andygrove commented on code in PR #2614:
URL: https://github.com/apache/datafusion-comet/pull/2614#discussion_r2449452211
##########
fuzz-testing/src/main/scala/org/apache/comet/fuzz/Meta.scala:
##########
@@ -35,100 +61,261 @@ object Meta {
(DataTypes.createDecimalType(10, 2), 0.2),
(DataTypes.DateType, 0.2),
(DataTypes.TimestampType, 0.2),
- // TimestampNTZType only in Spark 3.4+
- // (DataTypes.TimestampNTZType, 0.2),
+ (DataTypes.TimestampNTZType, 0.2),
(DataTypes.StringType, 0.2),
(DataTypes.BinaryType, 0.1))
- val stringScalarFunc: Seq[Function] = Seq(
- Function("substring", 3),
- Function("coalesce", 1),
- Function("starts_with", 2),
- Function("ends_with", 2),
- Function("contains", 2),
- Function("ascii", 1),
- Function("bit_length", 1),
- Function("octet_length", 1),
- Function("upper", 1),
- Function("lower", 1),
- Function("chr", 1),
- Function("init_cap", 1),
- Function("trim", 1),
- Function("ltrim", 1),
- Function("rtrim", 1),
- Function("string_space", 1),
- Function("rpad", 2),
- Function("rpad", 3), // rpad can have 2 or 3 arguments
- Function("hex", 1),
- Function("unhex", 1),
- Function("xxhash64", 1),
- Function("sha1", 1),
- // Function("sha2", 1), -- needs a second argument for number of bits
- Function("substring", 3),
- Function("btrim", 1),
- Function("concat_ws", 2),
- Function("repeat", 2),
- Function("length", 1),
- Function("reverse", 1),
- Function("instr", 2),
- Function("replace", 2),
- Function("translate", 2))
-
- val dateScalarFunc: Seq[Function] =
- Seq(Function("year", 1), Function("hour", 1), Function("minute", 1),
Function("second", 1))
+ private def createFunctionWithInputs(name: String, inputs: Seq[SparkType]):
Function = {
+ Function(name, Seq(FunctionSignature(inputs)))
+ }
+ private def createFunctions(name: String, signatures:
Seq[FunctionSignature]): Function = {
+ Function(name, signatures)
+ }
+
+ private def createUnaryStringFunction(name: String): Function = {
+ createFunctionWithInputs(name, Seq(SparkStringType))
+ }
+
+ private def createUnaryNumericFunction(name: String): Function = {
+ createFunctionWithInputs(name, Seq(SparkNumericType))
+ }
+
+ // Math expressions (corresponds to mathExpressions in QueryPlanSerde)
val mathScalarFunc: Seq[Function] = Seq(
- Function("abs", 1),
- Function("acos", 1),
- Function("asin", 1),
- Function("atan", 1),
- Function("Atan2", 1),
- Function("Cos", 1),
- Function("Exp", 2),
- Function("Ln", 1),
- Function("Log10", 1),
- Function("Log2", 1),
- Function("Pow", 2),
- Function("Round", 1),
- Function("Signum", 1),
- Function("Sin", 1),
- Function("Sqrt", 1),
- Function("Tan", 1),
- Function("Ceil", 1),
- Function("Floor", 1),
- Function("bool_and", 1),
- Function("bool_or", 1),
- Function("bitwise_not", 1))
+ createUnaryNumericFunction("abs"),
+ createUnaryNumericFunction("acos"),
+ createUnaryNumericFunction("asin"),
+ createUnaryNumericFunction("atan"),
+ createFunctionWithInputs("atan2", Seq(SparkNumericType, SparkNumericType)),
+ createUnaryNumericFunction("cos"),
+ createUnaryNumericFunction("exp"),
+ createUnaryNumericFunction("expm1"),
+ createFunctionWithInputs("log", Seq(SparkNumericType, SparkNumericType)),
+ createUnaryNumericFunction("log10"),
+ createUnaryNumericFunction("log2"),
+ createFunctionWithInputs("pow", Seq(SparkNumericType, SparkNumericType)),
+ createFunctionWithInputs("remainder", Seq(SparkNumericType,
SparkNumericType)),
+ createFunctions(
+ "round",
+ Seq(
+ FunctionSignature(Seq(SparkNumericType)),
+ FunctionSignature(Seq(SparkNumericType, SparkIntType)))),
+ createUnaryNumericFunction("signum"),
+ createUnaryNumericFunction("sin"),
+ createUnaryNumericFunction("sqrt"),
+ createUnaryNumericFunction("tan"),
+ createUnaryNumericFunction("ceil"),
+ createUnaryNumericFunction("floor"),
+ createFunctionWithInputs("unary_minus", Seq(SparkNumericType)))
+
+ // Hash expressions (corresponds to hashExpressions in QueryPlanSerde)
+ val hashScalarFunc: Seq[Function] = Seq(
+ createFunctionWithInputs("md5", Seq(SparkAnyType)),
+ createFunctionWithInputs("murmur3_hash", Seq(SparkAnyType)), // TODO can
take multiple columns
+ createFunctionWithInputs("sha2", Seq(SparkAnyType, SparkIntType)))
+ // String expressions (corresponds to stringExpressions in QueryPlanSerde)
+ val stringScalarFunc: Seq[Function] = Seq(
+ createUnaryStringFunction("ascii"),
+ createUnaryStringFunction("bit_length"),
+ createUnaryStringFunction("chr"),
+ createFunctionWithInputs("concat_ws", Seq(SparkStringType,
SparkStringType)),
Review Comment:
I added support for concat with two arguments. According to the
[docs](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.concat.html):
```
The function works with strings, numeric, binary and compatible array
columns.
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
