Redis Extendable Efficient Middleware
Project description
REEM
Author: Trishul Nagenalli, updates by Kris Hauser
About
REEM (Redis Extendable Efficient Middleware) is a centralized middleware package for communication across distributed systems (e.g., robots). It is designed to be a single-package solution for passing information anywhere in the system while emphasizing ease of use and efficiency.
To make it easy, we chose to model information as a nested data structure that closely resembles python dictionaries. To the user, working with a database feels like working with a python dictionary. Out of the box, REEM supports communicating all native python types and numpy arrays.
To make it fast, we used Redis (an in-memory key-value database) running ReJSON (enabling Redis to store JSON data) as a central information store. To get maximum performance, we give users the power to control exactly how information is passed between the local program and Redis by defining their own encoder/decoder objects.
REEM currently offers two communication paradigms:
- get/set database
- publish-subscribe
Installation
To install the python package (and its dependencies), run
pip install reem
You will also need to have access to a Redis server with RedisJSON enabled. See the setup tutorial for step-by-step instructions o installing and configuring a compatible local Redis server.
Tutorials and API documentation
See example.py or the docs on read the docs.
Version history
0.1.2:
- bug fixes in pub/sub implementation.
- Added
get(key,default_value)
to subkey accessor analogous to dict'sget
method.
0.1.1:
- added
get()
andset()
to KeyValueStore for direct access to Rejson's JSON.GET / JSON.SET.
0.1.0: fork by Kris Hauser
- Can now access items by array index.
- Much easier to work with items like normal Python objects. Can:
- Can treat accesssors as variables using
read()
andwrite()
, e.g.,var = server['key']; var.read(); var.write(x)
. (See bug fix note below). - Delete items via
del server['key']
ordel server['key']['subkey']
(uses Rejson's JSON.DEL) - Increment/decrement values via
server['key']['subkey'] += 1
or-=
(uses Rejson's JSON.NUMINCRBY). Note: does not work for Numpy arrays. - Multiply/divide values via
server['key']['subkey'] *= 2
or/=
(uses Rejson's JSON.NUMMULTBY). Note: does not work for Numpy arrays. - Append to arrays via
server['key'].append(x)
orserver['key']['subkey'].append(x)
(uses Rejson's JSON.ARRAPPEND) - Multiple append to arrays via
server['key'] += [x,y,z]
orserver['key']['subkey'] += [x,y,z]
(uses Rejson's JSON.ARRAPPEND) - Get sizes of arrays/objects using
len(server['key'])
orlen(server['key']['subkey'])
(uses Rejson's JSON.OBJLEN or JSON.ARRLEN) - Get types of items using
server['key'].type()
orserver['key']['subkey'].type()
(uses Rejson's JSON.TYPE)
- Can treat accesssors as variables using
- KeyValueStore, PublishSpace, SilentSubscriber, and CallbackSubscriber are now in the global reem namespace. Also, they can be given a string host rather than separately having to create a RedisInterface object.
- KeyValueStore and PublishSpace are now thread safe. (Note: not thoroughly tested yet).
- Objects retrieved by ['key'] no longer get clobbered when accessing ['subkey']. E.g.,
kvs['topkey'] = {'subkey':{'foo':3}} a = kvs['topkey'] b = a['subkey'] b['foo'] = 4 print(a.read()) #prior version unexpected prints {'subkey':{'foo':3},'foo':4}, new version prints {'subkey'{'foo':4}} as expected.
- Slight performance improvements for deeply nested accesses
- Python 2 version automatically returns json objects with keys / values as str instead of unicode.
0.0.x: original from Trishul Nagenalli
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distributions
Built Distribution
File details
Details for the file reem-0.1.2-py2.py3-none-any.whl
.
File metadata
- Download URL: reem-0.1.2-py2.py3-none-any.whl
- Upload date:
- Size: 20.6 kB
- Tags: Python 2, Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/3.4.1 importlib_metadata/4.3.1 pkginfo/1.7.0 requests/2.25.1 requests-toolbelt/0.9.1 tqdm/4.61.0 CPython/3.7.10
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | d5a300d4e4a5c92f2a1804f5a92d3be0a6d5b704d3d7a847aadff8606b452e0e |
|
MD5 | 5619d1cc81e808d06d455d0d7bf64eb4 |
|
BLAKE2b-256 | 3c5f9dc361bf0d00cdcf3213e2174b5b654bdaf0a148ec6bec29dadf0a1ae7dc |