Merge pull request #39056 from mt-caret/jumanpp

jumanpp: init at 1.02

1 files changed, 24 insertions, 0 deletions

diff --git a/pkgs/tools/text/jumanpp/default.nix b/pkgs/tools/text/jumanpp/default.nix
new file mode 100644
index 000000000000..c4cb12cb6938
--- /dev/null
+++ b/pkgs/tools/text/jumanpp/default.nix
@@ -0,0 +1,24 @@
+{ stdenv, fetchurl, cmake, protobuf }:
+stdenv.mkDerivation rec {
+  name = "jumanpp";
+  version = "2.0.0-rc2";
+
+  src = fetchurl {
+    url = "https://github.com/ku-nlp/${name}/releases/download/v${version}/${name}-${version}.tar.xz";
+    sha256 = "17fzmd0f5m9ayfhsr0mg7hjp3pg1mhbgknhgyd8v87x46g8bg6qp";
+  };
+  buildInputs = [ cmake protobuf ];
+
+  meta = with stdenv.lib; {
+    description = "A Japanese morphological analyser using a recurrent neural network language model (RNNLM)";
+    longDescription = ''
+      JUMAN++ is a new morphological analyser that considers semantic
+      plausibility of word sequences by using a recurrent neural network
+      language model (RNNLM).
+    '';
+    homepage = http://nlp.ist.i.kyoto-u.ac.jp/index.php?JUMAN++;
+    license = licenses.asl20;
+    maintainers = with maintainers; [ mt-caret ];
+    platforms = platforms.all;
+  };
+}