[jira] [Created] (CALCITE-5121) Rethrow RuntimeException as SQLException in Driver#onConnectionInit
Gregory Hart created CALCITE-5121:
-
Summary: Rethrow RuntimeException as SQLException in
Driver#onConnectionInit
Key: CALCITE-5121
URL: https://issues.apache.org/jira/browse/CALCITE-5121
Project: Calcite
Issue Type: Bug
Components: core
Affects Versions: 1.30.0
Reporter: Gregory Hart
The documentation for org.apache.calcite.avatica.Handler#onConnectionInit
states that a SQLException should be thrown "on error". However, the
implementation of onConnectInit in org.apache.calcite.jdbc.Driver allows for
other types of exceptions to be thrown. This also breaks the documentation
contract for java.sql.Driver#connect since a RuntimeException could be throw
instead of the expected SQLException.
*Steps to reproduce:*
DriverManager.getConnection("jdbc:calcite:schemaType=JDBC;schema.jdbcUrl=jdbc:invalid:");
*Expected result:*
A SQLException is thrown
*Actual result:*
A RuntimeException is thrown
{code:java}
java.lang.RuntimeException: Error instantiating JsonCustomSchema(name=adhoc)
at org.apache.calcite.model.ModelHandler.visit(ModelHandler.java:277)
at
org.apache.calcite.model.JsonCustomSchema.accept(JsonCustomSchema.java:66)
at org.apache.calcite.model.ModelHandler.visit(ModelHandler.java:200)
at org.apache.calcite.model.ModelHandler.(ModelHandler.java:106)
at org.apache.calcite.jdbc.Driver$1.onConnectionInit(Driver.java:101)
at
org.apache.calcite.avatica.UnregisteredDriver.connect(UnregisteredDriver.java:139)
at java.sql.DriverManager.getConnection(DriverManager.java:664)
at java.sql.DriverManager.getConnection(DriverManager.java:270){code}
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
[jira] [Created] (CALCITE-5120) Throw SQLException from UnregisteredDriver#connect if url is null
Gregory Hart created CALCITE-5120:
-
Summary: Throw SQLException from UnregisteredDriver#connect if url
is null
Key: CALCITE-5120
URL: https://issues.apache.org/jira/browse/CALCITE-5120
Project: Calcite
Issue Type: Bug
Components: avatica
Affects Versions: avatica-1.20.0
Reporter: Gregory Hart
The documentation for java.sql.Driver#connect states that a SQLException should
be thrown "if a database access error occurs or the url is null." However,
UnregisteredDriver throws a NullPointerException instead.
*Steps to reproduce:*
new org.apache.calcite.avatica.remote.Driver().connect(null, new Properties());
*Expected result:*
A SQLException is thrown
*Actual result:*
A NullPointerException is thrown
{noformat}
java.lang.NullPointerException
at
org.apache.calcite.avatica.UnregisteredDriver.acceptsURL(UnregisteredDriver.java:144)
at
org.apache.calcite.avatica.UnregisteredDriver.connect(UnregisteredDriver.java:130){noformat}
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
[jira] [Created] (CALCITE-5119) SqlNode#unparse hides an alias table when using CTE
Mitsunori Komatsu created CALCITE-5119:
--
Summary: SqlNode#unparse hides an alias table when using CTE
Key: CALCITE-5119
URL: https://issues.apache.org/jira/browse/CALCITE-5119
Project: Calcite
Issue Type: Bug
Components: core
Affects Versions: 1.30.0
Reporter: Mitsunori Komatsu
Hi team,
I found SqlNode#unparse hides an alias table when using CTE.
