We Keep Coding

How does Hibernate know which member variable to populate when an object has two members of the same class

I've got a MySQL database schema with 3 tables as follows:
mysql> describe results;
+--------------+------------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+--------------+------------------+------+-----+---------+----------------+
| id | int(10) unsigned | NO | PRI | NULL | auto_increment |
| run_at | datetime | YES | | NULL | |
| trials | int(10) unsigned | YES | | NULL | |
+--------------+------------------+------+-----+---------+----------------+
mysql> describe result_details;
+------------------------+------------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+------------------------+------------------+------+-----+---------+----------------+
| id | int(10) unsigned | NO | PRI | NULL | auto_increment |
| results_id | int(10) | NO | | NULL | |
| summarys_id | int(10) | YES | | NULL | |
+------------------------+------------------+------+-----+---------+----------------+
mysql> describe summarys;
+-------+------------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-------+------------------+------+-----+---------+----------------+
| mean | double | YES | | NULL | |
| id | int(10) unsigned | NO | PRI | NULL | auto_increment |
+-------+------------------+------+-----+---------+----------------+
Where a Result object can have several ResultDetail members. However, semantically, it makes sense to have one of these details stand out among the rest as a 'overall' detail. Therefore, I have the following classes:
Result.java (some members and methods removed for brevity)
#Entity
#Table(name="results")
public class Result extends BaseEntity {
#Column(name="run_at")
#Temporal(TemporalType.TIMESTAMP)
private Date runAt;
#Column(name="trials")
private Integer trials;
#OneToOne(mappedBy="result", cascade = {CascadeType.ALL})
private ResultDetails overallStats;
#OneToMany(mappedBy="result", cascade = {CascadeType.ALL})
private List<ResultDetails> resultDetails = new ArrayList<ResultDetails>();
}
ResultDetail.java
#Entity
#Table(name="result_details")
public class ResultDetails extends BaseEntity {
#ManyToOne
#JoinColumn(name = "results_id", nullable=false)
#NotNull
private Result result;
#OneToOne(cascade=CascadeType.ALL)
#JoinColumn(name="summarys_id", nullable=true)
private Summary summary;
}
When I create persistent entities from my main as follows:
public static void main (String [] args) {
Result result = new Result();
ResultDetails detail1 = new ResultDetails();
ResultDetails detail2 = new ResultDetails();
Summary s1 = new Summary();
Summary s2 = new Summary();
result.setRunAt(new Date());
result.setTrials(1000000);
detail1.setResult(result);
s1.setMean(3.0);
detail1.setSummary(s1);
result.setOverallStats(detail1);
detail2.setCybervarResult(result);
s2.setMean(11.0);
detail2.setSummary(s2);
result.addDetails(detail2);
Session session = HibernateUtil.getSessionFactory().openSession();
session.beginTransaction();
session.saveOrUpdate(result);
session.getTransaction().commit();
HibernateUtil.shutdown();
}
It adds the appropriate rows to the tables. However, when I retrieve the data as demonstrated by the following additional main file:
Session session = HibernateUtil.getSessionFactory().openSession();
Result result = session.get(CybervarResult.class, 6);
result.getOverallStats().getSummary();
result.getResultDetails().size();
HibernateUtil.shutdown();
System.out.println(result.getOverallStats().getSummary().getMean());
System.out.println(result.getResultDetails().get(0).getSummary().getMean());
Hibernate is able to correctly populate the 'overallStats' and 'resultDetails' objects. How is it able to differentiate the two rows in the result_details table? As far as I can tell, there is nothing to distinguish the two from each other. Does Hibernate implement hidden tables/rows to manage which member variables correspond to which rows? If I were to create this database from mysql queries instead of through the hibernate API, how would Hibernate know which row should be the 'overallStats' and which rows should belong to the 'resultDetails' collection?
For reference, the rows created look as follows:
mysql> select * from results;
+----+---------------------+---------+
| id | run_at | trials |
+----+---------------------+---------+
| 6 | 2017-11-13 09:27:52 | 1000000 |
+----+---------------------+---------+
mysql> select * from result_details;
+----+------------+-------------+
| id | results_id | summarys_id |
+----+------------+-------------+
| 10 | 6 | 14 |
| 11 | 6 | 15 |
+----+------------+-------------+
hibernate.cfg.xml
<property name="hibernate.dialect">org.hibernate.dialect.MySQLInnoDBDialect</property>
<property name="hibernate.connection.driver_class">com.mysql.jdbc.Driver</property>
<property name="hibernate.connection.url">jdbc:mysql://localhost:3306/usersdb</property>
<property name="hibernate.connection.username">test</property>
<property name="hibernate.connection.password">test</property>
<property name="hibernate.show_sql">true</property>
<property name="hibernate.format_sql">true</property>
<property name="hibernate.hbm2ddl.auto">validate</property>

