Skip to content

Instantly share code, notes, and snippets.

@pylemon
Forked from mayli/zipdb.py
Created May 30, 2012 05:10
Show Gist options
  • Save pylemon/2833894 to your computer and use it in GitHub Desktop.
Save pylemon/2833894 to your computer and use it in GitHub Desktop.
python: Use a zipfile store a dict like k-v database
#!/usr/bin/env python
# -*- coding: utf-8 -*-
#
# zipdb.py
# Use a zipfile store a dict like k-v database.
# Known bug: duplicate key(filenames) allowed
#
# Copyright 2012 mayli <[email protected]>
#
import zipfile
class ZipDB(object):
#dbfile=None
def __init__(self,filename='tmp.zipdb',flags='a',compression=True,allowZip64=True):
self.dbfile=zipfile.ZipFile(filename,
flags,compression=[zipfile.ZIP_STORED,zipfile.ZIP_DEFLATED][compression],
allowZip64=allowZip64)
def __setitem__(self, key, item):
"""
Method to put items into the database file.
"""
return self.dbfile.writestr(key,item)
def __getitem__(self, key):
"""
Method to get items from the databasde file using its key
"""
return self.dbfile.read(key)
def __len__(self):
"""
Returns the row count of the database file
"""
return len(self.dbfile.namelist())
def close(self):
"""
Closes the database file
"""
self.dbfile.close()
def keys(self):
"""
Returns a list of keys in the database file
"""
return self.dbfile.namelist()
def main():
def benchmark(db):
import time
s='.'*4*1024
count=100000
t1=time.time()
for i in xrange(count):
db[str(i)]=s
print count/(time.time()-t1),'Insertion per second'
t1=time.time()
for i in xrange(count):
tmp=db[str(i)]
print count/(time.time()-t1),'Reads per second'
# Usage
d=ZipDB()
# Benchmark
benchmark(d)#compressed 8880/14862/20MB on my laptop
d.close()
d=ZipDB('tmp_uncompress.zipdb',compression=False)
benchmark(d) #uncompressed 13802/18565/398MB on my laptop
d.close()
return 0
if __name__ == '__main__':
main()
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment