Skip to content

[SPARK-31205][SQL] support string literal as the second argument of date_add/date_sub functions #27965

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.

Sign up for GitHub

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

Closed
cloud-fan wants to merge 3 commits into from
Closed

[SPARK-31205][SQL] support string literal as the second argument of date_add/date_sub functions #27965

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
9caf98c
support string literal as the second argument of date_add/date_sub fu…
cloud-fan Mar 19, 2020
94e2fce
address comments
cloud-fan Mar 20, 2020
bd6be16
reduce scope
cloud-fan Mar 23, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
address comments
  • Loading branch information
@cloud-fan
cloud-fan committed Mar 20, 2020
commit 94e2fce05621ac90a3161cb2ae4a2205a2f1db0f
2 changes: 1 addition & 1 deletion docs/sql-migration-guide.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -113,7 +113,7 @@ license: |

### UDFs and Built-in Functions

- Since Spark 3.0, the `date_add` and `date_sub` functions only accepts int, smallint, tinyint as the 2nd argument, fractional and string types are not valid anymore, e.g. `date_add(cast('1964-05-23' as date), '12.34')` will cause `AnalysisException`. Note that, string literals are still allowed, but Spark will throw Analysis Exception if the string is not a valid integer. In Spark version 2.4 and earlier, if the 2nd argument is fractional or string value, it will be coerced to int value, and the result will be a date value of `1964-06-04`.
- Since Spark 3.0, the `date_add` and `date_sub` functions only accept int, smallint, tinyint as the 2nd argument, fractional and non-literal string are not valid anymore, e.g. `date_add(cast('1964-05-23' as date), 12.34)` will cause `AnalysisException`. Note that, string literals are still allowed, but Spark will throw Analysis Exception if the string content is not a valid integer. In Spark version 2.4 and earlier, if the 2nd argument is fractional or string value, it will be coerced to int value, and the result will be a date value of `1964-06-04`.
Copy link
Member

@dongjoon-hyun dongjoon-hyun Mar 20, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The new approach looks good.


- Since Spark 3.0, the function `percentile_approx` and its alias `approx_percentile` only accept integral value with range in `[1, 2147483647]` as its 3rd argument `accuracy`, fractional and string types are disallowed, e.g. `percentile_approx(10.0, 0.2, 1.8D)` will cause `AnalysisException`. In Spark version 2.4 and earlier, if `accuracy` is fractional or string value, it will be coerced to an int value, `percentile_approx(10.0, 0.2, 1.8D)` is operated as `percentile_approx(10.0, 0.2, 1)` which results in `10.0`.

Expand Down
2 changes: 1 addition & 1 deletion sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/TypeCoercion.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1049,7 +1049,7 @@ object TypeCoercion {
/**
* A special rule to support string literal as the second argument of date_add/date_sub functions,
* to keep backward compatibility as a temporary workaround.
* TODO: revisit the type coercion rules for string.
* TODO(SPARK-28589): implement ANSI type type coercion and handle string literals.
*/
object StringLiteralCoercion extends TypeCoercionRule {
Copy link
Member

@dongjoon-hyun dongjoon-hyun Mar 20, 2020 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This causes a behavior difference in arithmatic operations, too. Could you describe the following change in the PR description? New one looks reasonable to me.

2.4.5 and 3.0.0-preview2

scala> sql("select (cast('2020-03-28' AS DATE) + '1')").show
org.apache.spark.sql.AnalysisException: cannot resolve '(CAST('2020-03-28' AS DATE) + CAST('1' AS DOUBLE))' due to data type mismatch: differing types in '(CAST('2020-03-28' AS DATE) + CAST('1' AS DOUBLE))' (date and double).; line 1 pos 8;

This PR.

scala> sql("select (cast('2020-03-28' AS DATE) + '1')").show
+-------------------------------------+
|date_add(CAST(2020-03-28 AS DATE), 1)|
+-------------------------------------+
|                           2020-03-29|
+-------------------------------------+

Copy link
Member

@maropu maropu Mar 20, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You forgot to write a function name in the second query above?

scala> sql("select (cast('2020-03-28' AS DATE) + '1')").show
                 ^^^^

Copy link
Member

@dongjoon-hyun dongjoon-hyun Mar 22, 2020 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

No~ The both queries are the same. What I meant was it's the behavior of this PR; this PR extends expressions, too.

Copy link
Member

@maropu maropu Mar 22, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Oh, I see.

override protected def coerceTypes(plan: LogicalPlan): LogicalPlan = plan resolveExpressions {
Expand Down