From fa7fbe565ded463f2864f128e8d467d0fd53bdff Mon Sep 17 00:00:00 2001
From: Martin Puppe <dev@mpuppe.de>
Date: Thu, 9 Feb 2023 22:36:39 +0100
Subject: [PATCH 1/3] nixos/paperless: apply nixpkgs-fmt

---
 nixos/modules/services/misc/paperless.nix | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)

diff --git a/nixos/modules/services/misc/paperless.nix b/nixos/modules/services/misc/paperless.nix
index 667f16d98f82..71ba8fd7075f 100644
--- a/nixos/modules/services/misc/paperless.nix
+++ b/nixos/modules/services/misc/paperless.nix
@@ -24,12 +24,14 @@ let
     lib.mapAttrs (_: toString) cfg.extraConfig
   );
 
-  manage = let
-    setupEnv = lib.concatStringsSep "\n" (mapAttrsToList (name: val: "export ${name}=\"${val}\"") env);
-  in pkgs.writeShellScript "manage" ''
-    ${setupEnv}
-    exec ${pkg}/bin/paperless-ngx "$@"
-  '';
+  manage =
+    let
+      setupEnv = lib.concatStringsSep "\n" (mapAttrsToList (name: val: "export ${name}=\"${val}\"") env);
+    in
+    pkgs.writeShellScript "manage" ''
+      ${setupEnv}
+      exec ${pkg}/bin/paperless-ngx "$@"
+    '';
 
   # Secure the services
   defaultServiceConfig = {
@@ -170,7 +172,7 @@ in
 
     extraConfig = mkOption {
       type = types.attrs;
-      default = {};
+      default = { };
       description = lib.mdDoc ''
         Extra paperless config options.
 

From ebdfdae156da6d3a94b3e5119f792e33334d6edb Mon Sep 17 00:00:00 2001
From: Martin Puppe <dev@mpuppe.de>
Date: Thu, 9 Feb 2023 22:59:50 +0100
Subject: [PATCH 2/3] nixos/paperless: download NLTK data

Since version 1.10.0 paperless-ngx depends on the NLTK library which is
used to pre-process data for machine learning. NLTK needs certain
data for stemming, stopword removal etc. This data has to be downloaded
first. This commit introduces a new systemd service that does the
downloading.
---
 nixos/modules/services/misc/paperless.nix | 30 +++++++++++++++++++++++
 1 file changed, 30 insertions(+)

diff --git a/nixos/modules/services/misc/paperless.nix b/nixos/modules/services/misc/paperless.nix
index 71ba8fd7075f..4199e7713304 100644
--- a/nixos/modules/services/misc/paperless.nix
+++ b/nixos/modules/services/misc/paperless.nix
@@ -6,6 +6,7 @@ let
   pkg = cfg.package;
 
   defaultUser = "paperless";
+  nltkDir = "/var/cache/paperless/nltk";
 
   # Don't start a redis instance if the user sets a custom redis connection
   enableRedis = !hasAttr "PAPERLESS_REDIS" cfg.extraConfig;
@@ -15,6 +16,7 @@ let
     PAPERLESS_DATA_DIR = cfg.dataDir;
     PAPERLESS_MEDIA_ROOT = cfg.mediaDir;
     PAPERLESS_CONSUMPTION_DIR = cfg.consumptionDir;
+    PAPERLESS_NLTK_DIR = nltkDir;
     GUNICORN_CMD_ARGS = "--bind=${cfg.address}:${toString cfg.port}";
   } // optionalAttrs (config.time.timeZone != null) {
     PAPERLESS_TIME_ZONE = config.time.timeZone;
@@ -49,6 +51,7 @@ let
       cfg.dataDir
       cfg.mediaDir
     ];
+    CacheDirectory = "paperless";
     CapabilityBoundingSet = "";
     # ProtectClock adds DeviceAllow=char-rtc r
     DeviceAllow = "";
@@ -293,6 +296,33 @@ in
       };
     };
 
+    # Download NLTK corpus data
+    systemd.services.paperless-download-nltk-data = {
+      wantedBy = [ "paperless-scheduler.service" ];
+      before = [ "paperless-scheduler.service" ];
+      after = [ "network-online.target" ];
+      serviceConfig = defaultServiceConfig // {
+        User = cfg.user;
+        Type = "oneshot";
+        # Enable internet access
+        PrivateNetwork = false;
+        # Restrict write access
+        BindPaths = [];
+        BindReadOnlyPaths = [
+          "/nix/store"
+          "-/etc/resolv.conf"
+          "-/etc/nsswitch.conf"
+          "-/etc/ssl/certs"
+          "-/etc/static/ssl/certs"
+          "-/etc/hosts"
+          "-/etc/localtime"
+        ];
+        ExecStart = let pythonWithNltk = pkg.python.withPackages (ps: [ ps.nltk ]); in ''
+          ${pythonWithNltk}/bin/python -m nltk.downloader -d '${nltkDir}' punkt snowball_data stopwords
+        '';
+      };
+    };
+
     systemd.services.paperless-consumer = {
       description = "Paperless document consumer";
       # Bind to `paperless-scheduler` so that the consumer never runs

From 7be1ecb2dca7eb5e9119f406bdb8aacaff4dcd0d Mon Sep 17 00:00:00 2001
From: Martin Puppe <dev@mpuppe.de>
Date: Sun, 12 Feb 2023 16:31:50 +0100
Subject: [PATCH 3/3] paperless-ngx: update homepage URL

---
 pkgs/applications/office/paperless-ngx/default.nix | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/pkgs/applications/office/paperless-ngx/default.nix b/pkgs/applications/office/paperless-ngx/default.nix
index 61c607c422e8..c9cd38a37bf7 100644
--- a/pkgs/applications/office/paperless-ngx/default.nix
+++ b/pkgs/applications/office/paperless-ngx/default.nix
@@ -317,7 +317,7 @@ python.pkgs.buildPythonApplication rec {
 
   meta = with lib; {
     description = "Tool to scan, index, and archive all of your physical documents";
-    homepage = "https://paperless-ngx.readthedocs.io/";
+    homepage = "https://docs.paperless-ngx.com/";
     changelog = "https://github.com/paperless-ngx/paperless-ngx/releases/tag/v${version}";
     license = licenses.gpl3Only;
     maintainers = with maintainers; [ lukegb gador erikarvstedt ];