Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/grantjenks/python-diskcache

Python disk-backed cache (Django-compatible). Faster than Redis and Memcached. Pure-Python.
https://github.com/grantjenks/python-diskcache

cache filesystem key-value-store persistence python

Last synced: 15 days ago
JSON representation

Python disk-backed cache (Django-compatible). Faster than Redis and Memcached. Pure-Python.

Awesome Lists containing this project

README 

        
DiskCache: Disk Backed Cache

============================



`DiskCache`_ is an Apache2 licensed disk and file backed cache library, written

in pure-Python, and compatible with Django.



The cloud-based computing of 2023 puts a premium on memory. Gigabytes of empty

space is left on disks as processes vie for memory. Among these processes is

Memcached (and sometimes Redis) which is used as a cache. Wouldn't it be nice

to leverage empty disk space for caching?



Django is Python's most popular web framework and ships with several caching

backends. Unfortunately the file-based cache in Django is essentially

broken. The culling method is random and large caches repeatedly scan a cache

directory which slows linearly with growth. Can you really allow it to take

sixty milliseconds to store a key in a cache with a thousand items?



In Python, we can do better. And we can do it in pure-Python!



::



   In [1]: import pylibmc

   In [2]: client = pylibmc.Client(['127.0.0.1'], binary=True)

   In [3]: client[b'key'] = b'value'

   In [4]: %timeit client[b'key']



   10000 loops, best of 3: 25.4 µs per loop



   In [5]: import diskcache as dc

   In [6]: cache = dc.Cache('tmp')

   In [7]: cache[b'key'] = b'value'

   In [8]: %timeit cache[b'key']



   100000 loops, best of 3: 11.8 µs per loop



**Note:** Micro-benchmarks have their place but are not a substitute for real

measurements. DiskCache offers cache benchmarks to defend its performance

claims. Micro-optimizations are avoided but your mileage may vary.



DiskCache efficiently makes gigabytes of storage space available for

caching. By leveraging rock-solid database libraries and memory-mapped files,

cache performance can match and exceed industry-standard solutions. There's no

need for a C compiler or running another process. Performance is a feature and

testing has 100% coverage with unit tests and hours of stress.



Testimonials

------------



`Daren Hasenkamp`_, Founder --



    "It's a useful, simple API, just like I love about Redis. It has reduced

    the amount of queries hitting my Elasticsearch cluster by over 25% for a

    website that gets over a million users/day (100+ hits/second)."



`Mathias Petermann`_, Senior Linux System Engineer --



    "I implemented it into a wrapper for our Ansible lookup modules and we were

    able to speed up some Ansible runs by almost 3 times. DiskCache is saving

    us a ton of time."



Does your company or website use `DiskCache`_? Send us a `message

`_ and let us know.



.. _`Daren Hasenkamp`: https://www.linkedin.com/in/daren-hasenkamp-93006438/

.. _`Mathias Petermann`: https://www.linkedin.com/in/mathias-petermann-a8aa273b/



Features

--------



- Pure-Python

- Fully Documented

- Benchmark comparisons (alternatives, Django cache backends)

- 100% test coverage

- Hours of stress testing

- Performance matters

- Django compatible API

- Thread-safe and process-safe

- Supports multiple eviction policies (LRU and LFU included)

- Keys support "tag" metadata and eviction

- Developed on Python 3.10

- Tested on CPython 3.6, 3.7, 3.8, 3.9, 3.10

- Tested on Linux, Mac OS X, and Windows

- Tested using GitHub Actions



.. image:: https://github.com/grantjenks/python-diskcache/workflows/integration/badge.svg

   :target: https://github.com/grantjenks/python-diskcache/actions?query=workflow%3Aintegration



.. image:: https://github.com/grantjenks/python-diskcache/workflows/release/badge.svg

   :target: https://github.com/grantjenks/python-diskcache/actions?query=workflow%3Arelease



Quickstart

----------



Installing `DiskCache`_ is simple with `pip `_::



  $ pip install diskcache



You can access documentation in the interpreter with Python's built-in help

function::



  >>> import diskcache

  >>> help(diskcache)                             # doctest: +SKIP



The core of `DiskCache`_ is three data types intended for caching. `Cache`_

objects manage a SQLite database and filesystem directory to store key and

value pairs. `FanoutCache`_ provides a sharding layer to utilize multiple

caches and `DjangoCache`_ integrates that with `Django`_::



  >>> from diskcache import Cache, FanoutCache, DjangoCache

  >>> help(Cache)                                 # doctest: +SKIP

  >>> help(FanoutCache)                           # doctest: +SKIP

  >>> help(DjangoCache)                           # doctest: +SKIP



Built atop the caching data types, are `Deque`_ and `Index`_ which work as a

cross-process, persistent replacements for Python's ``collections.deque`` and

``dict``. These implement the sequence and mapping container base classes::



