Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[FEA] Test all types and supported expressions for hybrid scan filter pushdown #12124

Open
thirtiseven opened this issue Feb 13, 2025 · 0 comments
Assignees
Labels
test Only impacts tests

Comments

@thirtiseven
Copy link
Collaborator

Is your feature request related to a problem? Please describe.
In hybrid parquet scan, we are missing a lot of tests around predicate push down for various data types that we say we support: byte, short, int, long, float, double, date, decimal_32, decimal_64, string, struct, array, ...

We also support a large number of expressions.

val ansiOn = Seq(
classOf[Acos],
classOf[Acosh],
classOf[AddMonths],
classOf[Alias],
classOf[And],
classOf[ArrayAggregate],
classOf[ArrayContains],
classOf[ArrayDistinct],
classOf[ArrayExcept],
classOf[ArrayExists],
classOf[ArrayForAll],
classOf[ArrayIntersect],
classOf[ArrayJoin],
classOf[ArrayMax],
classOf[ArrayMin],
classOf[ArrayPosition],
classOf[ArrayRemove],
classOf[ArrayRepeat],
classOf[ArraySort],
classOf[ArraysZip],
classOf[Ascii],
classOf[Asin],
classOf[Asinh],
classOf[Atan],
classOf[Atan2],
classOf[Atanh],
classOf[AttributeReference],
classOf[BitLength],
classOf[BitwiseAnd],
classOf[BitwiseOr],
classOf[Cbrt],
classOf[Ceil],
classOf[Chr],
classOf[Concat],
classOf[Cos],
classOf[Cosh],
classOf[Crc32],
classOf[CreateNamedStruct],
classOf[DateAdd],
classOf[DateDiff],
classOf[DateFormatClass],
classOf[DateFromUnixDate],
classOf[DateSub],
classOf[DayOfMonth],
classOf[DayOfWeek],
classOf[DayOfYear],
classOf[ElementAt],
classOf[EqualNullSafe],
classOf[EqualTo],
classOf[Exp],
classOf[Expm1],
classOf[FindInSet],
classOf[Flatten],
classOf[Floor],
classOf[FromUTCTimestamp],
classOf[FromUnixTime],
classOf[GetJsonObject],
classOf[GetMapValue],
classOf[GreaterThan],
classOf[GreaterThanOrEqual],
classOf[Greatest],
classOf[Hex],
classOf[Hour],
classOf[If],
classOf[In],
classOf[IsNaN],
classOf[IsNotNull],
classOf[IsNull],
classOf[LastDay],
classOf[Least],
classOf[Left],
classOf[Length],
classOf[LengthOfJsonArray],
classOf[LessThan],
classOf[Levenshtein],
classOf[Like],
classOf[Literal],
classOf[Log],
classOf[Log10],
classOf[Log2],
classOf[Lower],
classOf[MapFromArrays],
classOf[MapKeys],
classOf[MapValues],
classOf[MapZipWith],
classOf[Md5],
classOf[MicrosToTimestamp],
classOf[MillisToTimestamp],
classOf[Minute],
classOf[MonotonicallyIncreasingID],
classOf[Month],
classOf[NaNvl],
classOf[NextDay],
classOf[Not],
classOf[Or],
classOf[Overlay],
classOf[Pi],
classOf[Pow],
classOf[Quarter],
classOf[Rand],
classOf[Remainder],
classOf[Reverse],
classOf[Rint],
classOf[Round],
classOf[Second],
classOf[Sha1],
classOf[Sha2],
classOf[ShiftLeft],
classOf[ShiftRight],
classOf[Shuffle],
classOf[Sin],
classOf[Size],
classOf[SortArray],
classOf[SoundEx],
classOf[SparkPartitionID],
classOf[Sqrt],
classOf[Stack],
classOf[StringInstr],
classOf[StringLPad],
classOf[StringRPad],
classOf[StringRepeat],
classOf[StringReplace],
classOf[StringToMap],
classOf[StringTrim],
classOf[StringTrimLeft],
classOf[StringTrimRight],
classOf[Substring],
classOf[SubstringIndex],
classOf[Tan],
classOf[Tanh],
classOf[ToDegrees],
classOf[ToRadians],
classOf[ToUnixTimestamp],
classOf[UnaryPositive],
classOf[Unhex],
classOf[UnixMicros],
classOf[UnixMillis],
classOf[UnixSeconds],
classOf[Upper],
classOf[Uuid],
classOf[WeekDay],
classOf[WeekOfYear],
classOf[WidthBucket],
classOf[Year],
classOf[ZipWith]
)
// Some functions are fully supported with ANSI mode off
lazy val ansiOff = Seq(
classOf[Remainder],
classOf[Multiply],
classOf[Add],
classOf[Subtract],
classOf[Divide],
classOf[Abs],
classOf[Pmod]
)

But we haven't tested them as part of predicate push down either.

Describe the solution you'd like
Add tests for all the data types and expressions that we support in hybrid parquet scan filter pushdown.

Additional context
Originally posted by @revans2 in #12112 (review)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
test Only impacts tests
Projects
None yet
Development

No branches or pull requests

2 participants