{code:java}
val planner = Frameworks.getPlanner(config)
val sql = """
WITH cte AS (
select 42 as i
)
SELECT i FROM (
SELECT [i] from cte
) AS [alias]
ORDER BY [alias].[i]
""".trimIndent()
val node = planner.parse(sql)
println(node.toSqlString { c ->
c.withDialect(PrestoSqlDialect.DEFAULT)
.withAlwaysUseParentheses(false)
.withSubQueryStyle(SqlWriter.SubQueryStyle.HYDE)
.withClauseStartsLine(false)
.withClauseEndsLine(false)
}) {code}
This code shows the following
{code:java}
WITH "cte" AS (SELECT 42 AS "i")
(SELECT "i" FROM (SELECT "i" FROM "cte") AS "alias")
ORDER BY "alias"."i"{code}
The `AS "alias"` is inside a block and the last `ORDER BY` can't find `alias`
and this query fails with Presto. I think the output should be as follows
{code:java}
WITH "cte" AS (SELECT 42 AS "i")
SELECT "i" FROM (SELECT "i" FROM "cte") AS "alias"
ORDER BY "alias"."i"{code}
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
[jira] [Created] (CALCITE-5118) SqlDatePartFunction#rewriteCall should check operands length
Benchao Li created CALCITE-5118:
---
Summary: SqlDatePartFunction#rewriteCall should check operands
length
Key: CALCITE-5118
URL: https://issues.apache.org/jira/browse/CALCITE-5118
Project: Calcite
Issue Type: Improvement
Components: core
Affects Versions: 1.30.0
Reporter: Benchao Li
Assignee: Benchao Li
Fix For: 1.31.0
{{SqlDatePartFunction#rewriteCall}} does not check operands length for now, and
it will be rewrite to {{EXTRACT}} unconditionally before validation.
E.g.
{code:sql}
select week(); -- will throw ArrayIndexOutOfBoundsException
select week(date '2022-04-27', 1); -- will just ignore second param
{code}
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
Re: [ANNOUNCE] Master branch renamed to main
Thanks for checking Francis! Hopefully that means we won't have any issues. -- Michael Mior mm...@apache.org Le mar. 26 avr. 2022 à 22:23, Francis Chuang a écrit : > Quick update. I just did a dry-run release for Avatica and it built fine > without needing to set the branch in vlsi-release-plugin's configuration. > > The plugin only pushes tags to git and I don't think it ever commits any > files, so I don't think it needs to know about us renaming our default > branch to main. I also ran `git branch -a` in my local copy of Avatica > and I don't see any branch or references to "master" after running a > dry-run release. > > On 27/04/2022 9:03 am, Michael Mior wrote: > > One outstanding issue is that I was unable to set the branch to main for > > stage-vote-release-plugin. I believe it may default to "master", so I'm > not > > sure if we'll run into issues on the next release. > > -- > > Michael Mior > > mm...@apache.org > > > > > > Le mar. 26 avr. 2022 à 16:30, Julian Hyde a écrit : > > > >> Thanks, Michael. It all looks good from my end. > >> > >> I had opened a pull request yesterday (when the branch was named > >> 'master') and it merged smoothly today (after the rename). > >> > >> Julian > >> > >> On Tue, Apr 26, 2022 at 7:54 AM Michael Mior wrote: > >>> > >>> The master branch has now been renamed to main. If you have an existing > >>> checkout, please follow the steps below (thanks to Francis!) to ensure > >> your > >>> local master branch is also correctly renamed. Note that if you do not > >> have > >>> any local changes, you can also simply clone a new copy of the > repository > >>> and the default branch will now be main. > >>> > >>> 1. Ensure you are in the repository's folder in your command line. > >>> 2. Stash any uncommitted changes: > >>>- git stash > >>> 3. Check out your local master branch: > >>>- git checkout master > >>> 4. Rename your local master branch to main: > >>>- git branch -m master main > >>> 5. Fetch latest commits and branches from remote: > >>>- git fetch > >>> 6. Remove the tracking branch: > >>>- git branch --unset-upstream > >>> 7. Create a new tracking branch (assuming the remote is called > >> origin): > >>>- git branch -u origin/main > >>> 8. Restore your uncommitted changes: > >>>- git stash pop > >>> > >>> -- > >>> Michael Mior > >>> mm...@apache.org > >> > > >
