Coalesce jsonArrayAgg to empty array in jOOQ - java

What is the equivalent for coalesce in the new jOOQ 3.14 SQL/JSON supporting version (in PostgreSQL)?
select coalesce(json_agg(t.*), '[]'::json)
from (select 'test' as mycol where 1 = 2) t;
The following unfortunately fails with the error "COALESCE types json and uuid[] cannot be matched".
coalesce(jsonArrayAgg(mycol), emptyArray<type>())

JSON.json() or JSONB.jsonb(), e.g.:
coalesce(jsonArrayAgg(mycol), inline(json("[]")))
coalesce(jsonbArrayAgg(mycol), inline(jsonb("[]")))
But you can also use the jsonArray() or jsonbArray() constructors:
coalesce(jsonArrayAgg(mycol), jsonArray())
coalesce(jsonbArrayAgg(mycol), jsonbArray())

Related

Union query with multiple selects post java 8

Here is a query that I want to try out in MySQL
SELECT A.x
FROM A
WHERE A.y = 'P'
UNION
SELECT A.x
FROM A
WHERE A.y = 'Q'
The above is a cut-down, much simpler version of the original query that I am trying. In my original query, each SELECT statement involves multiple tables with INNER JOIN
If the possible number of values in 'y' column of table 'A' that I need to query upon is 'n', then my query will involve doing 'n-1' unions on 'n' SELECT statements
I know that JOOQ can do union of multiple SELECT statements. But is there a good way to do this post Java 8 style? maybe using Steam.collect()?
This is what I have but wondering if I could do better
String firstValueToQuery = valuesToQuery.get(0);
Select<Record5<UUID, UUID, String, Integer, String>> selectQuery = getSelectQueryForValue(firstValueToQuery);
valuesToQuery.stream()
.skip(1)
.forEach(valueToQuery -> selectQuery.unionAll(getSelectQueryForValue(valueToQuery)));
selectQuery.fetchStream();
Here is how I implement getSelectQueryForValue
private Select<Record5<UUID, UUID, String, Integer, String>> getSelectQueryForValue(String valueToQuery) {
return jooq.select(
A.P,
A.Q,
A.R,
A.S,
A.T)
.from(A)
.where(A.Y.eq(valueToQuery));
}
PS: I understand that I could rather use the 'IN' clause like below
SELECT A.x
FROM A
WHERE A.y IN ('P','Q',...)
But with my current data distribution in the database, MySQL is using a sub-optimal query plan. Thus using UNION so that the database implicitly prefers a faster query plan by making use of the right index
The idiomatic approach here would be as follows (using JDK 9 API):
try (Stream<Record5<UUID, UUID, String, Integer, String>> stream = valuesToQuery
.stream()
.map(this::getSelectQueryForValue)
.reduce(Select::union)
.stream() // JDK 9 method
.flatMap(Select::fetchStream)) {
...
}
It uses the useful Optional.stream() method, which was added in JDK 9. In JDK 8, you could do this instead:
valuesToQuery
.stream()
.map(this::getSelectQueryForValue)
.reduce(Select::union)
.ifPresent(s -> {
try (Stream<Record5<UUID, UUID, String, Integer, String>> stream =
s.fetchStream()) {
...
}
})
I blogged about this in more detail here.

JOOQ - Select count inside select query

I have a problem while converting the following statement into jooq API:
SELECT t1.col1, t1.col2, t1.col3, (SELECT count(*) FROM table2 where table2.col2 = t1.col1)
FROM table1 t1
I tried it with DSL.count() and DSL.selectCount() but I failed while searching a way to add the where clause to the count subquery.
The database is PostgreSQL 9.6.
Lukas suggestion to use DSL.field is the better solution because it preserves the <T> type.
More typesafe version:
TableField<Table1Record, Long> col1 = TABLE1.COL1;
Field<Integer> count = DSL.field(DSL.selectCount().from(TABLE2).where(TABLE2.COL2.eq(col1)));
using(configuration).select(col1, count).from(TABLE1).fetch();
My first (less typesafe) solution:
TableField<Table1Record, Long> col1 = TABLE1.COL1;
Field count = DSL.selectCount().from(TABLE2).where(TABLE2.COL2.eq(col1)).asField("count");
using(configuration).select(col1, count).from(TABLE1).fetch();
Maybe there is a more elegant solution, but it works. The generated query looks like my original query.
Here is another example using DSL.field(...):
Field<Integer> COUNT = DSL.field("COUNT(*) OVER ()", Integer.class);
List<Map<String, Object>> records = DSL.select(ACCESSORY.ID,
ACCESSORY.NAME,
ACCESSORY.TYPE,
ACCESSORY.PRICE,
BRAND.NAME,
COUNT.as("total"))
.from(ACCESSORY)
.innerJoin(BRAND).onKey()
.fetchMaps();
The ResultSet will contain a column called total which will be treated as a type java.lang.Integer. This works with PostgreSQL 9.6.
A detailed description of COUNT(*) OVER () can be found here: here.

