Skip to content

In-memory, Hash-mapping, Disk-writable, Thread-safe database.

License

Notifications You must be signed in to change notification settings

umihico/microdb

Repository files navigation

microdb

In-memory, Hash-mapping, Disk-writable, Thread-safe database.
PyPI version

Installation with pip

$ pip install microdb

This database works as list of dictionaries.

>>> import microdb
>>> mdb = microdb.MicroDB("testdata/mdb.json", partition_keys=['job', 'name'])
>>> mdb.pprint_all()
[{'job': 'clean', 'name': 'Alice', 'status': 'undone'},
 {'job': 'study', 'name': 'Alice', 'status': 'undone'},
 {'job': 'clean', 'name': 'Bob', 'status': 'undone'},
 {'job': 'study', 'name': 'Bob', 'status': 'undone'}]
>>> mdb.pprint_all_as_grid()
[['name', 'status', 'job'],
  ['Alice', 'undone', 'clean'],
  ['Alice', 'undone', 'study'],
  ['Bob', 'undone', 'clean'],
  ['Bob', 'undone', 'study']]

Only 'upsert' method add data, or overwrite if the "key" is already in use.

>>> mdb.upsert({'job': 'clean', 'name': 'Alice', 'status': 'finished'})
>>> mdb.upsert({'job': 'clean', 'name': 'Eve', 'status': 'undone'})
>>> mdb.pprint_all()
[{'job': 'clean', 'name': 'Alice', 'status': 'finished'}, # overwritten
 {'job': 'study', 'name': 'Alice', 'status': 'undone'},
 {'job': 'clean', 'name': 'Bob', 'status': 'undone'},
 {'job': 'study', 'name': 'Bob', 'status': 'undone'},
 {'job': 'clean', 'name': 'Eve', 'status': 'undone'}] # new

Key is tupled values of partition_keys. Use 'gen_key' method to see.

>>> # mdb = microdb.MicroDB("testdata/mdb.json", partition_keys=['job', 'name'])
>>> mdb.gen_key({'job': 'clean', 'name': 'Bob', 'what': 'ever'})
('clean', 'Bob')

'get' method and 'in' operator as same as dict

>>> {'job': 'clean', 'name': 'Bob', 'what': 'ever'} in mdb
True
>>> mdb.get({'job': 'study', 'name': 'Alice','whaat':'everr'})
{'status': 'undone', 'job': 'study', 'name': 'Alice'}
>>> mdb.get({'wrong': 'key'},'when_not_exist') # same as dict.get
'when_not_exist'

Write on disk when you want.

>>> mdb.save()

>>> mdb.save_as_grid() # good to reduce amount when all keys are common
$ cat testdata/mdb.json
[['status', 'job', 'name'], ['undone', 'clean', 'Alice'], ['undone', 'study', 'Alice'], ['undone', 'clean', 'Bob'], ['undone', 'study', 'Bob']]

>>> mdb.save('another_file_name.json') # you can write anywhere
$ cat another_file_name.json
[{'job': 'clean', 'name': 'Alice', 'status': 'undone'}, {'job': 'study', 'name': 'Alice', 'status': 'undone'}, {'job': 'clean', 'name': 'Bob', 'status': 'undone'}, {'job': 'study', 'name': 'Bob', 'status': 'undone'}]

All operation is already thread-safe by threading.rlock(). In case if you need to block while operations in a row,

>>> with mdb.lock():
...    bob_clean_status=mdb.get({'job': 'clean', 'name': 'Bob'})['status']
...    if bob_clean_status != 'finished':
...       mdb.upsert({'job': 'pay_penalty', 'name': 'Bob','status':'undone'})

About

In-memory, Hash-mapping, Disk-writable, Thread-safe database.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published