We Keep Coding

Hibernate GROUP BY invalid for Oracle

I created an Hibernate query like this:
select new ProjectForUser
(p.projectId, p.name, p.description, p.client, p.startDate, p.endDate,
p.liveDate, p.projectState, p.overallRagStatus, p.scopeRagStatus, p.flt,
up.projectManager)
from UserProjectAssociation up left join up.project p
where up.user.id = :userId and up.project.projectState != 'ARCHIVED'
group by p.projectId
to retrieve projects for user. Projects are stored in one table, users in the other and UserProjectAssociation is a joining table with additional attribute - projectManager.
The whole query works fine for H2 database, but on oracle I get this error:
ORA-00979: not a GROUP BY expression
What is a correct way to use GROUP BY for Oracle?
Thanks!

The Problem. ORA-00979 occurs when the GROUP BY clause does not contain all the expressions in the SELECT clause. Any SELECT expression that is not included in the GROUP function must be listed in the GROUP BY clause. These are AVG, COUNT, MAX, MIN, SUM, STDDEV, and VARIANCE.
see: https://www.tekstream.com/resources/ora-00979-not-a-group-by-expression/

Delete from table on same select same table mariadb using jpa

I need delete from table on operation of same table .JPA query is
DELETE FROM com.model.ElectricityLedgerEntity a
Where a.elLedgerid IN
(SELECT P.elLedgerid FROM
(SELECT MAX(b.elLedgerid)
FROM com.model.ElectricityLedgerEntity b
WHERE b.accountId='24' and b.ledgerType='Electricity Ledger' and b.postType='ARREARS') P );
I got this error:
with root cause org.hibernate.hql.ast.QuerySyntaxException: unexpected
token: ( near line 1, column 109 [DELETE FROM
com.bcits.bfm.model.ElectricityLedgerEntity a Where a.elLedgerid IN (
SELECT P.elLedgerid FROM ( SELECT MAX(b.elLedgerid) FROM
com.bcits.ElectricityLedgerEntity b WHERE b.accountId='24'
and b.ledgerType='Electricity Ledger' and b.postType='ARREARS') P ) ]
at
org.hibernate.hql.ast.QuerySyntaxException.convert(QuerySyntaxException.java:54)
at
org.hibernate.hql.ast.QuerySyntaxException.convert(QuerySyntaxException.java:47)
at
org.hibernate.hql.ast.ErrorCounter.throwQueryException(ErrorCounter.java:82)
at
org.hibernate.hql.ast.QueryTranslatorImpl.parse(QueryTranslatorImpl.java:284)
Same query is running on mysql terminal ,but this is not working with jpa .Can any one tell me how i can write this query using jpa .

I don't understand why do you use Pbefore the last parenthesis...
The following code is not enough ?
DELETE FROM com.model.ElectricityLedgerEntity a
Where a.elLedgerid IN
(SELECT MAX(b.elLedgerid)
FROM com.model.ElectricityLedgerEntity b
WHERE b.accountId='24' and b.ledgerType='Electricity Ledger' and
b.postType='ARREARS')
Edit for bypassing mysql subquery limitations :
The new error java.sql.SQLException: You can't specify target table 'LEDGER' for update in FROM clause
is known in mysql when you use it with JPA. It's one MySQL limitation.
A recent stackoverflow question about it
In brief, you cannot "directly" updated/deleted a table that you query in a select clause
Now I understand why your original query did multiple subqueries seemingly not necessary (while it was useful for mysql) and had a "special" syntax.
I don't know tricks to solve this problem in JPA (I don't use the MySQL DBMS for a long time now).
At your place, I would do two queries. The first where you select the expected max elLedgerid and the second where you could delete line(s) with the id retrieved in the previous query.
You should not have performance issues if your sql model is well designed, the sql indexes well placed and the time to access to the database is correct.

You cannot do this in a single query with Hibernate. If you want to delete the max row(s) with Hibernate you will have to do so in two steps. First, you can find the max entry, then you can delete using that value in the WHERE clause.
But the query you wrote should actually run as a raw MySQL query. So why don't you try executing that query as a raw query:
String sql = "DELETE FROM com.model.ElectricityLedgerEntity a " +
"WHERE a.elLedgerid IN (SELECT P.elLedgerid FROM " +
"(SELECT MAX(b.elLedgerid) FROM com.model.ElectricityLedgerEntity b " +
"WHERE b.accountId = :account_id AND b.ledgerType = :ledger_type AND " +
" b.postType = :post_type) P );";
Query query = session.createSQLQuery(sql);
query.setParameter("account_id", "24");
query.setParameter("ledger_type", "Electricity Ledger");
query.setParameter("post_type", "ARREARS");