createNativeQuery will not return rows in java but will in db2 sql [duplicate]

I know I can pass a list to named query in JPA, but how about NamedNativeQuery? I have tried many ways but still can't just pass the list to a NamedNativeQuery. Anyone know how to pass a list to the in clause in NamedNativeQuery? Thank you very much!
The NamedNativeQuery is as below:
#NamedNativeQuery(
name="User.findByUserIdList",
query="select u.user_id, u.dob, u.name, u.sex, u.address from user u "+
"where u.user_id in (?userIdList)"
)
and it is called like this:
List<Object[]> userList = em.createNamedQuery("User.findByUserIdList").setParameter("userIdList", list).getResultList();
However the result is not as I expected.
System.out.println(userList.size()); //output 1
Object[] user = userList.get(0);
System.out.println(user.length); //expected 5 but result is 3
System.out.println(user[0]); //output MDAVERSION which is not a user_id
System.out.println(user[1]); //output 5
System.out.println(user[2]); //output 7
The above accepted answer is not correct and led me off track for many days !!
JPA and Hibernate both accept collections in native query using Query.
You just need to do
String nativeQuery = "Select * from A where name in :names"; //use (:names) for older versions of hibernate
Query q = em.createNativeQuery(nativeQuery);
q.setParameter("names", l);
Also refer the answers here which suggest the same (I picked the above example from one of them)
Reference 1
Reference 2 which mentioned which cases paranthesis works which giving the list as a parameter
*note that these references are about jpql queries, nevertheless the usage of collections is working with native queries too.
A list is not a valid parameter for a native SQL query, as it cannot be bound in JDBC. You need to have a parameter for each argument in the list.
where u.user_id in (?id1, ?id2)
This is supported through JPQL, but not SQL, so you could use JPQL instead of a native query.
Some JPA providers may support this, so you may want to log a bug with your provider.
Depending on your database/provider/driver/etc., you can, in fact, pass a list in as a bound parameter to a JPA native query.
For example, with Postgres and EclipseLink, the following works (returning true), demonstrating multidimensional arrays and how to get an array of double precision. (Do SELECT pg_type.* FROM pg_catalog.pg_type for other types; probably the ones with _, but strip it off before using it.)
Array test = entityManager.unwrap(Connection.class).createArrayOf("float8", new Double[][] { { 1.0, 2.5 }, { 4.1, 5.0 } });
Object result = entityManager.createNativeQuery("SELECT ARRAY[[CAST(1.0 as double precision), 2.5],[4.1, 5.0]] = ?").setParameter(1, test).getSingleResult();
The cast is there so the literal array is of doubles rather than numeric.
More to the point of the question - I don't know how or if you can do named queries; I think it depends, maybe. But I think following would work for the Array stuff.
Array list = entityManager.unwrap(Connection.class).createArrayOf("int8", arrayOfUserIds);
List<Object[]> userList = entityManager.createNativeQuery("select u.* from user u "+
"where u.user_id = ANY(?)")
.setParameter(1, list)
.getResultList();
I don't have the same schema as OP, so I haven't checked this exactly, but I think it should work - again, at least on Postgres & EclipseLink.
Also, the key was found in: http://tonaconsulting.com/postgres-and-multi-dimensions-arrays-in-jdbc/
Using hibernate, JPA 2.1 and deltaspike data I could pass a list as parameter in query that contains IN clause. my query is below.
#Query(value = "SELECT DISTINCT r.* FROM EVENT AS r JOIN EVENT AS t on r.COR_UUID = t.COR_UUID where " +
"r.eventType='Creation' and t.eventType = 'Reception' and r.EVENT_UUID in ?1", isNative = true)
public List<EventT> findDeliveredCreatedEvents(List<String> eventIds);
can be as simple as:
#Query(nativeQuery =true,value = "SELECT * FROM Employee as e WHERE e.employeeName IN (:names)")
List<Employee> findByEmployeeName(#Param("names") List<String> names);
currently I use JPA 2.1 with Hibernate
I also use IN condition with native query. Example of my query
SELECT ... WHERE table_name.id IN (?1)
I noticed that it's impossible to pass String like "id_1, id_2, id_3" because of limitations described by James
But when you use jpa 2.1 + hibernate it's possible to pass List of string values. For my case next code is valid:
List<String> idList = new ArrayList<>();
idList.add("344710");
idList.add("574477");
idList.add("508290");
query.setParameter(1, idList);
In my case ( EclipseLink , PostGreSQL ) this works :
ServerSession serverSession = this.entityManager.unwrap(ServerSession.class);
Accessor accessor = serverSession.getAccessor();
accessor.reestablishConnection(serverSession);
BigDecimal result;
try {
Array jiraIssues = accessor.getConnection().createArrayOf("numeric", mandayWorkLogQueryModel.getJiraIssues().toArray());
Query nativeQuery = this.entityManager.createNativeQuery(projectMandayWorkLogQueryProvider.provide(mandayWorkLogQueryModel));
