build: make python test loop easier:

build.sbt

@@ -60,8 +60,12 @@ cleanCondaEnvTask := {
     new File(".")) ! s.log
 }
 
+def isWindows: Boolean = {
+  sys.props("os.name").toLowerCase.contains("windows")
+}
+
 def osPrefix: Seq[String] = {
-  if (sys.props("os.name").toLowerCase.contains("windows")) {
+  if (isWindows) {
     Seq("cmd", "/C")
   } else {
     Seq()
@@ -98,6 +102,15 @@ generatePythonDoc := {
 
 }
 
+val pythonizedVersion = settingKey[String]("Pythonized version")
+pythonizedVersion := {
+  if (version.value.contains("-")){
+    version.value.split("-".head).head + ".dev1"
+  }else{
+    version.value
+  }
+}
+
 def uploadToBlob(source: String, dest: String,
                  container: String, log: ManagedLogger,
                  accountName: String="mmlspark"): Int = {
@@ -161,14 +174,6 @@ publishR := {
   singleUploadToBlob(rPackage.toString,rPackage.getName, "rrr", s.log)
 }
 
-def pythonizeVersion(v: String): String = {
-  if (v.contains("-")){
-    v.split("-".head).head + ".dev1"
-  }else{
-    v
-  }
-}
-
 packagePythonTask := {
   val s = streams.value
   (run in IntegrationTest2).toTask("").value
@@ -180,8 +185,7 @@ packagePythonTask := {
   Process(
     activateCondaEnv ++
       Seq(s"python", "setup.py", "bdist_wheel", "--universal", "-d", s"${pythonPackageDir.absolutePath}"),
-    pythonSrcDir,
-    "MML_PY_VERSION" -> pythonizeVersion(version.value)) ! s.log
+    pythonSrcDir) ! s.log
 }
 
 val installPipPackageTask = TaskKey[Unit]("installPipPackage", "install python sdk")
@@ -192,7 +196,7 @@ installPipPackageTask := {
   packagePythonTask.value
   Process(
     activateCondaEnv ++ Seq("pip", "install",
-      s"mmlspark-${pythonizeVersion(version.value)}-py2.py3-none-any.whl"),
+      s"mmlspark-${pythonizedVersion.value}-py2.py3-none-any.whl"),
     pythonPackageDir) ! s.log
 }
 
@@ -211,7 +215,6 @@ testPythonTask := {
       "mmlsparktest"
     ),
     new File("target/scala-2.11/generated/test/python/"),
-    "MML_VERSION" -> version.value
   ) ! s.log
 }
 
@@ -319,7 +322,7 @@ val settings = Seq(
   logBuffered in Test := false,
   buildInfoKeys := Seq[BuildInfoKey](
     name, version, scalaVersion, sbtVersion,
-    baseDirectory, datasetDir),
+    baseDirectory, datasetDir, pythonizedVersion),
   parallelExecution in Test := false,
   test in assembly := {},
   assemblyMergeStrategy in assembly := {

src/it/scala/com/microsoft/ml/spark/codegen/CodegenConfig.scala

@@ -59,6 +59,9 @@ object Config {
         |CNTK library, images, and text.
         |"\""
         |
+        |__version__ = "${BuildInfo.pythonizedVersion}"
+        |__spark_package_version__ = "${BuildInfo.version}"
+        |
         |$importString
         |""".stripMargin
   }

src/main/python/setup.py

@@ -3,10 +3,26 @@
 
 import os
 from setuptools import setup, find_packages
+import codecs
+import os.path
+
+def read(rel_path):
+    here = os.path.abspath(os.path.dirname(__file__))
+    with codecs.open(os.path.join(here, rel_path), 'r') as fp:
+        return fp.read()
+
+def get_version(rel_path):
+    for line in read(rel_path).splitlines():
+        if line.startswith('__version__'):
+            delim = '"' if '"' in line else "'"
+            return line.split(delim)[1]
+    else:
+        raise RuntimeError("Unable to find version string.")
+
 
 setup(
     name="mmlspark",
-    version=os.environ["MML_PY_VERSION"],
+    version=get_version("mmlspark/__init__.py"),
     description="Microsoft ML for Spark",
     long_description="Microsoft ML for Apache Spark contains Microsoft's open source " +
                      "contributions to the Apache Spark ecosystem",

src/test/python/mmlsparktest/spark.py

@@ -3,11 +3,12 @@
 
 from pyspark.sql import SparkSession, SQLContext
 import os
+import mmlspark
 
 spark = SparkSession.builder \
     .master("local[*]") \
     .appName("PysparkTests") \
-    .config("spark.jars.packages", "com.microsoft.ml.spark:mmlspark_2.11:" + os.environ["MML_VERSION"]) \
+    .config("spark.jars.packages", "com.microsoft.ml.spark:mmlspark_2.11:" + mmlspark.__spark_package_version__) \
     .config("spark.executor.heartbeatInterval", "60s") \
     .config("spark.sql.shuffle.partitions", 10) \
     .config("spark.sql.crossJoin.enabled", "true") \