Skip to main content

A bafflingly simple, JSON-backend configuration manager for python programs.

Project description

confjson

Build Status

What is it?

A bafflingly simple, JSON-backed configuration manager for Python programs.

Why is it?

I kept implementing roughly the same thing from scratch in my various projects, so I figured it was time to do it properly.

User guide

The confjson.Config class is similar to a ChainMap, and works by means of two JSON files in the same directory:

  • default.config.json
    • The default config file is read-only as far as confjson is concerned. To add default settings, manually edit the file.
    • If the file does not exist, there are no defaults.
    • This file will typically be version controlled along with the code.
  • user.config.json
    • Any changes made to the config (and saved) at runtime end up in this file.
    • If the file does not exist, it will be created when changes are saved.
    • User settings take priority over default settings.
    • This file should probably be added to .gitignore or such. The names of both files are configurable. The above names are the defaults.

Initialization

The path given when initializing the Config object can be either a directory or a file. If it refers to a file, confjson will look for config files in the containing directory. The reason for this is that it enables the pattern of using __file__ to find config files in the same directory as the program.

config = confjson.Config(__file__)
config = confjson.Config(".")
config = confjson.Config(
	"./configs",
	user_config_filename=f"{os.getenv('username')}.cfg",
	default_config_filename="global.cfg",
)

Data access

Items in the confjson config are accessed as in a dict.

if "username" in config["user"]:
	do_something(config["user"]["username"])
	config["user"]["something_count"] += 1

It is also possible to access items as attributes, unless they share a name with a method of the Config class.

config.my_key = "my value"

Persistence

The load() method (re-)loads the Config object with values from the backing JSON files. Loading is also performed on initialization, so this is mainly for discarding changes.

config.load()

The save() method saves any changed or added items to user.config.json only.

config.save()

Version history

1.3.0

  • Made filenames of both user.config.json and default.config.json configurable.
  • Added use_placeholders argument to Config class. When True, doesn't raise KeyError on attempted access to a nonexistent key, instead returning a placeholder. Assignments of keys to a placeholder will propagate up through the hierarchy and add real dicts as needed.

1.2.2

  • Added keys(), get(key, default) and __contains__(key) to ConfigItemProxy to fix a few problems caused by the previous change.

1.2.1

  • Improved attribute-style access and enabled mixed style access.

1.2.0

  • Enabled attribute-style access to config items.

1.1.0

  • Fixed handling of nested dicts.
  • Added check to prevent insertion of JSON-incompatible values right away rather than fail on save.
  • Added some more dict-like dunder methods.

1.0.1

  • Changed __init__ to also load the config files.

1.0.0

  • Initial release.

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

confjson-1.3.0.tar.gz (9.2 kB view hashes)

Uploaded Source

Built Distribution

confjson-1.3.0-py3-none-any.whl (13.3 kB view hashes)

Uploaded Python 3

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