Need help with norm?
Click the “chat” button below for chat support from the developer who created it, or find similar developers for support.

About the developer

167 Stars 26 Forks Other 152 Commits 21 Opened issues


Access a database in one line of code.

Services available


Need anything else?

Contributors list


Read a blog post on this project.

Norm is a simple way to access a JDBC database, usually in one line of code. It purges your code of the complex mess that is Hibernate, JPA, and ORM.

Lots of people think that complex ORMs are a bad idea.

Here it is:

List people = db.where("name=?", "Bob").results(Person.class);

Get Started

Configure your system.

Start with this sample code.


Norm is an extremely lightweight layer over JDBC. It gets rid of large amounts of boilerplate JDBC code. It steals some ideas from ActiveJDBC, which is a very nice system, but requires some very ugly instrumentation / byte code rewriting.


Sometimes the most important thing about writing software is knowing when to stop. A solution that gets you 90% of the way is often good enough, because the other 90% isn't worth the hassle. In this case, Norm gives you a fast and convenient way to do select, insert, update and delete, and when you need more, you just drop into straight SQL.

Norm returns results as a list of POJOs or as a

objects, whichever you prefer.

POJOs are fabulous, truly fabulous:

  • Populating them is really fast with the newer JVMs.
  • You can use them for declaratory data validation.
  • Using Jackson, you can serialize them to JSON.

which means that, yes, you can use the same class to fetch a record from a database and then create JSON from it.

Sample Code

There's a full example here.

Database db = new Database();

Person joe = new Person(); joe.firstName = "Joe"; joe.lastName = "Sixpack";


List people = db.where("lastname=?", "Sixpack").orderBy("lastName").results(Person.class);


class Person {
    public String firstName;
    public String lastName;
    // you can also use getters and setters

You can modify your database using .insert(), .upsert(), .update(), .delete(), and .sql().execute():

int rowsAffected = db.table("people").where("firstName=?", "Joe").delete();

// or just: int rowsAffected = db.delete(joe);

// rowsAffected will equal the number of rows inserted, updated, or deleted

When you need more than this, just use straight SQL. This is the best way to do joins:

List list1 = db.sql(
    "select lastname, sum(amount) from account, transaction " + 
    "where account.accountId = transaction.accountId " +
    "and date > ?", "2000-01-01")

You can also use straight SQL to modify the database:

db.sql("drop table people").execute();

Maps and Lists

Don't want to create a new POJO class for every query? No problem, just use a Map:

List list = db.sql("select * from people").results(HashMap.class);

HashMap, LinkedHashMap or any class that implements the Map interface will work.

Note that you must specify full sql, or at a minimum a table name, because the system won't be able to guess the table name from the Map class. Unless you've annotated it to that effect.


A single column result set can come back in the form of a list of primitives, or even as a single primitive.

Long count = db.sql("select count(*) from people").first(Long.class);

It's sometimes really useful to get a result in the form of a


Note that you have to specify the full sql when doing primitives because the system won't be able to guess the column or tables names from the primitive class.


Tell the system what to do with your POJOs by using a few annotations. Norm implements a subset of the

annotations, including @Table, @Id, @GeneratedValue, @Transient, @Column and @Enumerated.
public class Person {
    public long personId;

public String name;

@Column(name = "theColumnName")
public String renameThis;

public String thisFieldGetsIgnored;


specifies the table name. If it's not there, the table defaults to the class name.

specifies the primary key. The system uses this to identify the record to delete or update.

indicates that the field is marked AUTO_INCREMENT and will be generated on the server. This prevents the field from being inserted, and it fills in the value in the POJO after an insert.

tells the system to ignore the field. It doesn't get persisted to the database. (Note that this is
, not
. Different annotations.)

implements a subset of
will attach a property to a database column of a different name.
, and
apply when you call

specifies the type of the enumeration to be stored in the database. By default
is used for string representation. One can select
for integer representation.

Column-level annotations can go on either a public property or on a public getter for the property. Annotations on setters will be ignored.


If you need multiple database operations to succeed or fail as a unit, use a transaction. The basic scheme is to create a Transaction object, pass it to every query that needs it, and then .commit() or .rollback().

Transaction trans = db.startTransaction();
try {
} catch (Throwable t) {

Transaction is a pretty simple class, so if it doesn't do what you need, just subclass it and make it behave differently.

Custom Serialization

The older @DbSerializable and @DbSerializer annotations are now deprecated. Use @Convert and an AttributeConverter class instead.

Norm now supports JPA serialization. We'll add a bit more documentation here later, but for now you can learn all about it here:

and here:

In short, @Convert give you a way of converting a particular column datatype in your database to a particular datatype in your POJO. So, you can store a list of integers in your database as String, and in your POJO as List<Integer>.

Note that you can sometimes achieve the same purpose by using appropriate getters and setters on your POJO. Mark the ones that Norm should ignore with @Transient.

Pluggable SQL Flavors

You can specify the particular flavor of SQL for your database with

. By default, the
will handle most needs. As of version 0.8.1, there is also a
class that will handle MySql-style upserts. To implement your own flavor, subclass
and possibly
and do what you need.
Database db = new Database();
db.setSqlMaker(new MySqlMaker());

Some database-specific notes:

MySQL: Should work out of the box.

Postgres: Inexplicably, Postgres converts all column names to lowercase when you create a table, and forces you to use double quotes around column names if you want mixed or upper case. The workaround is to add an @Column(name="somelowercasename") annotation to the fields in your pojo.

H2: Does the opposite of Postgres. It forces all column names to upper case. Avoid the problem by adding the databasetoupper option to the jdbcUrl:



Here's the Maven dependency:


To specify the database connection parameters:

Database db = new Database();


System.setProperty("norm.jdbcUrl", "jdbc:mysql://localhost:3306/mydb?useSSL=false");
System.setProperty("norm.user", "root");
System.setProperty("norm.password", "rootpassword");

Internally, Norm uses the Hikari connection pool. Hikari allows you to use the jdbcUrl method or DataSource class names. Your database is bound to be on the list.

If you don't want to use system properties, or your DataSource needs some custom startup parameters, just subclass the Database class and override the .getDataSource() method. You can supply any DataSource you like.


Norm needs javax.persistence, but that's just for annotations.

It also has a dependency on HikariCP for connection pooling, but that's entirely optional. If you don't want it, add an

  to the Norm dependency in your project's pom. Then subclass Database and override the getDataSource() method. 

Finally, you'll need to include your JDBC driver as a dependency. Here's a sample for MySQL:


That's about it. Post any bugs or feature requests to the issue tracker.

We use cookies. If you continue to browse the site, you agree to the use of cookies. For more information on our use of cookies please see our Privacy Policy.