Quick Start Guide

Basic Usage

The ConditionalArgumentParser extends Python’s ArgumentParser to allow arguments that only appear based on other argument values.

Here’s a simple example:

from conditional_parser import ConditionalArgumentParser

# Create parser just like ArgumentParser
parser = ConditionalArgumentParser()

# Add regular argument
parser.add_argument("--use-regularization",
                   action="store_true",
                   help="Uses regularization if included.")

# Add conditional argument
parser.add_conditional(
    dest="use_regularization",    # Which argument to check
    cond=True,                    # Condition to meet
    "--regularizer-lambda",       # Argument to add when condition is met
    type=float,
    default=0.01,
    help="The lambda value for the regularizer."
)

Adding Conditional Arguments

The key method is add_conditional, which takes the following arguments:

1. dest: The destination (argument name) to check.

   The dest name is the same one that you would use to retrieve the argument value from the output of the parser. The nomenclature is intended to be consistent with argparse - when an argument is made in argparse, it is stored under the name designated in dest.

   import ArgumentParser from argparse
   parser = ArgumentParser()
   parser.add_argument("--use-regularization", action="store_true", help="Uses regularization if included.")
   args = parser.parse_args(["--use-regularization"])
   
   # Argument parser will store the value of "--use-regularization"
   # in the namespace under the name "use_regularization"
   # This is what the ``dest`` argument refers to.
   args.use_regularization  # True
   

   So if you want to add a conditional argument that is only added when the --use-regularization argument is included, you would set dest="use_regularization".

2. cond: The condition that dest has to meet for the conditional argument to be added.

   Either a value to compare against or a callable function that accepts the value of the dest as it’s single input argument and returns a boolean.

3. Additional arguments passed to add_argument as usual.

   These are the same arguments that you would pass to add_argument in argparse, and will be passed to the parser using parser.add_argument() whenever the condition is met.

Conditions can be:

• Simple equality: cond="value"

• Boolean check: cond=True

• Custom function: cond=lambda x: x in ["value1", "value2"]

Help Messages

Help messages show all possible conditional arguments with information about when they become available:

$ python script.py --help
usage: script.py [-h] [--use-regularization] [--regularizer-lambda LAMBDA]

optional arguments:
  -h, --help            show this help message and exit
  --use-regularization  Uses regularization if included.
  --regularizer-lambda  The lambda value for the regularizer. (Available when use_regularization=True)