Just want to extend existing answer:
In brief, you cannot "directly" updated/deleted a table that you query in a select clause
This was lifted with starting from MariaDB 10.3.1:
Same Source and Target Table
Until MariaDB 10.3.1, deleting from a table with the same source and target was not possible. From MariaDB 10.3.1, this is now possible. For example:
DELETE FROM t1 WHERE c1 IN (SELECT b.c1 FROM t1 b WHERE b.c2=0);

PostgreSQL query works in pgAdmin but not in Java EclipseLink

I have a SQL that counts rows per date truncate (months, days, hours). Think a history graph. Query works fine if executed in pgAdmin but fails in Java using EclipseLink.
pgAdmin query:
SELECT date_trunc( 'hour', delivered_at ),
COUNT(date_trunc( 'hour', delivered_at )) AS num
FROM messages
WHERE channel_type='EMAIL'
AND created_at>='2016-02-28 16:01:08.882'
AND created_at<='2016-02-29 16:01:08.882'
GROUP BY date_trunc( 'hour', delivered_at );
JPQL Named query
SELECT FUNCTION('date_trunc', 'hour', m.deliveredAt ),
COUNT(FUNCTION('date_trunc', 'hour', m.deliveredAt )) AS num
FROM Message m
WHERE m.channelType = :channelType
AND m.createdAt >= :fromDate
AND m.createdAt <= :toDate
GROUP BY FUNCTION('date_trunc', 'hour', m.deliveredAt )
EclipseLink debugging log:
SELECT date_trunc(?, delivered_at), COUNT(date_trunc(?, delivered_at)) FROM messages
WHERE (((channel_type = ?) AND (created_at >= ?)) AND (created_at <= ?)) GROUP BY date_trunc(?, delivered_at)
bind => [hour, hour, EMAIL, 2015-12-27 00:00:00.0, 2015-12-27 00:00:00.0, hour]
Error:
ERROR: column "messages.delivered_at" must appear in the GROUP BY
clause or be used in an aggregate function Position: 23
PostgreSQL log:
2016-03-01 13:22:08 CET ERROR: column "messages.delivered_at" must
appear in the GROUP BY clause or be used in an aggregate function at
character 23 2016-03-01 13:22:08 CET STATEMENT: SELECT date_trunc($1,
delivered_at), COUNT(delivered_at) FROM messages WHERE (((channel_type
= $2) AND (created_at >= $3)) AND (created_at <= $4)) GROUP BY date_trunc($5, delivered_at) 2016-03-01 13:22:08 CET LOG: execute
S_2: SELECT 1
If I replace the binded variables from EclipseLink logged query and execute it in pgAdmin the query works. What is going on here?
Edit: Just to clarify, it also works using em.createNativeQuery.

PostgreSQL can have trouble with resolving parameter binding, which in this case manifests as native SQL with parameters inline work, while JPA generated SQL which defaults to bound parameters fails.
One solution is to turn off parameter binding by passing "eclipselink.jdbc.bind-parameters" with a value of "false" as either a query hint for the specific query, or as a persistence unit property to turn off parameter binding by default for all queries.

On postgresql, you can use the following syntax
GROUP BY 1
That means you will group you agregate using the first selected attribute in the SELECT clause. That might be helpful there

HQL count distinct lower syntax

I want to lowercase all users usernames, then count how many distinct ones there are, but getting an error on this hql query:
select count(distinct lower(user.username)) from UserEntity as user
org.hibernate.hql.internal.ast.QuerySyntaxException: expecting CLOSE, found 'user' near line 1, column 29 [select count(distinct lower(user.username))
This works fine:
select count(distinct user.username) from UserEntity as user
But when I add the lower(...) it it fails, any help is much appreciated!

What about:
select count(lower(user.username)) from UserEntity as user group by lower(user.username)?
My guess is that you are messing with aggregate functions and statements: count(disctinct lower(...)) is a [aggregated function] - [statement] - [aggregated function] pattern which HQL doesn't support (or at least doesn't treat in documentation).

Thanks #Serge Ballesta, I ended up just running this as a SQL query instead of a HQL query, and it worked fine.

Java PreparedStatement can't identify placeholders in subquery

When using a Java PreparedStatement, the question-mark placeholders aren't being detected. It would throw an error "The column index is out of range: 1, number of columns: 0" when invoking statementName.setLong(1, 123). My example is from Postgres 8.4, but the problem occurs before the SQL has a chance to make it to the SQL server.
After comparing against some working prepared statements, I realized that the broken one contained a subquery similar to:
SELECT * FROM (
SELECT DISTINCT (name)
id,
name
FROM MyTable
WHERE id > ?
ORDER BY name) AS Level1
ORDER BY 1

The solution that worked for me was to convert the query to a CTE (Common Table Expression). The revised query looks like this:
WITH Level1 AS (
SELECT DISTINCT (name)
id,
name
FROM MyTable
WHERE id > ?
ORDER BY name)
SELECT *
FROM Level1
ORDER BY 1

In JDBC, the parameter indexes for prepared statements begin at 1 instead of 0.