Introduction

This package provides a simple programming interface for caching function outputs (i.e. memoization) to memory, disk or both. The API that exposes functionality for creating in-memory cache structures and accessing, writing and synchronizing these to disk. It supports maximum sizes (in number of objects or KiB of memory) and compression (through TranscodingStreams.jl codecs). Since this a work-in-progress, there are bound to be rough edges and little to no documentation. However, the interface is accessible enough to be productively employed at this stage.

Main features

• Non intrusive, easy to use @cache macro
• Fast and type stable if type annotations are used in the function definitions
• Loading/saving from/to disk with compression (.bzip2 and .gzip)
• Maximum in-memory cache size (either number of objects or bytes)
• Can save almost anything to disk (uses Serialization so it is slow)

Installation

In the shell of choice, using

$ git clone https://github.com/zgornel/Caching.jl

or, inside Julia

] add Caching

and for the latest master branch,

] add https://github.com/zgornel/Caching.jl#master