csharp/Adapter/Microsoft.Spark.CSharp/Sql/DataFrame.cs (4 lines): - line 178: //TODO - add this method if needed to convert Row to collection of T - line 207: public void Explain(bool extended = false) //TODO - GetQueryExecution is called in JVM twice if extendd = true - fix that - line 372: public DataFrame Join(DataFrame otherDataFrame, string joinColumnName) // TODO: need aliasing for self join - line 385: public DataFrame Join(DataFrame otherDataFrame, string[] joinColumnNames) // TODO: need aliasing for self join csharp/Adapter/Microsoft.Spark.CSharp/Interop/Ipc/PayloadHelper.cs (4 lines): - line 122: paramtersBytes.Add(SerDe.ToBytes(((IEnumerable)parameter).Count())); //TODO - Count() will traverse the collection - change interface? - line 132: paramtersBytes.Add(SerDe.ToBytes(((IEnumerable)parameter).Count())); //TODO - Count() will traverse the collection - change interface? - line 138: paramtersBytes.Add(SerDe.ToBytes(((IEnumerable)parameter).Count())); //TODO - Count() will traverse the collection - change interface? - line 182: internal static byte[] GetTypeId(Type type) //TODO - support other types csharp/Adapter/Microsoft.Spark.CSharp/Sql/Types.cs (2 lines): - line 96: else // {name: age, type: bigint,...} // TODO: validate more JTokenType other than Object - line 508: elementType as StructType ?? convertArrayTypeToStructTypeFunc(elementType, valueOfArray.Length)).Values; // TODO: this part may have some problems, not verified scala/src/main/org/apache/spark/api/csharp/SerDe.scala (2 lines): - line 17: // TODO look into the possibility of reusing SerDe from R implementation - line 228: // TODO: Types not handled right now include csharp/Adapter/Microsoft.Spark.CSharp/Proxy/Ipc/CatalogIpcProxy.cs (2 lines): - line 58: throw new NotImplementedException(); //TODO - implement - line 72: throw new NotImplementedException(); //TODO - implement csharp/Adapter/Microsoft.Spark.CSharp/Interop/Ipc/JvmBridge.cs (2 lines): - line 83: //TODO - add boolean instead of int in the backend - line 94: switch (typeAsChar) //TODO - add support for other types csharp/Adapter/Microsoft.Spark.CSharp/Sql/SparkSession.cs (1 line): - line 32: //setter is used only for testing...//TODO - refactor csharp/AdapterTest/DataFrameTest.cs (1 line): - line 25: //TODO - complete impl csharp/Adapter/Microsoft.Spark.CSharp/Streaming/Kafka.cs (1 line): - line 180: /// TODO: move to scala and merge into DynamicPartitionKafkaRDD.getPartitions to remove above limitation scala/perf/src/main/com/microsoft/spark/csharp/FreebaseDeletionsBenchmark.scala (1 line): - line 105: //TODO - add perf suite for subquery scala/src/main/org/apache/spark/api/csharp/CSharpBackendHandler.scala (1 line): - line 16: // TODO - work with SparkR devs to make this configurable and reuse RBackendHandler csharp/Adapter/Microsoft.Spark.CSharp/Sql/UdfRegistration.cs (1 line): - line 31: //TODO - the following section is a copy of the same functionality in SQLContext..refactoring needed csharp/Adapter/Microsoft.Spark.CSharp/Core/PipelinedRDD.cs (1 line): - line 30: //TODO - give generic types a better id csharp/Adapter/Microsoft.Spark.CSharp/Proxy/Ipc/StreamingContextIpcProxy.cs (1 line): - line 84: // TODO: We don't know whether accumulator variable is used before restart. We just start accumuator server for safety. csharp/Perf/Microsoft.Spark.CSharp/FreebaseDeletionsBenchmark.cs (1 line): - line 112: //TODO - add perf suite for subquery csharp/AdapterTest/SparkContextTest.cs (1 line): - line 24: //TODO - complete impl csharp/Adapter/Microsoft.Spark.CSharp/Sql/Catalog.cs (1 line): - line 33: // TODO Enable these convenience functions if needed csharp/Adapter/Microsoft.Spark.CSharp/Core/PairRDDFunctions.cs (1 line): - line 333: /// # TODO: add control over map-side aggregation csharp/Adapter/Microsoft.Spark.CSharp/Interop/Ipc/SerDe.cs (1 line): - line 44: public class SerDe //TODO - add ToBytes() for other types csharp/Adapter/Microsoft.Spark.CSharp/Proxy/Ipc/SqlContextIpcProxy.cs (1 line): - line 47: //TODO parameter Dictionary options is not used right now - it is meant to be passed on to data sources csharp/WorkerTest/WorkerTest.cs (1 line): - line 678: // TODO verification should not depends on the output of worker build/localmode/RunSamples.cmd (1 line): - line 28: @rem TODO: this check will fail if "--exe" only exists in the argument list of user application. python/perf/FreebaseDeletionsBenchmark.py (1 line): - line 92: # TODO - add perf suite for subquery csharp/Adapter/Microsoft.Spark.CSharp/Sql/SqlContext.cs (1 line): - line 60: //TODO - remove this constructor after fixing unit tests that reference this scala/perf/src/main/com/microsoft/spark/csharp/PerfBenchmark.scala (1 line): - line 49: //TODO - use annotation type scala/src/main/org/apache/spark/sql/api/csharp/SQLUtils.scala (1 line): - line 233: case _ => throw new Exception("Text files with more than 30 columns currently not supported") //TODO - if requirement comes up, generate code for additional columns csharp/Adapter/Microsoft.Spark.CSharp/Streaming/MapWithStateDStream.cs (1 line): - line 347: internal bool timingOut = false; // FIXME: set timingOut to true for those timeouted keys csharp/Worker/Microsoft.Spark.CSharp/Worker.cs (1 line): - line 215: // fetch names of includes - not used //TODO - complete the impl csharp/Adapter/Microsoft.Spark.CSharp/Core/Profiler.cs (1 line): - line 12: //TODO - complete the impl scala/src/main/org/apache/spark/api/csharp/CSharpBackend.scala (1 line): - line 58: //TODO - work with SparkR devs to make this configurable and reuse RBackend csharp/AdapterTest/SparkConfTest.cs (1 line): - line 23: //TODO - complete impl