add repr for FlattedValuesSerializer

marmbrus · Apr 7, 2015 · d2f053b · d2f053b
1 parent c6a2f8d
commit d2f053b
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 2 deletions.
diff --git a/python/pyspark/serializers.py b/python/pyspark/serializers.py
@@ -239,6 +239,9 @@ def _batched(self, iterator):
     def load_stream(self, stream):
         return self.serializer.load_stream(stream)
 
+    def __repr__(self):
+        return "FlattedValuesSerializer(%d)" % self.batchSize
+
 
 class AutoBatchedSerializer(BatchedSerializer):
     """
@@ -271,7 +274,7 @@ def __eq__(self, other):
         return (isinstance(other, AutoBatchedSerializer) and
                 other.serializer == self.serializer and other.bestSize == self.bestSize)
 
-    def __str__(self):
+    def __repr__(self):
         return "AutoBatchedSerializer(%s)" % str(self.serializer)
 
 

diff --git a/python/pyspark/shuffle.py b/python/pyspark/shuffle.py
@@ -313,7 +313,7 @@ def mergeCombiners(self, iterator, limit=None):
                 else:
                     batch *= 1.5
 
-        if get_used_memory() >= limit:
+        if limit and get_used_memory() >= limit:
             self._spill()
 
     def _spill(self):