Build updates

.sbtrc

@@ -1,6 +1,6 @@
 alias assemblyAndCopy=;assembly;copyAssembledJar
 alias assemblyOcrAndCopy=;ocr/assembly;copyAssembledOcrJar
 alias assemblyEvalAndCopy=;evaluation/assembly;copyAssembledEvalJar
-alias assemblyAllAndCopy=;assemblyAndCopy;assemblyOcrAndCopy;assemblyEvalAndCopy;copyAssembledEvalJar
+alias assemblyAllAndCopy=;assemblyEvalAndCopy;assemblyOcrAndCopy
 alias assemblyAndCopyForPyPi=;assembly;copyAssembledJarForPyPi
-alias publishSignedOcr=;ocr/assembly;ocr/publishSigned
+alias publishSignedOcr=;ocr/assembly;ocr/publishSigned

build.sbt

@@ -156,7 +156,6 @@ lazy val root = (project in file("."))
 
 
 val ocrMergeRules: String => MergeStrategy  = {
-
   case "versionchanges.txt" => MergeStrategy.discard
   case "StaticLoggerBinder" => MergeStrategy.discard
   case PathList("META-INF", fileName)
@@ -169,6 +168,19 @@ val ocrMergeRules: String => MergeStrategy  = {
   case _ => MergeStrategy.deduplicate
 }
 
+val evalMergeRules: String => MergeStrategy  = {
+  case "versionchanges.txt" => MergeStrategy.discard
+  case "StaticLoggerBinder" => MergeStrategy.discard
+  case PathList("META-INF", fileName)
+    if List("NOTICE", "MANIFEST.MF", "DEPENDENCIES", "INDEX.LIST").contains(fileName) || fileName.endsWith(".txt")
+  => MergeStrategy.discard
+  case PathList("META-INF", "services", _ @ _*)  => MergeStrategy.first
+  case PathList("META-INF", xs @ _*)  => MergeStrategy.first
+  case PathList("org", "apache", "spark", _ @ _*)  => MergeStrategy.discard
+  case PathList("apache", "commons", "logging", "impl",  xs @ _*)  => MergeStrategy.discard
+  case _ => MergeStrategy.deduplicate
+}
+
 assemblyMergeStrategy in assembly := {
   case PathList("apache.commons.lang3", _ @ _*)  => MergeStrategy.discard
   case PathList("org.apache.hadoop", _ @ _*)  => MergeStrategy.last
@@ -187,6 +199,15 @@ lazy val evaluation = (project in file("eval"))
     name := "spark-nlp-eval",
     version := "2.1.0",
 
+    assemblyMergeStrategy in assembly := evalMergeRules,
+
+    libraryDependencies ++= Seq( 
+      "org.apache.spark" %% "spark-core" % sparkVer,
+      "org.apache.spark" %% "spark-mllib" % sparkVer
+    ),
+
+    test in assembly := {},
+
     publishTo := Some(
       if (isSnapshot.value)
         Opts.resolver.sonatypeSnapshots
@@ -292,9 +313,10 @@ copyAssembledOcrJar := {
   println(s"[info] $jarFilePath copied to $newJarFilePath ")
 }
 
+// Includes spark-nlp, so use sparknlp.jar
 copyAssembledEvalJar := {
   val jarFilePath = (assemblyOutputPath in assembly in "evaluation").value
-  val newJarFilePath = baseDirectory( _ / "python" / "lib" /  "sparknlp-eval.jar").value
+  val newJarFilePath = baseDirectory( _ / "python" / "lib" /  "sparknlp.jar").value
   IO.copyFile(jarFilePath, newJarFilePath)
   println(s"[info] $jarFilePath copied to $newJarFilePath ")
 }

python/run-tests.py

@@ -29,6 +29,5 @@
 
 # Misc tests
 unittest.TextTestRunner().run(UtilitiesTestSpec())
-unittest.TextTestRunner().run(ConfigPathTestSpec())
 unittest.TextTestRunner().run(SerializersTestSpec())
 unittest.TextTestRunner().run(OcrTestSpec())

python/test/misc.py

@@ -17,15 +17,6 @@ def runTest():
         assert(regex_rule.rule() == "\w+")
 
 
-class ConfigPathTestSpec(unittest.TestCase):
-
-    @staticmethod
-    def runTest():
-        assert(get_config_path() == "./application.conf")
-        set_config_path("./somewhere/application.conf")
-        assert(get_config_path() == "./somewhere/application.conf")
-
-
 class SerializersTestSpec(unittest.TestCase):
     def setUp(self):
         self.test_dir = tempfile.mkdtemp()