Features

Lightweight. Priha requires only minimal dependencies to work.

Customizable. The Priha RepositoryProvider interface allows you to interface with almost any repository type you want. Priha ships with both a provider for JDBC databases as well as regular file system databases. In addition, there is an optional caching library powered by EhCache, and also a very fast memory-only provider. Priha also supports multiple providers, as each workspace can have its own provider.

Embeddable. Priha can be easily embedded in your own application, just as another JAR file, giving you the power of JCR with minimal work.

Level 1 JCR support. Priha has full support of JSR-170 Level 1.

Level 2 JCR support. Priha has full support of JSR-170 Level 2.

Locking. Priha supports JSR-170 locking feature.

Limitations

To be honest, Priha is not yet as good as commercial (or open source, for that matter) alternatives. At the moment the serious limitations are:

• Performance is not terribly good overall. For the most part, it's quite tolerable. However, JdbcProvider is quite snappy, and really limited by the performance of the underlying database.
• While all mandatory parts of JSR-170 are implemented, many optional parts are not.
• Parts of Priha have not yet been tested heavily. This means that it cannot be considered reliable at this point. While things seems to work generally, there are some edge cases (like running out of memory/disk space) where its behaviour is unknown.
• Scalability limits are currently unknown.
• JDBC support is limited. Currently only HSQLDB and H2 databases are supported, but support for others is easily added.
• Queries can be very slow in large repositories.