How do you select min(string_col) using a JPA Criteria query? - java

JPA Criteria seems to insist that min() (and max()) are only applicable to Numeric fields. However, this is NOT the case as SQL database can indeed select min() and max() values from a string column. I can also use JQL to select the min() or max() value of a string column.
However, when using Criteria queries, the compiler complains when min() is used with a string column:
Bound mismatch: The generic method min(Expression) of type CriteriaBuilder is not applicable for the arguments (subquery.get("string_column")). The inferred type String is not a valid substitute for the bounded parameter
This query works:
"select new com.epsilon.totalfact.web.api.model.UPCDetails" +
"(u.id, d.name as dimension, c.name as category, u.name, u.description," +
" u.keywords, u.userNotes, u.restrictedResellerFlag, u.hidden," +
" (select min(p.rateId) from u.platforms p)" +
" || case when (select min(p.rateId) from u.platforms p) != (select max(p.rateId) from u.platforms p) then concat(' ~ ',(select max(p.rateId) from u.platforms p)) else '' end" +
" as deRateId" +
" )" +
" from UPC u " +
" join u.category c" +
" join c.dimension d" +
" order by d.name, c.name, u.name",
and this criteria query has the above mentioned error:
CriteriaBuilder cb = entityManager.getCriteriaBuilder();
CriteriaQuery<UPCDetails> query = cb.createQuery(UPCDetails.class);
Root<UPC> u = query.from(UPC.class);
Join<UPC, Category> c = u.join("category");
Join<UPC, Dimension> d = c.join("dimension");
Subquery<String> sub = query.subquery(Number.class);
Root<UPC> subRoot = sub.from(UPC.class);
SetJoin<UPC,Platform> subPlatform = subRoot.join(UPC_.platforms);
sub.select(cb.min(subPlatform.get("rateId")));

mathguy provided the answer, you need to use least() and greatest() instead of min() and max() for strings and other types.

Related

JPA parameter outside a where clause

I would like to know if the is a way to build a JPA query with parameter outside of the where clause. This query works fine in my database manager.
There is my query :
#Query(value = "SELECT q.quote, q.author, q.display_at, count(ql.*) AS like, (SELECT '{:userUUID}'::uuid[] && ARRAY_AGG(ql.user_uuid)::uuid[]) AS liked " +
"FROM quotes q " +
"LEFT JOIN quotes_likes ql ON ql.quote_uuid = q.uuid " +
"WHERE display_at = :date " +
"GROUP BY q.quote, q.author, q.display_at;", nativeQuery = true)
Optional<QuoteOfTheDay> getQuoteOfTheDay(UUID userUUID, LocalDate date);
I have the following error when the query is called : ERROR: syntax error at or near ":"
By default, Spring Data JPA uses position-based parameter binding. We can also use named parameter with #Param annotation to give a method parameter a concrete name and bind the name in the query. As you are trying to use the named parameter try using the below snippet with #Param annotations.
Optional<QuoteOfTheDay> getQuoteOfTheDay(#Param("userUUID") UUID userUUID, #Param("date") LocalDate date);
ERROR: syntax error at or near means that you need to escape casting colons.
Every : needs to be replaced by \\:
(SELECT ARRAY[:userUUID]'\\:\\:uuid[] && ARRAY_AGG(ql.user_uuid)\\:\\:uuid[]) AS liked
OR use double colons
(SELECT ARRAY[:userUUID]::::uuid[] && ARRAY_AGG(ql.user_uuid)::::uuid[]) AS liked
OR use Cast instead of colons
(SELECT cast(ARRAY[:userUUID] as uuid[]) && cast(ARRAY_AGG(ql.user_uuid) as uuid[])) AS liked
The second problem is that you need to create an array with a query parameter.
Try ARRAY[:userUUID] instead of {:userUUID}
Full query example:
#Query(value = "SELECT q.quote, q.author, q.display_at, count(ql.*) AS like, (SELECT ARRAY[:userUUID]\\:\\:uuid[] && ARRAY_AGG(ql.user_uuid)\\:\\:uuid[]) AS liked " +
"FROM quotes q " +
"LEFT JOIN quotes_likes ql ON ql.quote_uuid = q.uuid " +
"WHERE display_at = :date " +
"GROUP BY q.quote, q.author, q.display_at;", nativeQuery = true)
Optional<QuoteOfTheDay> getQuoteOfTheDay(UUID userUUID, LocalDate date);
As you are trying to pass method parameters to the query using named parameters, We define these using the #Param annotation inside our repository method declaration.
Each parameter annotated with #Param must have a value string matching the corresponding JPQL or SQL query parameter name.
#Query(value = "SELECT q.quote, q.author, q.display_at, count(ql.*) AS like, (SELECT '{:userUUID}'::uuid[] && ARRAY_AGG(ql.user_uuid)::uuid[]) AS liked " +
"FROM quotes q " +
"LEFT JOIN quotes_likes ql ON ql.quote_uuid = q.uuid " +
"WHERE display_at = :date " +
"GROUP BY q.quote, q.author, q.display_at;", nativeQuery = true)
Optional<QuoteOfTheDay> getQuoteOfTheDay(#Param("userUUID") UUID userUUID,#Param("date") LocalDate date);
Refer this for more details on how to use the #Query annotation.

