pyutil-cfg

python utilility for configurations.

This package helps parsing .toml configurations and json-like configurations in .ini.

Usage

import pyutil_cfg as cfg

logger, config = cfg.init(name, filename)

File Formats

.toml

For the .toml files, the format follows TOML specification.

You can check production.tmpl.toml or test/data for more details.

.ini

For the .ini files, you can check production.tmpl.ini or test/data for more details.

Assume that you have the following development.ini:

[demo]
VAR_INT = 1
VAR_BOOL = true
VAR_DICT = {"A": 1, "B": "a"}
VAR_LIST = [
    {"A": 2, "B": "b"},
    {"A": 3, "B": "c"},
    {"A": 4, "B": "d"}]

Then with the following code:

import pyutil_cfg as cfg

logger, config = cfg.init('demo', 'development.ini')

logger is a logger with name = 'demo'

config ia as follow:

config = {
    "VAR_INT": 1,
    "VAR_BOOL": true,
    "VAR_DICT": {"A": 1, "B": "a"},
    "VAR_LIST": [
        {"A": 2, "B": "b"},
        {"A": 3, "B": "c"},
        {"A": 4, "B": "d"}
    ],
}

Advanced Usage

Separated log configuration filename

import pyutil_cfg as cfg


logger, config = cfg.init(name, filename, log_name, log_filename)

Additional customized config parameters (specified as dict)

import pyutil_cfg as cfg

extra_params = {'test_params': 'test'}

logger, config = cfg.init(name, filename, extra_params=extra_params)

Ok if extra parameters are already in the configuration file (default: True)

It is usual that we would like to temporarily change the configuration settings through command-line variables.

Because usually we still want to continue the process even if the extra parmeters are accidentally in the configuration file. Only warning logs if this parameter is set as False and some extra parameters are already in configuration file.

import pyutil_cfg as cfg

logger, config = cfg.init(name, filename, is_extra_params_in_file_ok=True)

Skip extra parameters if already in the configuration file. (default: False)

Skip extram parameters if they are already in the configuration file.

import pyutil_cfg as cfg

logger, config = cfg.init(name, filename, is_skip_extra_params_in_file=True)

Show configuration before returning. (default: False)

Setup the log-level to show the computed configurations before return. None as no-show.

Do not show the configuration before return.

import pyutil_cfg as cfg

logger, config = cfg.init(name, filename)

Show the configruration if the logger is set to logging.DEBUG

import pyutil_cfg as cfg

logger, config = cfg.init(name, filename, show_config=logging.DEBUG)

Show the configuration if the logger is set to logging.INFO or logging.DEBUG.

import pyutil_cfg as cfg

logger, config = cfg.init(name, filename, show_config=logging.INFO)

Misc

• Reason retaining pyutil_ prefix: cfg is an extremely common name. Retaining pyutil_ to avoid naming conflict.
• Reason not including .json: not a good file format for config.
• Reason not including .yaml: not the standard library.