[SPARK-53319][SQL] Support the time type by try_make_timestamp_ltz() #52063
Conversation
uros-db
changed the title
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/datetimeExpressions.scala
Outdated
Show resolved
Hide resolved
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/datetimeExpressions.scala
Show resolved
Hide resolved
|
changes look good, but I see the build is failing. Seems like it could be a transient issue, can you try rerunning the build? |
|
Docker Integration failure looks unrelated to these changes, re-running to confirm. |
|
cc @cloud-fan for review |
|
@cloud-fan Please review. |
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/datetimeExpressions.scala
Outdated
Show resolved
Hide resolved
…essions/datetimeExpressions.scala
| if (numArgs == 2 || numArgs == 3) { | ||
| // Overload for: date, time[, timezone]. | ||
| TryEval( | ||
| MakeTimestampFromDateTime( |
There was a problem hiding this comment.
shall we add the failOnError flag to MakeTimestampFromDateTime as well? TryEval is not very reliable as it may mistakenly catch errors from the function argument expressions.
There was a problem hiding this comment.
I don't think we have to do that here, because invalid date/time here would throw parse exceptions before that point anyways (both in ANSI and non-ANSI). In this sense, it's a different kind of situation compared to integer construction (year, month, day, hour, min) of MakeTimestamp below for example.
3 participants
What changes were proposed in this pull request?
In the PR, I propose to extend the
try_make_timestamp_ltzfunction, and accept a date and time fields.Syntax
Arguments
date: A date expression.time: A time expression.Returns
A
TIMESTAMP_LTZ.Examples
Why are the changes needed?
Users will be able to create a timestamp with local time zone by combining a time and a date.
Does this PR introduce any user-facing change?
Yes, this extends
try_make_timestamp_ltzto accept additional kinds of inputs.How was this patch tested?
Added new e2e SQL tests in corresponding golden files.
Was this patch authored or co-authored using generative AI tooling?
No.