  >>> from diskcache import Deque, Index

  >>> help(Deque)                                 # doctest: +SKIP

  >>> help(Index)                                 # doctest: +SKIP



Finally, a number of `recipes`_ for cross-process synchronization are provided

using an underlying cache. Features like memoization with cache stampede

prevention, cross-process locking, and cross-process throttling are available::



  >>> from diskcache import memoize_stampede, Lock, throttle

  >>> help(memoize_stampede)                      # doctest: +SKIP

  >>> help(Lock)                                  # doctest: +SKIP

  >>> help(throttle)                              # doctest: +SKIP



Python's docstrings are a quick way to get started but not intended as a

replacement for the `DiskCache Tutorial`_ and `DiskCache API Reference`_.



.. _`Cache`: http://www.grantjenks.com/docs/diskcache/tutorial.html#cache

.. _`FanoutCache`: http://www.grantjenks.com/docs/diskcache/tutorial.html#fanoutcache

.. _`DjangoCache`: http://www.grantjenks.com/docs/diskcache/tutorial.html#djangocache

.. _`Django`: https://www.djangoproject.com/

.. _`Deque`: http://www.grantjenks.com/docs/diskcache/tutorial.html#deque

.. _`Index`: http://www.grantjenks.com/docs/diskcache/tutorial.html#index

.. _`recipes`: http://www.grantjenks.com/docs/diskcache/tutorial.html#recipes



User Guide

----------



For those wanting more details, this part of the documentation describes

tutorial, benchmarks, API, and development.



* `DiskCache Tutorial`_

* `DiskCache Cache Benchmarks`_

* `DiskCache DjangoCache Benchmarks`_

* `Case Study: Web Crawler`_

* `Case Study: Landing Page Caching`_

* `Talk: All Things Cached - SF Python 2017 Meetup`_

* `DiskCache API Reference`_

* `DiskCache Development`_



.. _`DiskCache Tutorial`: http://www.grantjenks.com/docs/diskcache/tutorial.html

.. _`DiskCache Cache Benchmarks`: http://www.grantjenks.com/docs/diskcache/cache-benchmarks.html

.. _`DiskCache DjangoCache Benchmarks`: http://www.grantjenks.com/docs/diskcache/djangocache-benchmarks.html

.. _`Talk: All Things Cached - SF Python 2017 Meetup`: http://www.grantjenks.com/docs/diskcache/sf-python-2017-meetup-talk.html

.. _`Case Study: Web Crawler`: http://www.grantjenks.com/docs/diskcache/case-study-web-crawler.html

.. _`Case Study: Landing Page Caching`: http://www.grantjenks.com/docs/diskcache/case-study-landing-page-caching.html

.. _`DiskCache API Reference`: http://www.grantjenks.com/docs/diskcache/api.html

.. _`DiskCache Development`: http://www.grantjenks.com/docs/diskcache/development.html



Comparisons

-----------



Comparisons to popular projects related to `DiskCache`_.



Key-Value Stores

................



`DiskCache`_ is mostly a simple key-value store. Feature comparisons with four

other projects are shown in the tables below.



* `dbm`_ is part of Python's standard library and implements a generic

  interface to variants of the DBM database — dbm.gnu or dbm.ndbm. If none of

  these modules is installed, the slow-but-simple dbm.dumb is used.

* `shelve`_ is part of Python's standard library and implements a “shelf” as a

  persistent, dictionary-like object. The difference with “dbm” databases is

  that the values can be anything that the pickle module can handle.

* `sqlitedict`_ is a lightweight wrapper around Python's sqlite3 database with

  a simple, Pythonic dict-like interface and support for multi-thread

  access. Keys are arbitrary strings, values arbitrary pickle-able objects.

* `pickleDB`_ is a lightweight and simple key-value store. It is built upon

  Python's simplejson module and was inspired by Redis. It is licensed with the

  BSD three-clause license.



