A simple persistence library for redis
Switch branches/tags
Nothing to show
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.
src Added Short serialization support Jan 1, 2012
.gitignore Added .idea to gitignore Jan 1, 2012
README.md Update README.md Jan 1, 2012
pom.xml First commit Jan 1, 2012


#Jedi Jedi is a simple and small library for storing objects in redis. It uses Jedis internally so it is fully compatible with Redis 2.0.0.

This project is not ready for production yet.

It is not an orm framework

So it does not support orm features, like relation mappings, which Johm provided.


Jedi works fine with simple java pojos, just use very small amount of annotations

public class Book {

    private Long id;
    private String author;
    private String title;
    private Date publishedAt;

Done. Let's begin some crud staffs

###Basic CRUD

//initialize jedis pool
JedisPool jedisPool = new JedisPool("localhost", 6379);
Jedi jedi = new Jedi(jedisPool);

Book book = new Book();
book.setTile("Harry Potter");

// the generated redis hash will be { id: __auto_generated__, title: "Harry Potter", author: "J.K.Rowling" }

assertNotNull(book.getId()); //auto id generation
assertNotNull(jedi.get(Book.class, book.getId())); //ready for id lookup

Long id = book.getId();

//oh, we forgot to set the publish date
book.setPublishedAt(new Date());

// the redis hash will be updated to { id: __auto_generated__, title: "Harry Potter", author: "J.K.Rowling", publishedAt: __epoch__ }
// the java.util.Date will serialized to epoch, the Serialization will be covered later

jedi.delete(book); //see-ya, or you may need jedi.delete(Book.class, book.getId());

Directly talk to redis

jedi.withJedis(new JedisCallback() {
    public void execute(Jedis jedis) {
        //automatic jedis pool resource managment like JdbcTemplate does



jedi.withTransaction(new TransactionCallback() {
    public void execute(Transaction transaction) {
        //automatic transaltion start, commit and discard on exception

###Index support Let's add some indexes to Book

@Index(on = "author")
public class Book {

Then we can find books by their author

Iterator<Book> booksOfRowling = jedis.find(Book.class).where("author").is("J.K.Rowling").iterate(0, 1);

index rebuild is still in working progress. So for now, the old objects will not be indexed.

###Range support

@Index(on = "author", range = "publishedAt")
public class Book {

Then you can find books by author and its published date

Iterator<Book> booksOfRowlingAfter2001 = jedis.find(Book.class)
                                                .within(date2001InEpoch, Double.MAX_VALUE);

If you need to index author both in published date range and some other fields, you should declare it like:

        @Index(on = "author", range = "publishedAt"),
        @Index(on = "author", range = "someOtherField")
public class Book {

###Multi fields index support So what to do if we want to find books by its author and its title?

@Index(on = {"author", "title"})
//or you may prefer @Index(on = "author, title")
public class Book {

Then you can find them like:

jedi.find(Book.class).where("author").is("J.K.Rowling").and("title").is("Harry Potter").iterator(0, 1);

Maybe we should add some method like fetchOne for unique index? #Serialization The java object will be converted to String based hash, then perisist to redis. for now jedi supports types:

  • All basic primitive types and their wrappers
  • java.util.Date
  • String of couse

You can implemated your own StringSerializer, and register it using:

StringSerializers.regiser(yourType, yourSerializer);