Oracle - How to use START WITH in loop (hibernate select)

I am using hibernate for this select:
NativeQuery sqlQuery = session.createSQLQuery("select :id, min(a.time) " +
"from table1 a, (" +
" select parentid" +
" from (" +
" select LEVEL, parentid" +
" from table2 " +
" START WITH id = :id" +
" CONNECT BY NOCYCLE PRIOR parentid = id" +
" order by level desc)" +
" where rownum = 1" +
" ) b " +
"where a.id = b.parentid");
sqlQuery.setParameter("id", id);
List<Object[]> list = sqlQuery.list();
I need to use this for a lot of ids. Ideally I would pass a list of ids, run that in oracle and return the result. I cannot do that because of the START WITH clause.
This way I send a query to database for each id, which is really slow.
Is there any way to put a list of ids to the query, do this in loop and return back to the application with a list of results?
NativeQuery sqlQuery = session.createSQLQuery("...?...");
sqlQuery.setListParameter("ids", ids);
List<Object[]> list = sqlQuery.list();
Edit: I cannot use recursive with, because we use Oracle 10.
I don't know about the Oracle specific syntax, but with the SQL standard WITH RECURSIVE syntax, you can of course list multiple parameters. Just use the IN predicate e.g. for batches of 5 id IN (:id1, :id2, :id3, :id4, :id5) and then bind the values with setParameter("id1", ...). If you are interested, you can also formulate this with Blaze-Persistence on top of Hibernate by using the JPA model: https://persistence.blazebit.com/documentation/1.6/core/manual/en_US/#recursive-ctes
In your case this would look like the following:
#CTE
#Entity
public class ResultCte {
#Id
Long parentId;
Long rootId;
int level;
}
CriteriaBuilder<Tuple> cb = criteriaBuilderFactory.create(entityManager, Tuple.class);
cb.withRecursive(ResultCte.class)
.from(Entity2.class, "e2")
.where("e2.id").in(idList)
.bind("parentId").select("e2.parent.id")
.bind("rootId").select("e2.id")
.bind("level").select("1")
.unionAll()
.from(Entity2.class, "e2")
.from(ResultCte.class, "cte")
.where("e2.id").eqExpression("cte.parentId")
.bind("parentId").select("e2.parent.id")
.bind("rootId").select("cte.rootId")
.bind("level").select("cte.level + 1")
.end()
.from(Entity1.class "a")
.from(ResultCte.class, "cte")
.where("a.id").eqExpression("cte.parentId")
.select("cte.rootId")
.select("min(a.time)")

Spring Data paginate and sort a NamedNativeQuery (JPA-Hibernate-MySql)

