
 
   # Package import requires NLTK data to be downloaded
   # It is the only way to set NLTK_DATA environment variable,
@@ -72,8 +81,9 @@ buildPythonPackage rec {
   pythonImportsCheck = [ "type_infer" ];
 
   meta = with lib; {
+    changelog = "https://github.com/mindsdb/type_infer/releases/tag/${tag}";
     description = "Automated type inference for Machine Learning pipelines";
-    homepage = "https://pypi.org/project/type-infer/";
+    homepage = "https://github.com/mindsdb/type_infer";
     license = licenses.gpl3Only;
     maintainers = with maintainers; [ mbalatsko ];
   };