Configuration
The configurable items must be returned from the methods of the implementation class of
the org.seasar.doma.jdbc.Config
interface.
Configurable items
DataSource
Return a JDBC DataSource
from the getDataSource
method.
If you need local transactions provided by Doma, return a LocalTransactionDataSource
.
See also: Transaction
Note
The DataSource is a required configuration item.
DataSource’s name
Return a DataSource’s name from the getDataSourceName
method.
In the environment where multiple DataSources are used, the name is important.
You have to give an unique name to each DataSource.
The default value is the full qualified name of the implementation class of Config
.
SQL dialect
Return a Dialect
from the getDialect
method.
You have to choose an appropriate dialect for the database you use.
Doma provides following dialects:
Database |
Dialect Name |
Description |
---|---|---|
DB2 |
Db2Dialect |
|
H2 Database Engine 1.2.126 |
H212126Dialect |
H2 Database Engine 1.2.126 |
H2 Database |
H2Dialect |
H2 Database Engine 1.3.171 and above |
HSQLDB |
HsqldbDialect |
|
Microsoft SQL Server 2008 |
Mssql2008Dialect |
Microsoft SQL Server 2008 |
Microsoft SQL Server |
MssqlDialect |
Microsoft SQL Server 2012 and above |
MySQL |
MySqlDialect |
MySQL 5 and 8 |
Oracle Database 11g |
Oracle11Dialect |
Oracle Database 11g |
Oracle Database |
OracleDialect |
Oracle Database 12g and above |
PostgreSQL |
PostgresDialect |
|
SQLite |
SqliteDialect |
These dialect are located in the org.seasar.doma.jdbc.dialect
package.
MysqlDialect supports MySQL 5 by default. To use it as a dialect for MySQL 8, instantiate MysqlDialect by specifying the version as follows:
MysqlDialect dialect = new MysqlDialect(MysqlDialect.MySqlVersion.V8);
Note
The SQL dialect is a required configuration item.
Logger
Return a JdbcLogger
from the getJdbcLogger
method.
Doma provides following JdbcLogger:
org.seasar.doma.jdbc.UtilLoggingJdbcLogger
The default JdbcLogger is UtilLoggingJdbcLogger which uses java.util.logging
.
SQL File Repository
Return a SqlFileRepository
from the getSqlFileRepository
method.
Doma provides following SqlFileRepositories:
org.seasar.doma.jdbc.GreedyCacheSqlFileRepository
org.seasar.doma.jdbc.NoCacheSqlFileRepository
The default SqlFileRepository is GreedyCacheSqlFileRepository which caches the result of SQL parsing without limitation.
Controlling REQUIRES_NEW transaction
Return a RequiresNewController
from the getRequiresNewController
method.
RequiresNewController may begin new transactions to makes transaction locked time shorter.
This feature is used only when you use @TableGenerator
which generates identities with the table.
The default RequiresNewController does nothing.
Loading classes
Return a ClassHelper
from the getClassHelper
method.
When the application server and framework you use loads classes with their specific way, consider to create your own ClassHelper.
The default ClassHelper loads classes with Class#forName
mainly.
Choosing SQL format contained in exception messages
Return a SqlLogType
from the getExceptionSqlLogType
.
The default SqlLogType contains the formatted SQL in exception messages.
Handling duplicate columns
Return a DuplicateColumnHandler
from the getDuplicateColumnHandler
method.
In result set mappings, if a duplicate column for an entity class is found,
the DuplicateColumnHandler
handles the situation.
The default DuplicateColumnHandler
does nothing.
To throw a DuplicateColumnException
, return a ThrowingDuplicateColumnHandler
.
Handling unknown columns
Return a UnknownColumnHandler
from the getUnknownColumnHandler
method.
In result set mappings, if an unknown column to an entity class is found,
the UnknownColumnHandler handles the situation.
The default UnknownColumnHandler throws an UnknownColumnException
.
Naming convention for tables and columns
Return a Naming
from the getNaming
method.
The naming element of @Entity
have preference over this value.
When you specify explicit value to the name elements of @Table
and @Column
,
the naming convention is not applied to them.
The default Naming does nothing.
Naming convention for keys of java.util.Map
Return a MapKeyNaming
from the getMapKeyNaming
method.
The MapKeyNaming is used when the result set is mapped to java.util.Map<String, Object>
.
The default MapKeyNaming does nothing.
Local transaction manager
Return a LocalTransactionManager
from the getTransactionManager
method.
The getTransactionManager
method throws UnsupportedOperationException
as default.
See also: Transaction
Adding SQL identifiers to the SQLs as a comment
Return a Commenter
from the getCommenter
method.
Doma provides following commenter:
org.seasar.doma.jdbc.CallerCommenter
The default Commenter does nothing.
Command implementors
Return a CommandImplementors
from the getCommandImplementors
method.
For example, the CommandImplementors provides you a hook to execute JDBC API.
Query implementors
Return a QueryImplementors
from the getQueryImplementors
method.
For example, the QueryImplementors provides you a hook to rewrite SQL statements.
Query timeout
Return the query timeout (second) from the getQueryTimeout
method.
This value is used as default in Queries.
Max rows
Return the max rows from the getMaxRows
method.
This value is used as default in Select.
Fetch size
Return the fetch size from the getFetchSize
method.
This value is used as default in Select.
Batch size
Return the batch size from the getBatchSize
method.
This value is used as default in Batch insert,
Batch update and Batch delete.
Providing entity listeners
Return a EntityListenerProvider
from the getEntityListenerProvider
method.
When you want to get entity listeners from a dependency injection container,
create your own EntityListenerProvider.
The default EntityListenerProvider get the entity listener from the accepted supplier.
