public void TestOption()
{
// arrange
mockDataFrameWriterProxy.Setup(m => m.Options(It.IsAny <Dictionary <string, string> >()));
var dataFrameWriter = new DataFrameWriter(mockDataFrameWriterProxy.Object);
const string key = "path";
const string value = "path_value";
// Act
dataFrameWriter.Option(key, value);
// Assert
mockDataFrameWriterProxy.Verify(m => m.Options(
It.Is <Dictionary <string, string> >(dict => dict[key] == value && dict.Count == 1)), Times.Once);
}
public void TestSignaturesV2_3_X()
{
{
DataFrameWriter dfw = _spark
.Read()
.Schema("age INT, name STRING")
.Json($"{TestEnvironment.ResourceDirectory}people.json")
.Write();
Assert.IsType <DataFrameWriter>(dfw.Mode(SaveMode.Ignore));
Assert.IsType <DataFrameWriter>(dfw.Mode("overwrite"));
Assert.IsType <DataFrameWriter>(dfw.Format("json"));
Assert.IsType <DataFrameWriter>(dfw.Option("stringOption", "value"));
Assert.IsType <DataFrameWriter>(dfw.Option("boolOption", true));
Assert.IsType <DataFrameWriter>(dfw.Option("longOption", 1L));
Assert.IsType <DataFrameWriter>(dfw.Option("doubleOption", 3D));
Assert.IsType <DataFrameWriter>(
dfw.Options(
new Dictionary <string, string>
{
{ "option1", "value1" },
{ "option2", "value2" }
}));
Assert.IsType <DataFrameWriter>(dfw.PartitionBy("age"));
Assert.IsType <DataFrameWriter>(dfw.PartitionBy("age", "name"));
Assert.IsType <DataFrameWriter>(dfw.BucketBy(3, "age"));
Assert.IsType <DataFrameWriter>(dfw.BucketBy(3, "age", "name"));
Assert.IsType <DataFrameWriter>(dfw.SortBy("name"));
}
using (var tempDir = new TemporaryDirectory())
{
DataFrameWriter dfw = _spark
.Read()
.Csv($"{TestEnvironment.ResourceDirectory}people.csv")
.Write();
// TODO: Test dfw.Jdbc without running a local db.
dfw.Option("path", tempDir.Path).SaveAsTable("TestTable");
dfw.InsertInto("TestTable");
dfw.Option("path", $"{tempDir.Path}TestSavePath1").Save();
dfw.Save($"{tempDir.Path}TestSavePath2");
dfw.Json($"{tempDir.Path}TestJsonPath");
dfw.Parquet($"{tempDir.Path}TestParquetPath");
dfw.Orc($"{tempDir.Path}TestOrcPath");
dfw.Text($"{tempDir.Path}TestTextPath");
dfw.Csv($"{tempDir.Path}TestCsvPath");
}
}
public void TestSignaturesV2_3_X()
{
{
DataFrameWriter dfw = _spark
.Read()
.Schema("age INT, name STRING")
.Json($"{TestEnvironment.ResourceDirectory}people.json")
.Write();
Assert.IsType <DataFrameWriter>(dfw.Mode(SaveMode.Ignore));
Assert.IsType <DataFrameWriter>(dfw.Mode("overwrite"));
Assert.IsType <DataFrameWriter>(dfw.Format("json"));
Assert.IsType <DataFrameWriter>(dfw.Option("stringOption", "value"));
Assert.IsType <DataFrameWriter>(dfw.Option("boolOption", true));
Assert.IsType <DataFrameWriter>(dfw.Option("longOption", 1L));
Assert.IsType <DataFrameWriter>(dfw.Option("doubleOption", 3D));
Assert.IsType <DataFrameWriter>(
dfw.Options(
new Dictionary <string, string>
{
{ "option1", "value1" },
{ "option2", "value2" }
}));
Assert.IsType <DataFrameWriter>(dfw.PartitionBy("age"));
Assert.IsType <DataFrameWriter>(dfw.PartitionBy("age", "name"));
Assert.IsType <DataFrameWriter>(dfw.BucketBy(3, "age"));
Assert.IsType <DataFrameWriter>(dfw.BucketBy(3, "age", "name"));
Assert.IsType <DataFrameWriter>(dfw.SortBy("name"));
}
using (var tempDir = new TemporaryDirectory())
{
DataFrameWriter dfw = _spark
.Read()
.Csv($"{TestEnvironment.ResourceDirectory}people.csv")
.Write();
// TODO: Test dfw.Jdbc without running a local db.
dfw.Save($"{tempDir.Path}TestSavePath1");
dfw.Json($"{tempDir.Path}TestJsonPath");
dfw.Parquet($"{tempDir.Path}TestParquetPath");
dfw.Orc($"{tempDir.Path}TestOrcPath");
dfw.Text($"{tempDir.Path}TestTextPath");
dfw.Csv($"{tempDir.Path}TestCsvPath");
dfw.Option("path", tempDir.Path).SaveAsTable("TestTable");
dfw.InsertInto("TestTable");
// In Spark 3.1.1+ setting the `path` Option and then calling .Save(path) is not
// supported unless `spark.sql.legacy.pathOptionBehavior.enabled` conf is set.
// .Json(path), .Parquet(path), etc follow the same code path so the conf
// needs to be set in these scenarios as well.
dfw.Option("path", $"{tempDir.Path}TestSavePath2").Save();
}
}