python312Packages.markitdown: 0-unstable-2024-12-15

pkgs/development/python-modules/markitdown/default.nix

@@ -0,0 +1,65 @@
+{
+  lib,
+  buildPythonPackage,
+  fetchFromGitHub,
+  hatchling,
+  beautifulsoup4,
+  mammoth,
+  markdownify,
+  numpy,
+  openpyxl,
+  pandas,
+  pathvalidate,
+  pdfminer-six,
+  puremagic,
+  pydub,
+  python-pptx,
+  requests,
+  speechrecognition,
+  youtube-transcript-api,
+}:
+
+buildPythonPackage rec {
+  pname = "markitdown";
+  version = "unstable-2024-12-15";
+  pyproject = true;
+
+  src = fetchFromGitHub {
+    owner = "microsoft";
+    repo = "markitdown";
+    rev = "81e3f24acd0049a59cd2dcb2d01d0a98cc57c734";
+    hash = "sha256-ejj7ARvLDzB1WRNkc1zUtzYAujZVysB0MkY/PQKv/nQ=";
+  };
+
+  build-system = [
+    hatchling
+  ];
+
+  dependencies = [
+    beautifulsoup4
+    mammoth
+    markdownify
+    numpy
+    openpyxl
+    pandas
+    pathvalidate
+    pdfminer-six
+    puremagic
+    pydub
+    python-pptx
+    requests
+    speechrecognition
+    youtube-transcript-api
+  ];
+
+  pythonImportsCheck = [
+    "markitdown"
+  ];
+
+  meta = {
+    description = "Python tool for converting files and office documents to Markdown";
+    homepage = "https://github.com/microsoft/markitdown";
+    license = lib.licenses.mit;
+    maintainers = with lib.maintainers; [ ];
+  };
+}

pkgs/top-level/python-packages.nix

@@ -7851,6 +7851,8 @@ self: super: with self; {
 
   markdownify  = callPackage ../development/python-modules/markdownify { };
 
+  markitdown  = callPackage ../development/python-modules/markitdown { };
+
   marko = callPackage ../development/python-modules/marko { };
 
   markupsafe = callPackage ../development/python-modules/markupsafe { };