MySQL before insert trigger not invoked when object saved by hibernate orm

I have a before insert trigger in MySQL that works when row is inserted directly (through DB client) into database but DOES NOT work through Hibernate ORM.
Some basic info
MySQL 5.6.26 on Debian Jessie 8.6
MySQL Connector Java 5.1.40
Hibernate ORM 5.2.4
the Product table
+----------------+---------------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+----------------+---------------------+------+-----+---------+----------------+
| id | int(10) unsigned | NO | PRI | NULL | auto_increment |
| prd_no | varchar(6) | YES | UNI | NULL | |
| prd_ty | varchar(20) | NO | | NULL | |
| prd_name | varchar(50) | NO | | NULL | |
| prd_name_short | varchar(10) | NO | | NULL | |
| prd_cat | varchar(20) | NO | | NULL | |
| prd_tax_ty | varchar(20) | NO | | NULL | |
| prd_price_ty | varchar(20) | NO | | NULL | |
| prd_norm_avail | tinyint(3) unsigned | NO | | 1 | |
| ppp_id | int(10) unsigned | YES | MUL | NULL | |
+----------------+---------------------+------+-----+---------+----------------+
My trigger
DELIMITER $$
CREATE TRIGGER tg_product_insPrdNo BEFORE INSERT ON product
FOR EACH ROW
BEGIN
IF NEW.prd_ty = 'ns' THEN
SET NEW.prd_no = fn_getPrdNoNextAI();
END IF;
END $$
DELIMITER ;
Function that gets called by the trigger
DELIMITER //
CREATE FUNCTION fn_getPrdNoNextAI()
RETURNS INTEGER SIGNED
BEGIN
DECLARE last_prd_no INTEGER SIGNED;
SELECT (MAX(CAST(product.prd_no AS SIGNED))) INTO last_prd_no FROM product;
IF last_prd_no IS NULL THEN
SET last_prd_no = 0;
END IF;
RETURN last_prd_no+1;
END //
DELIMITER ;
Java code (abbreviated) for inserting row
Session session = AES_Server.getSessionFactory().openSession();
Transaction tx = null;
int prdID = -1;
try {
tx = session.beginTransaction();
prdID = (int) session.save(prd);
tx.commit();
} catch (HibernateException he) {
if (tx != null)
tx.rollback();
} finally {
session.close();
}
return prdID;
Basically, my DB trigger will only auto-insert value for prd_no field if prd_ty is of value 'ns'. This works fine when inserting directly into the DB but when using hibernate to insert, no value is inserted.
Any pointers would be much appreciated. Cheers.

