Some projects relying on a relational database do just fine using a high-level ORM, some have need for more low-level control over the queries. Using JDBC directly for application development can be really cumbersome. Not because it's a bad API, the problem is it's so low-level it's really more suitable for tool- than application development.
There are many abstraction layer libraries over JDBC that add the much needed syntax sugar. Even so, sometimes I find it difficult to find a suitable one for the applications I work on. Either because they constrain in some way or have other unwanted properties (like transitive dependencies).
In some cases, I implemented abstractions of my own and a pattern has started to emerge. I decided to collect those ideas and make it a library for easier reuse. A library that makes the most common SQL querying operations as simple, declarative as possible, is light-weight, has no dependencies. Most similar libraries don't make use of a fluent API which makes way for some interesting possibilities in terms of both readability and flexibility. Also I wanted to put some emphasis on easy integration to popular persistence / transaction management tools since the goal is to complement them, not necessarily replace them. As a result, it blends quite well with JPA, Spring, Guice, JEE, etc.
Some of FluentJdbc's features are listed below:
- Easy integration to a project (DataSource and other alternatives)
- Execution of
select/insert/update/delete/alter/... statements as one-liners - Parameter mapping (named, positional, supporting java.time, extension with custom types)
- Accessing generated keys of insert/update queries
- Automatic result -> pojo mapping
- Transaction handling
Check out FluentJdbc's github page for more details. The artifact is up in maven central, if someone wants to give it a shot:
<dependency>
<groupid>org.codejargon</groupid>
<artifactid>fluentjdbc</artifactid>
<version>0.9.9</version>
</dependency>
Note: Java 8 only