nativeQuery.setParameter(1,mandayWorkLogQueryModel.getPsymbol());
nativeQuery.setParameter(2,jiraIssues);
nativeQuery.setParameter(3,mandayWorkLogQueryModel.getFrom());
nativeQuery.setParameter(4,mandayWorkLogQueryModel.getTo());
result = (BigDecimal) nativeQuery.getSingleResult();
} catch (Exception e) {
throw new DataAccessException(e);
}
return result;
Also in query cannot use IN(?) because you will get error like :
Caused by: org.postgresql.util.PSQLException: ERROR: operator does not exist: numeric = numeric[]
'IN(?)' must be swapped to '= ANY(?)'
My solution was based on Erhannis concept.
In jpa, it worked for me
#Query(nativeQuery =true,value = "SELECT * FROM Employee as e WHERE e.employeeName IN (:names)")
List<Employee> findByEmployeeName(#Param("names") List<String> names);
Tried in JPA2 with Hibernate as provider and it seems hibernate does support taking in a list for "IN" and it works. (At least for named queries and I believe it will be similar with named NATIVE queries)
What hibernate does internally is generate dynamic parameters, inside the IN same as the number of elements in the passed in list.
So in you example above
List<Object[]> userList = em.createNamedQuery("User.findByUserIdList").setParameter("userIdList", list).getResultList();
If list has 2 elements the query will look like
select u.user_id, u.dob, u.name, u.sex, u.address from user u "+
"where u.user_id in (?, ?)
and if it has 3 elements it looks like
select u.user_id, u.dob, u.name, u.sex, u.address from user u "+
"where u.user_id in (?, ?, ?)
you should do this:
String userIds ="1,2,3,4,5";
List<String> userIdList= Stream.of(userIds.split(",")).collect(Collectors.toList());
Then, passes like parameter inside your query, like this:
#NamedNativeQuery(name="User.findByUserIdList", query="select u.user_id, u.dob, u.name, u.sex, u.address from user u where u.user_id in (?userIdList)")
It's not possible with standard JPA. Hibernate offers the proprietary method setParameterList(), but it only works with Hibernate sessions and is not available in JPA's EntityManager.
I came up with the following workaround for Hibernate, which is not ideal but almost standard JPA code and has some nice properties to it.
For starters you can keep the named native query nicely separated in a orm.xml file:
<named-native-query name="Item.FIND_BY_COLORS" result-class="com.example.Item">
<query>
SELECT i.*
FROM item i
WHERE i.color IN ('blue',':colors')
AND i.shape = :shape
</query>
</named-native-query>
The placeholder is wrapped in single quotes, so it's a valid native JPA query. It runs without setting a parameter list and would still return correct results when other matching color parameters are set around it.
Set the parameter list in your DAO or repository class:
#SuppressWarnings("unchecked")
public List<Item> findByColors(List<String> colors) {
String sql = getQueryString(Item.FIND_BY_COLORS, Item.class);
sql = setParameterList(sql, "colors", colors);
return entityManager
.createNativeQuery(sql, Item.class)
.setParameter("shape", 'BOX')
.getResultList();
}
No manual construction of query strings. You can set any other parameter as you normally would.
Helper methods:
String setParameterList(String sql, String name, Collection<String> values) {
return sql.replaceFirst(":" + name, String.join("','", values));
}
String getQueryString(String queryName, Class<?> resultClass) {
return entityManager
.createNamedQuery(queryName, resultClass)
.unwrap(org.hibernate.query.Query.class) // Provider specific
.getQueryString();
}
So basically we're reading a query string from orm.xml, manually set a parameter list and then create the native JPA query. Unfortunately, createNativeQuery().getResultList() returns an untyped query and untyped list even though we passed a result class to it. Hence the #SuppressWarnings("unchecked").
Downside: Unwrapping a query without executing it may be more complicated or impossible for JPA providers other than Hibernate. For example, the following might work for EclipseLink (untested, taken from Can I get the SQL string from a JPA query object?):
Session session = em.unwrap(JpaEntityManager.class).getActiveSession();
DatabaseQuery databaseQuery = query.unwrap(EJBQueryImpl.class).getDatabaseQuery();
databaseQuery.prepareCall(session, new DatabaseRecord());
Record r = databaseQuery.getTranslationRow();
String bound = databaseQuery.getTranslatedSQLString(session, r);
String sqlString = databaseQuery.getSQLString();
An alternative might be to store the query in a text file and add code to read it from there.
You can pass in a list as a parameter, but:
if you create a #NamedNativeQuery and use .createNamedQuery(), you don't use named param, you used ?1(positional parameter). It starts with 1, not 0.
if you use .createNativeQuery(String), you can use named param.
You can try this :userIdList instead of (?userIdList)
#NamedNativeQuery(
name="User.findByUserIdList",
query="select u.user_id, u.dob, u.name, u.sex, u.address from user u "+
"where u.user_id in :userIdList"
)

