Have pyani use a user-controlled config file

[widdowquinn]

Summary:

The current config arrangement is hacky and awful. It would be better to have a user-controlled config file that specifies colour palettes, application locations and so on. This could be defined in environment variables and/or a config.py or config.json file.

An extension of this could be used for pipeline control of the pyani chain.

[baileythegreen]

Proposed format: .toml for readability.

Options to include:

• custom paths for executables†
• colour scheme used for plots†
• input and output directories
• modifiable parameters for the algorithms†
• algorithm itself?

† should have a default value on which to fall back

[widdowquinn]

Default values could continue to be defined as they are now, in a top-level config.py (or even TOML) file, e.g.:

import os

from dotenv import load_dotenv

BASEDIR = os.path.abspath(os.path.dirname(__file__))
load_dotenv(os.path.join(BASEDIR, ".env"))


class Config(object):
    # Database configuration
    TEST_DBPATH = os.path.join(BASEDIR, "testdb.sqlite")
    DBPATH = os.environ.get("SRE_DBPATH") or TEST_DBPATH

    # Flask app configuration
    SREDB_HOST = "127.0.0.1"
    SREDB_PORT = "5000"
    FLASK_SECRET_KEY = os.environ.get("FLASK_SECRET_KEY") or os.urandom(24)
    SQLALCHEMY_DATABASE_URI = os.environ.get(
        "SREDB_DBURL"
    ) or "sqlite:///" + os.path.join(BASEDIR, DBPATH)
    SQLALCHEMY_TRACK_MODIFICATIONS = False

values are then imported as a Config object:

from config import Config

which can be used to populated, for example, the argparse default values and/or anything that's not present in the user-defined TOML.

[baileythegreen]

I was assuming the default values would be in pyani_config.py. Anything similar thing would be fine, but defaults definitely don't need to be in a file the user would generally modify.

[widdowquinn]

I was assuming the default values would be in pyani_config.py.

Yes - it seems a sensible place.