JarCastingSpring's RowMapper implementation using annotated constructors
Creates RowMappers for classes having their constructor arguments annotated with JSR330 @Named annotation. Uses constructors for objects instantiation. Supports class hierarchies using discriminator column. May be used as an alternative for BeanPropertyRowMapper
Library depends on spring-jdbc.
Library is available in Maven cental.
Javadocs for the latest release are available here.
Library usage
Maven dependency (available in central repository):
<dependency>
<groupId>com.alexkasko.springjdbc</groupId>
<artifactId>springjdbc-constructor-mapper</artifactId>
<version>1.0.1</version>
</dependency>
Single class example:
// class with annotated constructor
private static class MyClass {
private final String foo;
private final int bar;
private MyClass(@Named("foo") String foo, @Named("bar") Integer bar) {
this.id = id;
this.bar = null != bar ? bar : -1;
}
}
// mapper creation
RowMapper<MyClass> mapper = NamedConstructorMapper.forClass(MyClass.class);
// standard row mapper usage
MyClass obj = jt.queryForObject("select * from my_table where id = 42", mapper);
Subclasses example:
// builder takes discriminator column name, subclasses are registered with their discriminators
RowMapper<Parent> mapper = NamedConstructorMapper.<Parent>builder("discriminator_column")
.addSubclass("Foo", Foo.class)
.addSubclass("Bar", Bar.class)
.build();
####annotated constructors
To support constructor invocation with unordered row data constructor arguments must be named. There are different ways to access constructor names in runtime (see paranamer). This project uses JSR330 javax.inject.Named annotations, that must be applied to all constructor arguments. Constructors without @Named annotations on arguments will be ignored. Constructors with @Named annotations must have all they arguments annotated with not blank values without duplicates.
####choosing from multiple constrcutors
Constructor will be invoken only if row data contains columns for each constructor argument with the same names. Additional columns in row data will be ignored. If class has multiple annotated constructors, they will be checked against row data based on arguments list size in descending order.
####reflection introspection and caching
All reflection introspection is done on mapper instantiation. After that mapper instances hold references to constructors and their arguments names and invoke constructors for incoming row data. NamedConstructorMapper doesn't cache mapper creation: each NamedConstructorMapper.forClass call do reflection introspection. So it's better to create mappers for each class only once (using static final field or application level cache).
####subclasses mapping
NamedConstructorMapper supports subclasses mapping using discriminator column. Subclasses mapper chooses constructor by discriminator column value.
####columns case sensivity
All column names to @Named values comparisons are case-insensitive using Locale.ENGLISH, so all column names and @Named values must be locale insensitive.
License information
This project is released under the Apache License 2.0
Changelog
1.0.1 (2013-05-06)
- static-import-friendly alias for single classes
1.0 (2012-11-11)
- initial version