I'm using NamedNativeQueries with SqlResultSetMappings in a Spring Data (JPA Hibernate MySQL) application, and I've been successful with the Pagination, but not with the sorting.
I've tried two forms of queries:
#NamedNativeQuery(
name = "DatasetDetails.unallocatedDetailsInDataset",
resultClass = DatasetDetails.class,
resultSetMapping = "DatasetDetails.detailsForAllocation",
query = "SELECT dd.id, fk_datasets_id, fk_domains_id, fk_sources_id, dom.name AS domain, " +
"src.name AS source " +
"FROM datasets AS d " +
"JOIN datasets_details AS dd ON dd.fk_datasets_id = d.id " +
"JOIN sources AS src ON src.id = dd.fk_sources_id " +
"JOIN domains AS dom ON dom.id = dd.fk_domains_id " +
"WHERE fk_datasets_id = :datasetId " +
"AND dd.id NOT IN (" +
"SELECT fk_datasets_details_id from allocations_datasets_details) \n/* #page */\n"),
and the second is simply using the count notation on a second query instead of using the #page notation.
#NamedNativeQuery(
name = "DatasetDetails.unallocatedDetailsInDataset.count",
resultClass = DatasetDetails.class,
resultSetMapping = "DatasetDetails.detailsForAllocation",
query = "SELECT count(*)
....
Both methods work for pagination, but the sorting is ignored.
Here is the repository:
public interface DatasetDetailsRepository extends PagingAndSortingRepository<DatasetDetails, Long> {
#Query(nativeQuery = true)
List<DatasetDetails> unallocatedDetailsInDataset(#Param("datasetId") long datasetId,
#Param("page") Pageable page);
}
And the pageable gets assembled like this:
Sort sort = Sort.by(Sort.Order.asc(DatasetDetails.DOMAIN), Sort.Order.asc(DatasetDetails.SOURCE));
Pageable page = PageRequest.of(page, limit, sort);
No errors are thrown, but the sorting simply doesn't get done and no ORDER BY is generated.
Explicitly adding something like ORDER BY #{#page} won't compile.
I encountered the same problem, where I had to dynamically filter/sort using a NamedNativeQuery by different columns and directions; apparently the Sorting was ignored. I found this workaround, which is not necessarily nice but it does the job:
For the repository:
List<MyEntity> findMyEntities(
#Param("entityId") long entityId,
#Param("sortColumn") String sortColumn,
#Param("sortDirection") String sortDirection,
Pageable page);
The native queries look like this:
#NamedNativeQueries({
#NamedNativeQuery(name = "MyEntity.findMyEntities",
query = "select e.field1, e.field2, ..." +
" from my_schema.my_entities e" +
" where condition1 and condtition2 ..." +
" order by " +
" CASE WHEN :sortColumn = 'name' and :sortDirection = 'asc' THEN e.name END ASC," +
" CASE WHEN :sortColumn = 'birthdate' and :sortDirection = 'asc' THEN e.birthdate END ASC," +
" CASE WHEN :sortColumn = 'name' and :sortDirection = 'desc' THEN e.name END DESC," +
" CASE WHEN :sortColumn = 'birthdate' and :sortDirection = 'desc' THEN e.birthdate END DESC" +
),
#NamedNativeQuery(name = "MyEntity.findMyEntities.count",
query = "select count(*) from my_schema.my_entities e" +
" where condition1 and condtition2 ..." +
" and :sortColumn = :sortColumn and :sortDirection = :sortDirection"
)
})
Notice in the count query I use the 2 redundant conditions for :sortColumn and :sortDirection, because once specified as #Param in the repository function, you need to use them in the actual query.
When calling the function, in my service I had a boolean which dictates the direction and a string that dictates the sorting column like this:
public Page<MyEntity> serviceFindFunction(Long entityId, String sortColumn, Boolean sortDirection, Integer pageNumber, Integer pageSize) {
String sortDir = (sortDirection) ? 'asc' : 'desc';
Pageable pageable = new PageRequest(pageNumber, pageSize); // Spring Data 1.0 syntax
// for Spring Data 2.0, as you were using, simply:
// Pageable pageable = PageRequest.of(pageNumber, pageSize);
return entityRepository.findMyEntities(entityId, sortColumn, sortDir, pageable)
}
The 2 things that I don't like about this are the redundant usage of the sortColumn and sortDirection params in the count query, and the way I wrote the order by statement. The reasoning for having separate CASE statements is because I had different data types for the columns that I sorted by, and if they are incompatible (e.g. nvarchar and date), the query will fail with the error:
Conversion failed when converting date and/or time from character string
I could also probably nest the conditionals, i.e. first making a case for the direction, the making an inner case for the columns, but my SQL skills only went this far.
Hope this helps! Any feedback or improvements are very welcomed.

