Spring-boot：JPA查询（JPQL）不起作用，但在更改为本机查询时有效

Question

I'm new on Spring-boot. I created a simple JPA app that use web and H2 and JPA dependency. I want to have embedded database with h2 and use JPA to connect with and have some requests.

I created the local table in schema.sql and put this code:

drop TABLE  if EXISTS  stock;
CREATE TABLE stock(stockId INT ,companyName VARCHAR(20),price DECIMAL (10,2)
);

INSERT INTO stock VALUES (1,'apple',345.01);
INSERT INTO stock VALUES (2,'google',8999.00);
INSERT INTO stock VALUES (3,'kashky',56.60);

I also created the stock like this:

@Entity
public class Stock {

    @Id
    private int stockId;

    private String companyName;

    private double price;

... with all setter and getters

}

and this is my application.properties

spring.jpa.hibernate.ddl-auto=none
server.port=9090

Java Persistence Query language (JPQL) do not work but when I change to native query it it works fine.

this code that should return all my records and does not works:

@SpringBootApplication
@RestController
public class SpringDataTestApplication {

    @Autowired
    private EntityManager em;

    @RequestMapping("/stocks")
    public List<Stock> stocks(){
        return em.createQuery("select s from Stock s").getResultList();
    }

if I change query to Native mode it is works:

@RequestMapping("/stocks")
public List<Stock> stocks(){
    return em.createNativeQuery("select * from Stock").getResultList();
}

and this is the error:

This application has no explicit mapping for /error, so you are seeing this as a fallback.

Wed Jan 25 12:51:38 IRST 2017
There was an unexpected error (type=Internal Server Error, status=500).
org.hibernate.exception.SQLGrammarException: could not prepare statement

why this happen? I think I have some problem with JPA cause in later I can not even use StockRepository that i do not mention in code but i think all of this is chained to gather.

Answer 1

I would guess that you would need to add dialect infos in your application.properties file, eg for hibernate:

<property name="hibernate.dialect" value="org.hibernate.dialect.H2Dialect"/>

I guess spring has its own knobs for that task. Knowing the dialect, your program knows the "flavor" of the SQL to create. Native SQL works, because there is no translation to be done.

Answer 2

I finally find the answer. It is an unwritten convention that exist and when you use camel-case properties you had to use under lined fields in schema.sql file.

so with my entity file like this:

@Entity
public class Stock {

    @Id
    private int stockId;

    private String companyName;

    private double price;

... with all setter and getters

}

I changed schema.sql like this and error gone and it's worked fine.

drop TABLE  if EXISTS  stock;
CREATE TABLE stock(stock_id INT ,company_name VARCHAR(20),price DECIMAL (10,2)
);

INSERT INTO stock VALUES (1,'apple',345.01);
INSERT INTO stock VALUES (2,'google',8999.00);
INSERT INTO stock VALUES (3,'kashky',56.60);

I also check that even stock_Id and company_Name meet the JPA convention and works. so you just need to put Underline between camel-cased properties name.