Merge pull request FRosner#131 from Gerrrr/issue/111

Check if the ddq jar is available in pyddq

python/pyddq/core.py

@@ -1,8 +1,10 @@
-from reporters import ConsoleReporter
-from streams import ByteArrayOutputStream
-import jvm_conversions as jc
+from pyddq.reporters import ConsoleReporter
+from pyddq.streams import ByteArrayOutputStream
+from pyddq.exceptions import JavaClassNotFoundException
+import pyddq.jvm_conversions as jc
 
 from pyspark import sql
+import py4j.java_gateway as jg
 
 
 class Check(object):
@@ -31,6 +33,8 @@ def __init__(self, dataFrame, displayName=None, cacheMethod=None,
             self.jvmCheck = jvmCheck
         else:
             ddq_check = self._jvm.de.frosner.ddq.core.Check
+            if not isinstance(ddq_check, jg.JavaClass):
+                raise JavaClassNotFoundException("de.frosner.ddq.core.Check")
             self.jvmCheck = ddq_check(
                 self._dataFrame._jdf,
                 self._jvm_display_name,

python/pyddq/exceptions.py

@@ -0,0 +1,16 @@
+class JavaClassNotFoundException(Exception):
+    """
+    Raise if required Java class is not found by py4j
+    """
+
+    def __init__(self, java_class):
+        Exception.__init__(self)
+        self.java_class = java_class
+
+    def __str__(self):
+        return "%s. Did you forget to add the jar to the class path?" % (
+            self.java_class
+        )
+
+    def __repr__(self):
+        return "%s: %s" % (self.__class__.__name__, self.java_class)

python/tests/unit/test_check.py

@@ -1,40 +1,52 @@
 import unittest
-from mock import Mock
+from mock import Mock, call
 from pyddq.core import Check
-
+from pyddq.exceptions import JavaClassNotFoundException
+from utils import get_df
 
 class ConstructorTest(unittest.TestCase):
+    def setUp(self):
+        self.df = get_df()
+
     def test_default_args(self):
-        df = Mock()
-        check = Check(df)
-        ddq_check = check._jvm.de.frosner.ddq.core.Check
-        ddq_check.assert_called_with(
-            df._jdf,
-            getattr(ddq_check, "apply$default$2")(),
-            getattr(ddq_check, "apply$default$3")(),
-            getattr(ddq_check, "apply$default$4")(),
-            getattr(ddq_check, "apply$default$5")(),
-        )
+        check = Check(self.df)
+        ddq_core = check._jvm.de.frosner.ddq.core
+        ddq_check = ddq_core.Check
+
+        ddq_core.assert_has_calls([
+            call.Check(
+                self.df._jdf,
+                getattr(ddq_check, "apply$default$2")(),
+                getattr(ddq_check, "apply$default$3")(),
+                getattr(ddq_check, "apply$default$4")(),
+                getattr(ddq_check, "apply$default$5")()
+            )
+        ])
 
     def test_passed_args(self):
-        df = Mock()
         display_name = Mock()
         cache_method = Mock()
         id = Mock()
 
-        df._sc._jvm.scala.Some.apply = Mock(
+        self.df._sc._jvm.scala.Some.apply = Mock(
             side_effect=["Some(displayName)", "Some(cacheMethod)"]
         )
-        check = Check(df, display_name, cache_method, id)
+        check = Check(self.df, display_name, cache_method, id)
+        ddq_core = check._jvm.de.frosner.ddq.core
         ddq_check = check._jvm.de.frosner.ddq.core.Check
 
-        ddq_check.assert_called_with(
-            df._jdf,
-            "Some(displayName)",
-            "Some(cacheMethod)",
-            getattr(ddq_check, "apply$default$4")(),
-            id
-        )
+        ddq_core.assert_has_calls([
+            call.Check(
+                self.df._jdf,
+                "Some(displayName)",
+                "Some(cacheMethod)",
+                getattr(ddq_check, "apply$default$4")(),
+                id
+            )
+        ])
+
+    def test_ddq_jar(self):
+        self.assertRaises(JavaClassNotFoundException, Check, Mock())
 
 if __name__ == '__main__':
     unittest.main()

python/tests/unit/test_constraints.py

@@ -1,13 +1,16 @@
 import unittest
 from mock import Mock
+from py4j.java_gateway import JavaClass
 from pyddq.core import Check
+from utils import get_df
 
 
 class ConstraintTest(unittest.TestCase):
     COLUMN_NAME = "column name"
 
     def setUp(self):
-        self.check = Check(Mock())
+        df = get_df()
+        self.check = Check(df)
         self.jvmCheck = self.check.jvmCheck
 
     def test_hasUniqueKey(self):

python/tests/unit/utils.py

@@ -0,0 +1,11 @@
+from mock import Mock
+from py4j.java_gateway import JavaClass
+
+def get_df():
+    df = Mock()
+    df._sc._jvm.de.frosner.ddq.core.Check = Mock(spec=JavaClass)
+    setattr(df._sc._jvm.de.frosner.ddq.core.Check, "apply$default$2", Mock())
+    setattr(df._sc._jvm.de.frosner.ddq.core.Check, "apply$default$3", Mock())
+    setattr(df._sc._jvm.de.frosner.ddq.core.Check, "apply$default$4", Mock())
+    setattr(df._sc._jvm.de.frosner.ddq.core.Check, "apply$default$5", Mock())
+    return df