pyclilib 0.3.1

Build CLI Tools with ease

pyclilib

Python Library to create CLI tools using Subject-Verb-Object pattern

# [my-namespace] not yet implemented
my-cmd [my-namespace] my-action my-resource [pos_args] [--options]

Quickstart 1: CLI without Namespaces

1. pip install pyclilib
2. Create a hello world CLI

import clilib

@clilib.decorator.resource
class MyResource():
    @clilib.decorator.verb
    def get(self):
        print("Hello world")

clilib.init("mycli")
clilib.run("mycli")

3. Run your program

$ python example.py get my-resource
Hello world

Quickstart 2: CLI with args

1. pip install pyclilib
2. Create a hello world CLI with args.

import clilib

# Under the hood, clilib uses argparse.
# the @arg decorator uses its "add_argument" method

@clilib.decorator.resource
@clilib.decorator.arg("--environment", "-e", type=str, default="dev")
class MyResource():
    @clilib.decorator.verb
    @clilib.decorator.arg("name", type=str)
    def get(self, args):
        print(f"Get: {args.name}, {args.environment}")

    @clilib.decorator.verb
    def list(self, args):
        print(f"List: {args.environment}")


clilib.init("mycli")
clilib.run("mycli")

3. Run your program

$ python example.py get my-resource Foo -e stage
Get: Foo, stage

$ python example.py list my-resource
List: dev

Quickstart 3: CLI with Namespaces (not implemented yet)

1. pip install clilib
2. Create the runner, example.py

import clilib

clilib.init("mycli")
clilib.run("mycli")

3. Namespaces are sub-packages of an expected local namespace package

$ mkdir -p namespace namespace/my-namespace
$ touch namespace/__init__.py namespace/my-namespace/__init__.py

# namespace/my-namespace/__init__.py
import clilib

@clilib.decorator.resource
class MyResource():
    @clilib.decorator.verb
    def get(self):
        print("Hello World")

4. Run your program

$ python example.py my-namespace get my-resource
Hello World

Notes

• You can either have cli namespace action resource OR cli action resource in the future

Roadmap

• Global Config -- Potential
• Namespace
• Conditional Args
• CLI without resource ex. git init

TODO

• How to make log level configurable? Since decorators run during "compile"(?) time, setting the level through clilib.run("foo", logging.DEBUG) will set the level after the decorators do their thing. Same story with clilib._log_level = logging.DEBUG