This is mainly just me playing around with git, reST, python packages, and cython; but it is also a fully functioning key value store with both memory and disk implementations using sqlite and some other key value stores.

All storage back ends have dict-like interfaces as well as a default LRU cache based off of OrderedDict.

python 3.5+

build dependencies

• cython
• ujson (optional)

Install with:

pip install https://github.com/Bigjaked/py-kv-store/raw/master/dist/persist_kv_store-0.93.tar.gz

• Vedis (memory and disk)
• sqlite3 (memory and disk)
• Python OrderedDict subclass

Here is an example of how each benchmark is run

from timeit import default_timer

# time writes
set_start = default_timer()
for i in range(0, its):
    db['key-{i}'.format(i=i)] = i
set_time = default_timer() - set_start

# time reads
get_start = default_timer()
for i in range(0, its):
    a = db['key-{i}'.format(i=i)]
get_time = default_timer() - get_start

Benchmark was run on a MSI GS60 32GB DDR4, i7-6700HQ Main storage is a Samsung 950 PRO 512GB