HQL left outer join with missing mapping

I am using hibernate/hql to create my querys.
Now I got a problem I'm stuck on for quite some hours now, to understand the situation here is
my sorrounding:
I got 3 Tables I need to get Informations from, with the connection/assignment tables I got 5 overall.
The Informations I need are the Key, the Type and the SourceFile, but the special case here is
that the sql will be done while importing new data, so I want to first check if the data is
already existent, or partially existent.
My query needs to always give me the key, no matter what Type or SourceFile
if the Key itself is already in the Database, also I then only want the key and not the other
Informations. (Key matches but SourceFile and Type do not, so I want only the key back)
If the Key is existent with the exact same Type and SourceFile I want to get all Informations.
The Tables are:
(heads up: FK_K_ID is saved as a object with the name key,
FK_S_ID is saved as Source and FK_T_ID is saved as Type)
Key:
+-------------+--------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-------------+--------------+------+-----+---------+----------------+
| K_ID | bigint(20) | NO | PRI | NULL | auto_increment |
| key | varchar(255) | NO | | NULL | |
| deleted | boolean | NO | | FALSE | |
+-------------+--------------+------+-----+---------+----------------+
KeyType:
+-------------+--------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-------------+--------------+------+-----+---------+----------------+
| KT_ID | bigint(20) | NO | PRI | NULL | auto_increment |
| FK_K_ID | bigint(20) | NO | | NULL | |
| FK_T_ID | bigint(20) | NO | | NULL | |
| deleted | boolean | NO | | FALSE | |
+-------------+--------------+------+-----+---------+----------------+
Type:
+-------------+--------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-------------+--------------+------+-----+---------+----------------+
| T_ID | bigint(20) | NO | PRI | NULL | auto_increment |
| name | varchar(255) | NO | | NULL | |
| description | varchar(255) | NO | | NULL | |
| deleted | boolean | NO | | FALSE | |
+-------------+--------------+------+-----+---------+----------------+
KeySource:
+-------------+--------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-------------+--------------+------+-----+---------+----------------+
| KS_ID | bigint(20) | NO | PRI | NULL | auto_increment |
| FK_K_ID | bigint(20) | NO | | NULL | |
| FK_S_ID | bigint(20) | NO | | NULL | |
| deleted | boolean | NO | | FALSE | |
+-------------+--------------+------+-----+---------+----------------+
Source:
+-------------+--------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-------------+--------------+------+-----+---------+----------------+
| S_ID | bigint(20) | NO | PRI | NULL | auto_increment |
| sourceFile | varchar(255) | NO | | NULL | |
| deleted | boolean | NO | | FALSE | |
+-------------+--------------+------+-----+---------+----------------+
and here is what I tried so far:
from KeyType kt
right outer join kt.Key k
with k.name in (...)
where k.deleted = false
and ( kt.Type in (...) or kt is null )
The problem with this one is, it kind of does what I want. but I get all keys in the database and the KeyType only where it matches, else null. I don't want to get all keys I just want to get my keys I asked for.
from KeyType kt
right outer join kt.Key k
with k.name in (...)
where k.deleted = false
and (k.name in (...) and kt.Type in (...) or kt is null )
I don't even know what I tried here to be honest. I guess I tried to optimize the first query to only give the keys I asked for.
from KeyType kt
right outer join fetch kt.Key k
where k.deleted = false
and (k.name in (...) and kt.Type in (...) or kt is null )
I also tried to use fetch like this or in other variations but didn't work like I need it to.
from Key k
left outer join KeyType kt on kt.Key.id = k.id
left outer join KeySource ks on ks.Key.id = k.id
inner join Source s on ks.Source.id = s.id
where k.deleted = false
and k.name in (...)
and ( kt.appType in (...) or kt is null )
and ( s.SourceFile in (...) or s is null )
well as expected this doesn't work. I mean its quite simpel to see that it can't work but you try
a lot if you can't get a hang on it.
I tried many more combinations and variations of querys but with no luck. The first query is the closest I got.
I Hope somebody can help me.
PS: I can't change the mapping or the entitys now. I have to work with what I got.
UPDATE:
Alright so I am very close to solving the problem. My query now looks like this:
select k, case when kt.Type not in (...) then null
else 1 end
from KeyType kt
join kt.Key k
where k.name in (...)
Now the only thing I want to to is exchange the 1 with the actual object. But if I do this I get the error "org.hibernate.exception.GenericJDBCException: could not execute query" (running on an oracle db)
Can someone tell me how to solve it?

For my case and sorrounding the way I wanted to do it is not possible.
So I asked coworkers again and we came to the solution we have to do it in single querys.
Just saying this for anyone passing by this post with the same table configuration/problem.

Not equal query in hql not work

I am tring the 'not equal' query in hql.
#Override
public Student findStudentsByYear(String year) {
String queryString = "from Student where year<>:year ";
Query query = sessionFactory.getCurrentSession().createQuery(queryString);
query.setParameter("year", year);
return (Student)query.uniqueResult();
}
but it is not working properly.How to write this query correctly
My student table is
+-------------+--------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-------------+--------------+------+-----+---------+----------------+
| studentId | bigint(20) | NO | PRI | NULL | auto_increment |
| course | varchar(255) | YES | | NULL | |
| dob | varchar(255) | YES | | NULL | |
| email | varchar(255) | YES | | NULL | |
| faculty | varchar(255) | YES | | NULL | |
| firstName | varchar(255) | YES | | NULL | |
| gender | int(11) | YES | | NULL | |
| homeAddress | varchar(255) | YES | | NULL | |
| lastName | varchar(255) | YES | | NULL | |
| linkedIn | varchar(255) | YES | | NULL | |
| university | varchar(255) | YES | | NULL | |
| year | varchar(255) | YES | | NULL | |
| user_userId | bigint(20) | YES | MUL | NULL | |
+-------------+--------------+------+-----+---------+----------------+

