[
https://issues.apache.org/jira/browse/CALCITE-5538?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17692319#comment-17692319
]
Mihai Budiu commented on CALCITE-5538:
--------------------------------------
I have tried to create a test case, and I notice that the SQL parser actually
goes through DateTimeUtils.parsePrecisionDateTimeLiteral, which parses
correctly values after the decimal point.
So this is not really a problem of the SQL parser, but only a problem of this
internal class.
I still think it's worth fixing, though, but since this internal class has no
unit tests, I don't know how to write a test that exercises this path.
> TimestampString rejects timestamp literals that end with 0 after the period
> ---------------------------------------------------------------------------
>
> Key: CALCITE-5538
> URL: https://issues.apache.org/jira/browse/CALCITE-5538
> Project: Calcite
> Issue Type: Bug
> Components: core
> Affects Versions: 1.33.0
> Reporter: Mihai Budiu
> Priority: Trivial
>
> The root cause is that the TimestampString constructor validates a string
> argument with the following regular expression:
> ??"[0-9][0-9][0-9][0-9]{-}[0-9][0-9]{-}[0-9][0-9]"
> + " "
> + "[0-9][0-9]:[0-9][0-9]:[0-9][0-9](.[0-9]*[1-9])?"??
> Unfortunately this rejects perfectly legal timestamp strings such as
> "2023-02-21 10:10:10.000".
> The fix is trivial, if we agree that this is a bug. Is there a deeper reason
> for this validation?
> There seem to be no unit tests for this TimestampString constructor.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
