Skip to main content

Addict dictionary enhanced with ability to track changes -- currently only supports field additions

Project description

addict-tracking-changes

Python versions Tests Status Coverage Status Flake8 Status

Documentation PyPI Downloads Downloads per week GitHub stars

Introduction

HEADS UP Before using the library carefully read the Known bugs and Caveats section below.

Originally, this repository was a fork of addict by Mats Julian Olsen. Overtime, it has substatially diverged in functionality and codebase that it made sense to breakout as its own repository.

The original addict: provides an alternative and succient interface to manipulate a dictionary. This is especially useful when dealing with heavily nested dictionaries. As example (taken from https://github.com/mewwts/addict) a dictionary created using standard python dictionary interface looks as follows:

body = {
    'query': {
        'filtered': {
            'query': {
                'match': {'description': 'addictive'}
            },
            'filter': {
                'term': {'created_by': 'Mats'}
            }
        }
    }
}`

can be summarized to following three lines:

body = Dict()
body.query.filtered.query.match.description = 'addictive'
body.query.filtered.filter.term.created_by = 'Mats'

This repo builds on original addict and adds contraptions to track key additions in the dictionary. This features comes in quite handy in building reactive webapps where one has to respond to all the changes made on the browser. Addict-tracking-changes is the underpinning data-structure in ofjustpy: a python based webframework build from justpy

The functions relevant to tracking changed history are: get_changed_history and clear_changed_history. The get_changed_history returns an iterator of XPath style paths like /a/b/c/e (see Demo example).

Usage and examples

Installation

This project is not on PyPI. Its a simple package with no third party dependency. Simply clone from github and include the source directory in your PYTHONPATH.

Demo example

from addict import Dict
body = Dict()
body.query.filtered.query.match.description = 'addictive'
body.query.filtered.filter.term.created_by = 'Mats'

for changed_path in body.get_changed_history():
    #<work with changed_path>

body.clear_changed_history()

Behaviour when values are instances of container types

addict works as expected when the values of keys are simple data types (such as string, int, float, etc.). However, for container types such as dict, list, tuples, etc. the behaviour is somewhat differs.

  • dicts are treated as opaque object just like simple data types
from addict import Dict

mydict = Dict()
mydict.a.b.c = {'kk': 1}
mydict.a.b.e = {'dd': 1}

for _ in mydict.get_changed_history():
    print(_) 

will print

/a/b/c
/a/b/e

and not

/a/b/cc/kk
/a/b/e/dd
  • lists are seen as container, i.e., get_changed_history will report path for each element of the list
from addict import Dict

mydict = Dict()

mydict.a.b = [1, [1]]
mydict.a.c = [2, [2, [3]]]

get_changed_history will report following paths:

/a/b/0,
/a/b/1/0,
/a/c/0,
/a/c/1/0,
/a/c/1/1/"
  • tuple, namedtuple, sets tuple behave same as dict and are treated as opaque data structure

Known bugs and Caveats

  1. Only tracks field additions. Deletions and updates are not tracked.
  2. freeze doesn't guards against deletions
  3. building dict from another dict as shown in following expression wont' work
cjs_cfg = Dict(other_dict, track_changes=True)

instead use

cjs_cfg = Dict(track_changes = True)
with open("other_dict.pickle", "rb") as fh:
    x = pickle.load(fh)
for _ in oj.dictWalker(x):
    oj.dnew(cjs_cfg, _[0], _[1])
  1. Use TrackedList for tracking nested list
from addict_tracking_changes import Dict, TrackedList


trackerprop = Dict(track_changes = True)
trackerprop.a.b = [1, TrackedList([1], track_changes=True)]
trackerprop.a.c = [2, TrackedList([2, TrackedList([3], track_changes=True)
                                   ], track_changes=True)
                   ]

which when asked changed history will output as follows:

print(list(trackerprop.get_changed_history()))

output

['/a/b/0', '/a/b/1/0', '/a/c/0', '/a/c/1/0', '/a/c/1/1/0']

APIs

API Description
Dict(*args, **kwargs) Initializes a new Dict object
to_dict(self) Converts the Dict object to a regular dictionary
freeze(self, shouldFreeze=True) Freezes the Dict object, making it immutable
unfreeze(self) Unfreezes the Dict object, making it mutable again
get_changed_history(self, prefix="", path_guards=None) Returns an iterator with the changed history of keys
clear_changed_history(self) Clears the changed history of the Dict object
set_tracker(self, track_changes=False) Sets or resets the change tracker for the Dict object

EndNotes

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

addict-tracking-changes-1.0.5.tar.gz (34.6 kB view details)

Uploaded Source

Built Distribution

addict_tracking_changes-1.0.5-py3-none-any.whl (10.2 kB view details)

Uploaded Python 3

File details

Details for the file addict-tracking-changes-1.0.5.tar.gz.

File metadata

File hashes

Hashes for addict-tracking-changes-1.0.5.tar.gz
Algorithm Hash digest
SHA256 7d5f8565a28cc40943c390dc926397d969acd3d83af2aba2b2d0bf07dded7e7e
MD5 3601e3a68e3ca90f0701c05c5858347e
BLAKE2b-256 077ae45e74991c44ff3caa92b28ae840cbad6b6a7dde1a11ff788147e4678a44

See more details on using hashes here.

File details

Details for the file addict_tracking_changes-1.0.5-py3-none-any.whl.

File metadata

File hashes

Hashes for addict_tracking_changes-1.0.5-py3-none-any.whl
Algorithm Hash digest
SHA256 800b298460024c7af9aaa4e0fa3b9bd56f87a736a79d382f17613786b4b52a38
MD5 1e15544a575962546cfcf12254bd079e
BLAKE2b-256 4990575717b5340472271efa17ead799b43b4b7cbc281f6f59bfdbb8a198dc90

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page