SQL Builder Settings
Return a SqlBuilderSettings
from the getSqlBuilderSettings
method.
SqlBuilderSettings
controls the following aspects of SQL building:
Whether to remove block comments from SQL
Whether to remove line comments from SQL
Whether to remove blank lines from SQL
Whether to enable IN list padding
IN list padding is a feature that pads the parameters in an SQL IN clause with the last parameter when the number of parameters is less than a power of 2. This feature helps ensure that the same SQL statement is more likely to be generated regardless of the number of parameters, which can have positive effects on SQL caching and related performance optimizations.
By default, no special controls are applied.
Statistic Manager
Return a StatisticManager
from the getStatisticManager
method.
StatisticManager
manages statistical information related to SQL execution.
It retains the following information for each SQL statement:
execution count
execution maximum time in milliseconds
execution minimum time in milliseconds
total execution time in milliseconds
average execution time in milliseconds
Collection of statistical information is disabled by default. To enable it, do the following:
Config config = ...
config.getStatisticManager().setEnabled(true);
To disable it, call setEnabled(false)
.
The default implementation collects statistical information indefinitely while enabled.
To prevent memory exhaustion, either call the clear
method of StatisticManager
periodically
or create an appropriate implementation class for StatisticManager
.
Loading JDBC drivers
All JDBC drivers are loaded automatically by the service provider mechanism.
Warning
But in the specific environment, the mechanism doesn’t work appropriately. For example, when you use Apache Tomcat, you will find the case. See also: DriverManager, the service provider mechanism and memory leaks
Configuration definition
Simple definition
The simple definition is appropriate in following cases:
The configuration instance isn’t managed in the dependency injection container
Local transactions is used
public class DbConfig implements Config {
private static final DbConfig CONFIG = new DbConfig();
private final Dialect dialect;
private final LocalTransactionDataSource dataSource;
private final TransactionManager transactionManager;
private DbConfig() {
dialect = new H2Dialect();
dataSource = new LocalTransactionDataSource(
"jdbc:h2:mem:tutorial;DB_CLOSE_DELAY=-1", "sa", null);
transactionManager = new LocalTransactionManager(
dataSource.getLocalTransaction(getJdbcLogger()));
}
@Override
public Dialect getDialect() {
return dialect;
}
@Override
public DataSource getDataSource() {
return dataSource;
}
@Override
public TransactionManager getTransactionManager() {
return transactionManager;
}
public static DbConfig singleton() {
return CONFIG;
}
}
You can use the above DbConfig
class as follows:
EmployeeDao dao = new EmployeeDaoImpl(DbConfig.singleton());
The above EmployeeDao
interface must be annotated with the @Dao
annotation as follows:
@Dao
public interface EmployeeDao {
@Select
Employee selectById(Integer id);
}
More Simple definition
You can build the configuration more easily by using org.seasar.doma.jdbc.SimpleConfig
.
SimpleConfig
determines the Dialect
based on the connection string and manages transactions using local transactions.
Here is an example of building a Config
using SimpleConfig
.
Config config = SimpleConfig.builder("jdbc:h2:mem:tutorial;DB_CLOSE_DELAY=-1", "sa", null)
.naming(Naming.SNAKE_UPPER_CASE)
.queryTimeout(10)
.build();
You can use the above config
instance as follows:
EmployeeDao dao = new EmployeeDaoImpl(config);
Note
SimpleConfig
is primarily intended for use in sample or test code.
Advanced definition
The advanced definition is appropriate in following cases:
The configuration instance is managed as a singleton object in the dependency injection container
The transaction manager is provided from the application server or framework you use
Suppose the dialect
and the dataSource
are injected by the dependency injection container:
public class DbConfig implements Config {
private Dialect dialect;
private DataSource dataSource;
@Override
public Dialect getDialect() {
return dialect;
}
public void setDialect(Dialect dialect) {
this.dialect = dialect;
}
@Override
public DataSource getDataSource() {
return dataSource;
}
public void setDataSource(DataSource dataSource) {
this.dataSource = dataSource;
}
}
To inject the instance of the above class to your DAO implementation instance,
you have to annotate your DAO interfaces with @AnnotateWith
:
@Dao
@AnnotateWith(annotations = {
@Annotation(target = AnnotationTarget.CONSTRUCTOR, type = javax.inject.Inject.class),
@Annotation(target = AnnotationTarget.CONSTRUCTOR_PARAMETER, type = javax.inject.Named.class, elements = "\"config\"") })
public interface EmployeeDao {
@Select
Employee selectById(Integer id);
}
@Dao
@AnnotateWith(annotations = {
@Annotation(target = AnnotationTarget.CONSTRUCTOR, type = javax.inject.Inject.class),
@Annotation(target = AnnotationTarget.CONSTRUCTOR_PARAMETER, type = javax.inject.Named.class, elements = "\"config\"") })
public interface DepartmentDao {
@Select
Department selectById(Integer id);
}
To avoid annotating your DAO interfaces with @AnnotateWith
repeatedly,
annotate the arbitrary annotation with it only once:
@AnnotateWith(annotations = {
@Annotation(target = AnnotationTarget.CONSTRUCTOR, type = javax.inject.Inject.class),
@Annotation(target = AnnotationTarget.CONSTRUCTOR_PARAMETER, type = javax.inject.Named.class, elements = "\"config\"") })
public @interface InjectConfig {
}
Then, you can annotate your DAO interfaces with the above @InjectConfig
annotation:
@Dao
@InjectConfig
public interface EmployeeDao {
@Select
Employee selectById(Integer id);
}
@Dao
@InjectConfig
public interface DepartmentDao {
@Select
Department selectById(Integer id);
}