level

  1. level
  2. Description
    1. Interface API
      1. Basic read and write
      2. Atomic updates (batches)
      3. Range queries (streams)
        1. Stream Examples
      4. Synchronous Writes
    2. Creating an interface
    3. Implementations
  3. Source code / issues
  4. Changelog
    1. 3.0.0
    2. 2.0.0

Description

Provides a high-level API to leveldb implementations. Use in combination with an implementation egg (eg, leveldb).

Interface API

This module exposes an interface, which other eggs provide implementations of. The API described below is what the interface provides.

Basic read and write

[procedure] (db-get db key)

Returns the value of key in db as a string. Causes an exception if the key does not exist.

[procedure] (db-get/default db key default)

Same as db-get but returns default on missing keys instead of raising an exception.

[procedure] (db-put db key value #!key (sync #f))

Stores value under key in datbase db. If the sync option can be set to #t to make the write operation not return until the data being written has been pushed all the way to persistent storage. See the Synchronous Writes section for more information.

[procedure] (db-delete db key #!key (sync #f))

Removes the value associated with key from db. If the sync option can be set to #t to make the write operation not return until the data being written has been pushed all the way to persistent storage. See the Synchronous Writes section for more information.

Atomic updates (batches)

[procedure] (db-batch db ops #!key (sync #f))

When making multiple changes that rely on each other you can apply a batch of operations atomically using db-batch. The ops argument is a list of operations which will be applied in order (meaning you can create then later delete a value in the same batch, for example).

    
(define myops '((put "abc" "123")
                (put "def" "456")
                (delete "abc")))

;; apply all operations in myops
(db-batch db myops)

The first item in an operation should be the symbol put or delete, any other value will give an error. The next item is the key and in the case of put the third item is the value.

Apart from its atomicity benefits, db-batch may also be used to speed up bulk updates by placing lots of individual mutations into the same batch.

Range queries (streams)

[procedure] (db-keys db #!key start end limit reverse fillcache)

Allows forward and backward iteration over the keys in alphabetical order. Returns a lazy sequence of all keys from start to end (up to limit). This uses the lazy-seq egg.

[procedure] (db-values db #!key start end limit reverse fillcache)

Allows forward and backward iteration over the keys in alphabetical order. Returns a lazy sequence of all values pairs from start to end (up to limit). This uses the lazy-seq egg.

[procedure] (db-pairs db #!key start end limit reverse fillcache)

Allows forward and backward iteration over the keys in alphabetical order. Returns a lazy sequence of all key/value pairs from start to end (up to limit). This uses the lazy-seq egg.

Stream Examples
(lazy-map display (db-pairs db start: "foo:" end: "foo::" limit: 10)))

You can turn the lazy-seq into a list using lazy-seq->list, just be warned that it will evaluate the entire key range and should be avoided unless you know the number of values is small (eg, when using limit).

    
(db-batch db '((put "foo" "1")
               (put "bar" "2")
               (put "baz" "3")))

(lazy-seq->list (db-pairs db limit: 2)) ;; => (("foo" "1") ("bar" "2"))
(lazy-seq->list (db-values db)) ;; => ("1" "2" "3")
(lazy-seq->list (db-keys db)) ;; => ("foo" "bar" "baz")

Synchronous Writes

Note: this information is mostly copied from the LevelDB docs

By default, each write to leveldb is asynchronous: it returns after pushing the write from the process into the operating system. The transfer from operating system memory to the underlying persistent storage happens asynchronously. The sync flag can be turned on for a particular write to make the write operation not return until the data being written has been pushed all the way to persistent storage. (On Posix systems, this is implemented by calling either fsync(...) or fdatasync(...) or msync(..., MS_SYNC) before the write operation returns.)

Asynchronous writes are often more than a thousand times as fast as synchronous writes. The downside of asynchronous writes is that a crash of the machine may cause the last few updates to be lost. Note that a crash of just the writing process (i.e., not a reboot) will not cause any loss since even when sync is false, an update is pushed from the process memory into the operating system before it is considered done.

db-batch provides an alternative to asynchronous writes. Multiple updates may be placed in the same batch and applied together using a sync: #t. The extra cost of the synchronous write will be amortized across all of the writes in the batch.

Creating an interface

If you want to provide your own storage impelmentation, import this egg and define the interface as follows:

    
(use level)

(define myleveldb
  (implementation level-api

    (define (level-get db key) ...)
    (define (level-get/default db key) ...)
    (define (level-put db key value #!key (sync #f)) ...)
    (define (level-delete db key #!key (sync #f)) ...)
    (define (level-batch db ops #!key (sync #f)) ...)

    (define (level-stream db
                          #!key
                          start
                          end
                          limit
                          reverse
                          (key #t)
                          (value #t)
                          fillcache)
      ..)))

Implementations

Source code / issues

https://github.com/caolan/chicken-level

Changelog

3.0.0

2.0.0