.. _`dbm`: https://docs.python.org/3/library/dbm.html

.. _`shelve`: https://docs.python.org/3/library/shelve.html

.. _`sqlitedict`: https://github.com/RaRe-Technologies/sqlitedict

.. _`pickleDB`: https://pythonhosted.org/pickleDB/



**Features**



================ ============= ========= ========= ============ ============

Feature          diskcache     dbm       shelve    sqlitedict   pickleDB

================ ============= ========= ========= ============ ============

Atomic?          Always        Maybe     Maybe     Maybe        No

Persistent?      Yes           Yes       Yes       Yes          Yes

Thread-safe?     Yes           No        No        Yes          No

Process-safe?    Yes           No        No        Maybe        No

Backend?         SQLite        DBM       DBM       SQLite       File

Serialization?   Customizable  None      Pickle    Customizable JSON

Data Types?      Mapping/Deque Mapping   Mapping   Mapping      Mapping

Ordering?        Insert/Sorted None      None      None         None

Eviction?        LRU/LFU/more  None      None      None         None

Vacuum?          Automatic     Maybe     Maybe     Manual       Automatic

Transactions?    Yes           No        No        Maybe        No

Multiprocessing? Yes           No        No        No           No

Forkable?        Yes           No        No        No           No

Metadata?        Yes           No        No        No           No

================ ============= ========= ========= ============ ============



**Quality**



================ ============= ========= ========= ============ ============

Project          diskcache     dbm       shelve    sqlitedict   pickleDB

================ ============= ========= ========= ============ ============

Tests?           Yes           Yes       Yes       Yes          Yes

Coverage?        Yes           Yes       Yes       Yes          No

Stress?          Yes           No        No        No           No

CI Tests?        Linux/Windows Yes       Yes       Linux        No

Python?          2/3/PyPy      All       All       2/3          2/3

License?         Apache2       Python    Python    Apache2      3-Clause BSD

Docs?            Extensive     Summary   Summary   Readme       Summary

Benchmarks?      Yes           No        No        No           No

Sources?         GitHub        GitHub    GitHub    GitHub       GitHub

Pure-Python?     Yes           Yes       Yes       Yes          Yes

Server?          No            No        No        No           No

Integrations?    Django        None      None      None         None

================ ============= ========= ========= ============ ============



**Timings**



These are rough measurements. See `DiskCache Cache Benchmarks`_ for more

rigorous data.



================ ============= ========= ========= ============ ============

Project          diskcache     dbm       shelve    sqlitedict   pickleDB

================ ============= ========= ========= ============ ============

get                      25 µs     36 µs     41 µs       513 µs        92 µs

set                     198 µs    900 µs    928 µs       697 µs     1,020 µs

delete                  248 µs    740 µs    702 µs     1,717 µs     1,020 µs

================ ============= ========= ========= ============ ============



Caching Libraries

.................



* `joblib.Memory`_ provides caching functions and works by explicitly saving

  the inputs and outputs to files. It is designed to work with non-hashable and

  potentially large input and output data types such as numpy arrays.

* `klepto`_ extends Python’s `lru_cache` to utilize different keymaps and

  alternate caching algorithms, such as `lfu_cache` and `mru_cache`. Klepto

  uses a simple dictionary-sytle interface for all caches and archives.



.. _`klepto`: https://pypi.org/project/klepto/

.. _`joblib.Memory`: https://joblib.readthedocs.io/en/latest/memory.html



Data Structures

...............



* `dict`_ is a mapping object that maps hashable keys to arbitrary

  values. Mappings are mutable objects. There is currently only one standard

  Python mapping type, the dictionary.

* `pandas`_ is a Python package providing fast, flexible, and expressive data

  structures designed to make working with “relational” or “labeled” data both

  easy and intuitive.

* `Sorted Containers`_ is an Apache2 licensed sorted collections library,

  written in pure-Python, and fast as C-extensions. Sorted Containers

  implements sorted list, sorted dictionary, and sorted set data types.



.. _`dict`: https://docs.python.org/3/library/stdtypes.html#typesmapping

.. _`pandas`: https://pandas.pydata.org/

.. _`Sorted Containers`: http://www.grantjenks.com/docs/sortedcontainers/



Pure-Python Databases

.....................



* `ZODB`_ supports an isomorphic interface for database operations which means

  there's little impact on your code to make objects persistent and there's no

  database mapper that partially hides the datbase.

* `CodernityDB`_ is an open source, pure-Python, multi-platform, schema-less,

  NoSQL database and includes an HTTP server version, and a Python client

  library that aims to be 100% compatible with the embedded version.

* `TinyDB`_ is a tiny, document oriented database optimized for your

  happiness. If you need a simple database with a clean API that just works

  without lots of configuration, TinyDB might be the right choice for you.



.. _`ZODB`: http://www.zodb.org/

.. _`CodernityDB`: https://pypi.org/project/CodernityDB/

.. _`TinyDB`: https://tinydb.readthedocs.io/



Object Relational Mappings (ORM)

................................



* `Django ORM`_ provides models that are the single, definitive source of

  information about data and contains the essential fields and behaviors of the

  stored data. Generally, each model maps to a single SQL database table.

* `SQLAlchemy`_ is the Python SQL toolkit and Object Relational Mapper that

  gives application developers the full power and flexibility of SQL. It

  provides a full suite of well known enterprise-level persistence patterns.