JOOQ - convert result into Pojo

I have seen that JOOQ can automatically return a POJO when we use .selectFrom(TABLE) or .fetchInto(POJO.class);
But is it possible to convert the result of a complex query into multiple POJO ?
Example :
This query will return an array of all columns into tables Support and Box. It is possible to convert them into a Support and Box Pojo ?
Result<Record> results = query.select()
.from(BOX)
.join(SUPPORT)
.on(SUPPORT.ID.equal(BOX.SUPPORT_ID))
.where(SUPPORT.ID.equal("XXXX"))
.orderBy(BOX.ID)
.fetch();
I have tested the method .intoGroups(SUPPORT.ID, Box.class) , it works fine. But I doesn't have the support object.
Instantiate to SelectSeekStep1
With aliases it's more convenient:
Box b = BOX.as("b");
Support s = SUPPORT.as("s");
SelectSeekStep1<Integer, Integer> sql = query.select(b.ID, s.ID /* other columns */)
.from(b)
.join(s)
.on(s.ID.eq(b.SUPPORT_ID))
.where(s.ID.eq("XXXX"))
.orderBy(b.ID)
;
Then just fetch what/as you need:
List<BoxRecord> boxes = sql.fetchInto(BOX);
SupportRecord support = sql.limit(1).fetchOneInto(SUPPORT);
For future readers, if you want to achieve the same behaviour with insert methods you should use:
insertInto(BOX)
.set(BOX.COLUMN1, UInteger.valueOf(1))
.set(BOX.COLUMN2, "test")
.returning()
.fetchOne()
.into(<POJO_class>.class);

How to use the condition "NULLS LAST" in an order by in a readAllQuery in TopLink

Hi need to make a query in toplink that uses the NULLS LAST clause in an order by. Basically this is the query i'd like to do:
select * from VW_SEGNA_PRZZ_DEP_INFO where ente_cd = '7316' and refr_cd = '000070434' order by dp_dt_timestamp_modifica DESC NULLS LAST;
And this is my code:
ReadAllQuery query = new ReadAllQuery();
query.setReferenceClass(VwSegnaPrzzDepInfo.class);
Expression exp = new ExpressionBuilder();
exp =
exp.and(new ExpressionBuilder().get("enteCd").equal(getEnteCd()));
exp =
exp.and(new ExpressionBuilder().get("refrCd").equal(spdf.getRefrCd()));
query.setSelectionCriteria(exp);
query.addDescendingOrdering("dpDtTimestampModifica");
I don't know how to insert the NULLS LAST clause. I'm using toplink 3.0 an Oracle 9i as a database.
You can use,
query.addOrdering(Expression)
To create the Expression you will need to create your own ExpressionOperator that prints itself as "DESC NULLS LAST" (refer to the desc operator as an example).

Categories