Spring jpa native query sorting adding prefix to order by field name

I have problem with sorting.
Repository method:
#Query(nativeQuery = true,
value = "SELECT D.ID as dealerId , D.NAME as dealerName, K.ID as kpiId, " +
"K.NAME as kpiName FROM REGION R, DEALER D, KPI K "
+ "WHERE R.IMPORTER_ID = :importerId "
+ "AND D.REGION_ID = R.ID "
+ "AND K.IMPORTER_ID = :importerId ")
Page<DealersKpiTargets> getDealersKpiTargets(#Param("importerId") Long importerId, Pageable pageable);
Pageable object:
Page request [number: 0, size 20, sort: name: DESC]
Hibernate log:
Hibernate: SELECT D.ID as dealerId , D.NAME as dealerName, K.ID as kpiId, K.NAME as kpiName FROM REGION R, DEALER D, KPI K WHERE R.IMPORTER_ID = ? AND D.REGION_ID = R.ID AND K.IMPORTER_ID = ? order by R.name desc limit ?
I don't understand where R.name prefix came from, in the order by part in Hibernate (towards the end).
For reference, I am using:
spring-data-jpa version 2.0.7.RELEASE
spring-boot-starter-data-jpa version 2.0.2.RELEASE
UPDATE
I have solved this by changing the query from the native query to jpa query and it's working. And I changed cartesian to join version.
#Query("SELECT dealerEntity.id AS dealerId , dealerEntity.name AS dealerName, kpiEntity.id AS kpiId, " +
"kpiEntity.name AS kpiName FROM KpiEntity kpiEntity "
+ "JOIN RegionEntity regionEntity ON regionEntity.importerEntity = kpiEntity.importerEntity "
+ "JOIN DealerEntity dealerEntity ON dealerEntity.importerEntity = regionEntity.importerEntity "
+ "WHERE kpiEntity.importerEntity = :importerEntity ")
Page<DealersKpiTargets> getDealersKpiTargets(#Param("importerEntity") ImporterEntity importerEntity, Pageable pageable);
here is jira ticket with more details which can be key for resolution (https://jira.spring.io/browse/DATAJPA-1613).
QueryUtils.ALIAS_MATCH
(?<=from)(?:\s)+([._[\P\\{Z}&&\P\\{Cc}&&\P\\{Cf}&&\P\\{P}]]+)(?:\sas)*(?:\s)+(?!(?:where|group\s*by|order\s*by))(\w+)
responsible to incorrect alias extraction. The solution for my case was rewrite native query, so it doesn't match the provided regexp.
This may be a little late to answer this question. But thought to share how I got around this issue.
For native queries, it seems like hibernate tries to use the alias of the first table used in the query when it applies the sorting criteria. In your case, the first table alias is R hence you see R.name desc in the query generated by hibernate.
One way to get around this issue is to wrap your query in a select clause and name it as R, like
"SELECT * FROM(SELECT D.ID as dealerId , D.NAME as dealerName, K.ID as kpiId, " +
"K.NAME as kpiName FROM REGION R, DEALER D, KPI K "
+ "WHERE R.IMPORTER_ID = :importerId "
+ "AND D.REGION_ID = R.ID "
+ "AND K.IMPORTER_ID = :importerId ) R"
This way at runtime hibernate would apply the sort criteria on top of your query which corresponds to R now.
It has Sort class for this you can use this maybe. Besides, it is easy to use.
https://docs.spring.io/spring-data/jpa/docs/current/reference/html/#jpa.query-methods.sorting
I faced similar issue especially in case of complex queries where there is ORDER BY with in the query. I was getting syntax error because a , was getting added before ORDER BY.
The way I solved this issue was to create a VIEW with the SELECT query having necessary fields required for result set and WHERE condition (so you can run query with params in WHERE condition against the VIEW). And write native query to SELECT FROM the VIEW
CREATE VIEW my_view AS (// your complex select query with required fields);
#Query("SELECT field1 AS alias1, field2 AS alias2 FROM my_view "
+ "WHERE field3 = :param1 AND field4 = :param2")
Page<MyDto> myFunction(#Param("param1") Long param1, #Param("param1") String param2, Pageable pageable);

How to map Results when Querying raw SQL using Ebean

Using Postgres Tables created by Ebean, I would like to query these tables with a hand-written statement:
SELECT r.name,
r.value,
p.name as param1,
a.name as att1,
p2.name as param2,
a2.name as att2
FROM compatibility c
JOIN attribute a ON c.att1_id = a.id
JOIN attribute a2 ON c.att2_id = a2.id
JOIN PARAMETER p ON a.parameter_id = p.id
JOIN PARAMETER p2 ON a2.parameter_id = p2.id
JOIN rating r ON c.rating_id = r.id
WHERE p.problem_id = %d
OR p2.problem_id = %d
Each of the joined tables represent one of my model classes.
The query executes fine, but I don't know how I would proceed:
How do I even execute the query using Play 2.2. and Ebean?
How can I map this query to an iterable object? Do I need to create a Model class which contains all the fields from the query, or can I use some sort of HashMap?
How can I parameterize the query in a safe way?
To execute this query you need to use RawSql class. You will also have to create class to which results will be casted.
Here is a code of exemplary result class:
import javax.persistence.Entity;
import com.avaje.ebean.annotation.Sql;
#Entity
#Sql
public class Result {
String name;
Integer value;
String param1;
String param2;
String att1;
String att2;
}
And example of executing this query:
String sql
= " SELECT r.name,"
+ " r.value,"
+ " p.name as param1,"
+ " a.name as att1,"
+ " p2.name as param2,"
+ " a2.name as att2"
+ " FROM compatibility c"
+ " JOIN attribute a ON c.att1_id = a.id"
+ " JOIN attribute a2 ON c.att2_id = a2.id"
+ " JOIN PARAMETER p ON a.parameter_id = p.id"
+ " JOIN PARAMETER p2 ON a2.parameter_id = p2.id"
+ " JOIN rating r ON c.rating_id = r.id"
+ " WHERE p.problem_id = %d"
+ " OR p2.problem_id = %d"
RawSql rawSql =
RawSqlBuilder
.parse(sql)
.columnMapping("r.name", "name")
.columnMapping("r.value", "value")
.create();
Query<Result> query = Ebean.find(Result.class);
query.setRawSql(rawSql)
.where().gt("amount", 10);
List<Result> list = query.findList();

Categories