I know it's late but if anyone is having similar problem, you can use this:
Criteria criteria = session.createCriteria(Student.class);
criteria.add(Restrictions.ne("year", year));
List<Student> result = criteria.list();
Or this:
List<Student> result = session.createQuery ("from Student where year!=:year").setParameter("year", year).list();
I'm not sure what the problem is in above example as Samantha did not provide any information what so ever but my guess is that the uniqueResult() is causing trouble because this query returns a list and not one result.

Java / Hibernate: Could not resolve property with nested object criterias

I'm having problems with a Hibernate criteria. I'm trying to make a Criteria where I look at the id of a member object of the class the query returns.
For example:
Criteria crit = session.createCriteria(Enquiry.class);
crit.add(Expression.eq("lecture.admin.id", userId));`
The result of this is an exception:
org.hibernate.QueryException: could not resolve property: lecture.admin.id of: xxx.yyy.Enquiry
The Enquiry class does contain the lecture variable, which in turn contains the admin variable. I have tried using lecture.id and that works fine.
Is there a limit to the number of levels you can go down the object hierarchy like this?
Thanks!
Code snippets:
public class Lecture extends TransferItem {
private User admin;
public User getAdmin() {
return admin;
}
}
The 'User' class extends the Person class, which in turn extends an Itemclass, which has the getId()method:
public Integer getId() {
if (id != null) {
return id;
}
return TransferBean.NOT_SET;
}
From the Hibernate mapping XML:
<class name="User" table="user">
<id column="user_id" name="id">
<generator class="increment"/>
</id>
...
<class name="Lecture" table="lecture">
<many-to-one class="User" column="user_fk" lazy="false" name="admin"/>`
This is the user table:
mysql> show columns from user;
+-----------------+--------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+-----------------+--------------+------+-----+---------+-------+
| user_id | int(11) | NO | PRI | | |
| firstname | varchar(50) | YES | | NULL | |
| lastname | varchar(50) | YES | | NULL | |
| signature | varchar(16) | YES | | NULL | |
| email_signature | varchar(256) | YES | | NULL | |
| password | varchar(32) | YES | | NULL | |
| phone | varchar(16) | YES | | NULL | |
| email | varchar(255) | YES | UNI | NULL | |
| lecturer_fk | int(11) | YES | MUL | NULL | |
| access | int(11) | YES | | NULL | |
| deleted | tinyint(1) | YES | | NULL | |
+-----------------+--------------+------+-----+---------+-------+
11 rows in set (0.02 sec)`

You can't use nested paths directly in Criteria API (unlike HQL). Instead, you need to create nested criteria instances or define aliases on each "entity.property" pair starting with the first non-root entity:
Criteria criteria = session.createCriteria(Enquiry.class)
.createAlias("lecture", "l")
.createAlias("l.admin", "a")
.add( Restrictions.eqProperty("a.id", userId) );
Note that the very first property is not prefixed as it belongs to the root entity (Enquiry), the others are prefixed with previous level alias. Details are in the documentation.
Also note that id is a special property when it comes to associations; in your case user_fk is a column located in lecture table. It should, therefore, be possible (provided that the mappings you've posted are accurate) to rewrite the above criteria as:
Criteria criteria = session.createCriteria(Enquiry.class)
.createAlias("lecture", "l")
.add( Restrictions.eqProperty("l.admin.id", userId) );
thus eliminating extra join.

Criteria crit = session.createCriteria(Enquiry.class)
crit.createAlias("lecture.admin", "lectureAdmin");
crit.add(Expression.eq("lectureAdmin.id", userId));
You can indeed hit issues when you go down too deep in the object graph. I usually get around this by creating an alias as shown above.