From 128321bca98a63ad89fa1080416f3dc5d37fbae4 Mon Sep 17 00:00:00 2001
From: happysalada <raphael@megzari.com>
Date: Mon, 28 Aug 2023 16:45:49 +0800
Subject: [PATCH] unstructured-api: init at 0.0.39

---
 pkgs/servers/unstructured-api/default.nix | 68 +++++++++++++++++++++++
 pkgs/top-level/all-packages.nix           |  2 +
 2 files changed, 70 insertions(+)
 create mode 100644 pkgs/servers/unstructured-api/default.nix

diff --git a/pkgs/servers/unstructured-api/default.nix b/pkgs/servers/unstructured-api/default.nix
new file mode 100644
index 00000000000..521ff17e573
--- /dev/null
+++ b/pkgs/servers/unstructured-api/default.nix
@@ -0,0 +1,68 @@
+{
+  lib,
+  stdenvNoCC,
+  fetchFromGitHub,
+  python3,
+  makeWrapper,
+  nix-update-script,
+  symlinkJoin,
+  nltk-data,
+}:
+let
+  pythonEnv = python3.withPackages (packages: with packages; [
+    unstructured-api-tools
+    unstructured
+    pydantic
+    click
+    ratelimit
+    requests
+    pypdf
+    pycryptodome
+    safetensors
+    uvicorn
+  ] ++ packages.unstructured.optional-dependencies.local-inference);
+  version = "0.0.39";
+  unstructured_api_nltk_data = symlinkJoin {
+    name = "unstructured_api_nltk_data";
+
+    paths = [ nltk-data.punkt nltk-data.averaged_perceptron_tagger ];
+  };
+in stdenvNoCC.mkDerivation {
+  pname = "unstructured-api";
+  inherit version;
+
+  src = fetchFromGitHub {
+    owner = "Unstructured-IO";
+    repo = "unstructured-api";
+    rev = version;
+    hash = "sha256-fk0YkGllggi0eWdp9ytHy4/9rChkcDnQvEvVAp1+RJw=";
+  };
+
+  nativeBuildInputs = [ makeWrapper ];
+
+  installPhase = ''
+    runHook preInstall
+
+    mkdir -p $out $out/bin $out/lib
+    cp -r . $out/lib
+
+    makeWrapper ${pythonEnv}/bin/uvicorn $out/bin/unstructured-api \
+      --set NLTK_DATA ${unstructured_api_nltk_data} \
+      --prefix PYTHONPATH : $out/lib \
+      --add-flags "prepline_general.api.app:app"
+
+    runHook postInstall
+  '';
+
+  passthru = {
+    updateScript = nix-update-script { };
+  };
+
+  meta = with lib; {
+    description = "open-source toolkit designed to make it easy to prepare unstructured data like PDFs, HTML and Word Documents for downstream data science tasks";
+    homepage = "https://github.com/Unstructured-IO/unstructured-api";
+    changelog = "https://github.com/Unstructured-IO/unstructured-api/releases/tag/${version}";
+    license = licenses.asl20;
+    maintainers = with maintainers; [ happysalada ];
+  };
+}
diff --git a/pkgs/top-level/all-packages.nix b/pkgs/top-level/all-packages.nix
index e00fbc33de8..a84f95f282a 100644
--- a/pkgs/top-level/all-packages.nix
+++ b/pkgs/top-level/all-packages.nix
@@ -27486,6 +27486,8 @@ with pkgs;
     inherit (darwin.apple_sdk.frameworks) Cocoa WebKit;
   };
 
+  unstructured-api = callPackage ../servers/unstructured-api { };
+
   urserver = callPackage ../servers/urserver { };
 
   uxplay = callPackage ../servers/uxplay { };