NixOS · Dec 14, 2018 · Dec 17, 2018 · Dec 18, 2018 · Dec 19, 2018 · Dec 19, 2018
diff --git a/nixos/lib/testing.nix b/nixos/lib/testing.nix
@@ -116,7 +116,7 @@ in rec {
 
       vms = map (m: m.config.system.build.vm) (lib.attrValues nodes);
 
-      ocrProg = tesseract_4.override { enableLanguages = [ "eng" ]; };
+      ocrProg = tesseract4.override { enableLanguages = [ "eng" ]; };
 
       imagemagick_tiff = imagemagick_light.override { inherit libtiff; };
 

diff --git a/pkgs/applications/graphics/tesseract/4.x.nix b/pkgs/applications/graphics/tesseract/4.x.nix
diff --git a/pkgs/applications/graphics/tesseract/default.nix b/pkgs/applications/graphics/tesseract/default.nix
@@ -1,67 +1,18 @@
-{ stdenv, fetchFromGitHub, autoreconfHook, pkgconfig
-, leptonica, libpng, libtiff, icu, pango, opencl-headers
-# Supported list of languages or `null' for all available languages
-, enableLanguages ? null
-# if you want just a specific list of languages, optionally specify a hash
-# to make tessdata a fixed output derivation.
-, enableLanguagesHash ? (if enableLanguages == null # all languages
-                         then "1h48xfzabhn0ldbx5ib67cp9607pr0zpblsy8z6fs4knn0zznfnw"
-                         else null)
-}:
+{ callPackage, lowPrio }:
 
-let tessdata = stdenv.mkDerivation ({
-  name = "tessdata";
-  src = fetchFromGitHub {
-    owner = "tesseract-ocr";
-    repo = "tessdata";
-    rev = "3cf1e2df1fe1d1da29295c9ef0983796c7958b7d";
-    # when updating don't forget to update the default value fo enableLanguagesHash
-    sha256 = "1v4b63v5nzcxr2y3635r19l7lj5smjmc9vfk0wmxlryxncb4vpg7";
-  };
-  buildCommand = ''
-    cd $src;
-    for lang in ${if enableLanguages==null then "*.traineddata" else stdenv.lib.concatMapStringsSep " " (x: x+".traineddata") enableLanguages} ; do
-      install -Dt $out/share/tessdata $src/$lang ;
-    done;
-  '';
-  preferLocalBuild = true;
-  } // (stdenv.lib.optionalAttrs (enableLanguagesHash != null) {
-  # when a hash is given, we make this a fixed output derivation.
-  outputHashMode = "recursive";
-  outputHashAlgo = "sha256";
-  outputHash = enableLanguagesHash;
-  }));
+let
+  base3 = callPackage ./tesseract3.nix {};
+  base4 = callPackage ./tesseract4.nix {};
+  languages = callPackage ./languages.nix {};
 in
-
-stdenv.mkDerivation rec {
-  name = "tesseract-${version}";
-  version = "3.05.00";
-
-  src = fetchFromGitHub {
-    owner = "tesseract-ocr";
-    repo = "tesseract";
-    rev = version;
-    sha256 = "11wrpcfl118wxsv2c3w2scznwb48c4547qml42s2bpdz079g8y30";
+{
+  tesseract3 = callPackage ./wrapper.nix {
+    tesseractBase = base3;
+    languages = languages.v3;
   };
 
-  enableParallelBuilding = true;
-
-  nativeBuildInputs = [ pkgconfig autoreconfHook ];
-  buildInputs = [ leptonica libpng libtiff icu pango opencl-headers ];
-
-  LIBLEPT_HEADERSDIR = "${leptonica}/include";
-
-  postInstall = ''
-    for i in ${tessdata}/share/tessdata/*; do
-      ln -s $i $out/share/tessdata;
-    done
-  '';
-
-  meta = {
-    description = "OCR engine";
-    homepage = https://github.com/tesseract-ocr/tesseract;
-    license = stdenv.lib.licenses.asl20;
-    maintainers = with stdenv.lib.maintainers; [viric];
-    platforms = with stdenv.lib.platforms; linux ++ darwin;
-  };
+  tesseract4 = lowPrio (callPackage ./wrapper.nix {
+    tesseractBase = base4;
+    languages = languages.v4;
+  });
 }
diff --git a/pkgs/applications/graphics/tesseract/fetch-language-hashes b/pkgs/applications/graphics/tesseract/fetch-language-hashes
@@ -0,0 +1,35 @@
+#!/usr/bin/env bash
+
+# Usage:
+#   ./fetch-language-hashes <tessdataRev> [<language code>…]
+#
+#   Fetches all languages if no language codes are given.
+#
+# Example:
+#   ./fetch-language-hashes 4.0.0 eng spa
+#
+#   Output:
+#     eng = "0iy0...";
+#     spa = "15kw...";
+
+set -e
+
+(( $# >= 1 )) || exit 1
+tessdataRev=$1
+shift
+
+if (( $# > 0 )); then
+    langCodes="$@"
+else
+    repoPage=$(curl -fs https://github.com/tesseract-ocr/tessdata/tree/$tessdataRev || {
+                   >&2 echo "Invalid tessdataRev: $tessdataRev"
+                   exit 1
+               })
+    langCodes=$(echo $(echo "$repoPage" | grep -ohP "(?<=/)[^/]+?(?=\.traineddata)" | sort))
+fi
+
+for lang in $langCodes; do
+    url=https://github.com/tesseract-ocr/tessdata/raw/$tessdataRev/$lang.traineddata
+    hash=$(nix-prefetch-url $url 2>/dev/null)
+    echo "$lang = \"$hash\";"
+done