Cupboard is an embedded database library for Clojure, intended to make writing Clojure objects to disk, and retrieving them, a relatively simple task. Cupboard is almost completely schemaless, supports unique and non-unique indices, querying by one or more indices, and ACID transactional semantics.
At its lowest level, Cupboard uses Berkeley DB JE, and contains a fairly
complete Clojure wrapper for the JE API. This API is available in the
cupboard.bdb.je package; some projects may find this code useful on its
cupboard.core package provides a higher-level abstraction.
This code makes some struct-maps representing books, saves them into a database, and updates them.
(ns examples.cupboard.simple (:use clojure.contrib.pprint) (:require [cupboard.core :as cb]) (:use [cupboard.utils])) (cb/open-cupboard! "/tmp/books") (cb/defpersist book ((:isbn :index :unique) (:title :index :any) (:author :index :any))) (cb/make-instance book ["0393039854", "The First Folio", "Shakespeare, William"]) (cb/make-instance book ["0393925870", "The Canterbury Tales", "Chaucer"]) (cb/make-instance book ["143851557X", "Troilus and Criseyde", "Chaucer"]) (cb/make-instance book ["019280619X", "Paradise Lost", "Milton, John"]) (println "this returns maps") (pprint (cb/query (starts-with :isbn "039"))) (println "this returns struct-map values") (pprint (cb/query (= :author "Chaucer") :struct book)) (println "this updates the :author slot") (cb/query (= :author "Chaucer") :callback #(cb/passoc! % :author "Chaucer, Geoffrey")) (println "this verifies that the update worked") (pprint (cb/query (starts-with :author "Chaucer"))) (cb/close-cupboard!)
This example is available in
src/examples/cupboard/gutenberg.clj uses a few more of Cupboard's features.
build.xml contains an Ant task which downloads its dependencies,
including Clojure and clojure-contrib.
- Clojure (version 1.0.0 tested)
- clojure-contrib (1.0-compatible branch, Git revision e0080e64 tested)
- Berkeley DB JE (version 3.3.87 tested)
- Joda Time (version 1.6 tested)
- Clone the Cupboard Git repository.
ant package. This pulls down all dependencies and puts them into the
libdirectory. Then it compiles Cupboard's source into Java class files and packages everything into a jar file.
- Copy the Cupboard jar and its dependencies to wherever you normally copy your Clojure dependencies. Make sure to add them to your project's classpath.
Cupboard's documentation is still a work in progress. Please refer to the
directory in the source distribution.
Cupboard is distributed under the MIT license, and so has few restrictions itself. Its Berkeley DB dependency, however, is licensed by Oracle — see Oracle's Berkeley DB Licensing Information page for details.