* `Peewee`_ is a simple and small ORM. It has few (but expressive) concepts,

  making it easy to learn and intuitive to use. Peewee supports Sqlite, MySQL,

  and PostgreSQL with tons of extensions.

* `SQLObject`_ is a popular Object Relational Manager for providing an object

  interface to your database, with tables as classes, rows as instances, and

  columns as attributes.

* `Pony ORM`_ is a Python ORM with beautiful query syntax. Use Python syntax

  for interacting with the database. Pony translates such queries into SQL and

  executes them in the database in the most efficient way.



.. _`Django ORM`: https://docs.djangoproject.com/en/dev/topics/db/

.. _`SQLAlchemy`: https://www.sqlalchemy.org/

.. _`Peewee`: http://docs.peewee-orm.com/

.. _`SQLObject`: http://sqlobject.org/

.. _`Pony ORM`: https://ponyorm.com/



SQL Databases

.............



* `SQLite`_ is part of Python's standard library and provides a lightweight

  disk-based database that doesn’t require a separate server process and allows

  accessing the database using a nonstandard variant of the SQL query language.

* `MySQL`_ is one of the world’s most popular open source databases and has

  become a leading database choice for web-based applications. MySQL includes a

  standardized database driver for Python platforms and development.

* `PostgreSQL`_ is a powerful, open source object-relational database system

  with over 30 years of active development. Psycopg is the most popular

  PostgreSQL adapter for the Python programming language.

* `Oracle DB`_ is a relational database management system (RDBMS) from the

  Oracle Corporation. Originally developed in 1977, Oracle DB is one of the

  most trusted and widely used enterprise relational database engines.

* `Microsoft SQL Server`_ is a relational database management system developed

  by Microsoft. As a database server, it stores and retrieves data as requested

  by other software applications.



.. _`SQLite`: https://docs.python.org/3/library/sqlite3.html

.. _`MySQL`: https://dev.mysql.com/downloads/connector/python/

.. _`PostgreSQL`: http://initd.org/psycopg/

.. _`Oracle DB`: https://pypi.org/project/cx_Oracle/

.. _`Microsoft SQL Server`: https://pypi.org/project/pyodbc/



Other Databases

...............



* `Memcached`_ is free and open source, high-performance, distributed memory

  object caching system, generic in nature, but intended for use in speeding up

  dynamic web applications by alleviating database load.

* `Redis`_ is an open source, in-memory data structure store, used as a

  database, cache and message broker. It supports data structures such as

  strings, hashes, lists, sets, sorted sets with range queries, and more.

* `MongoDB`_ is a cross-platform document-oriented database program. Classified

  as a NoSQL database program, MongoDB uses JSON-like documents with

  schema. PyMongo is the recommended way to work with MongoDB from Python.

* `LMDB`_ is a lightning-fast, memory-mapped database. With memory-mapped

  files, it has the read performance of a pure in-memory database while

  retaining the persistence of standard disk-based databases.

* `BerkeleyDB`_ is a software library intended to provide a high-performance

  embedded database for key/value data. Berkeley DB is a programmatic toolkit

  that provides built-in database support for desktop and server applications.

* `LevelDB`_ is a fast key-value storage library written at Google that

  provides an ordered mapping from string keys to string values. Data is stored

  sorted by key and users can provide a custom comparison function.



.. _`Memcached`: https://pypi.org/project/python-memcached/

.. _`MongoDB`: https://api.mongodb.com/python/current/

.. _`Redis`: https://redis.io/clients#python

.. _`LMDB`: https://lmdb.readthedocs.io/

.. _`BerkeleyDB`: https://pypi.org/project/bsddb3/

.. _`LevelDB`: https://plyvel.readthedocs.io/



Reference

---------



* `DiskCache Documentation`_

* `DiskCache at PyPI`_

* `DiskCache at GitHub`_

* `DiskCache Issue Tracker`_



.. _`DiskCache Documentation`: http://www.grantjenks.com/docs/diskcache/

.. _`DiskCache at PyPI`: https://pypi.python.org/pypi/diskcache/

.. _`DiskCache at GitHub`: https://github.com/grantjenks/python-diskcache/

.. _`DiskCache Issue Tracker`: https://github.com/grantjenks/python-diskcache/issues/



License

-------



Copyright 2016-2023 Grant Jenks



Licensed under the Apache License, Version 2.0 (the "License"); you may not use

this file except in compliance with the License.  You may obtain a copy of the

License at



    http://www.apache.org/licenses/LICENSE-2.0



Unless required by applicable law or agreed to in writing, software distributed

under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR

CONDITIONS OF ANY KIND, either express or implied. See the License for the

specific language governing permissions and limitations under the License.



.. _`DiskCache`: http://www.grantjenks.com/docs/diskcache/