class-cli 0.3.3

Converts a python class into a CLI program

Class CLI

A Python module for converting a Python class into a CLI program.

Using the module encourages for well documented code while not demanding it. It allows for separation in implementation between method execution and it's argument validations.

Capabilities

• Full command and argument auto-completion and suggestions
• Basic type validation for entered arguments while typing
• Full input line validation
• Useful help messages for all methods and arguments
• Can combine CLI Class Objects together into a single Program
• Logging support
• Command history
• Execution of commands from a text file, line by line

---

---

API

The module exposes an API in the form of decorators.

First import the CLI class

from class_cli import CLI

create an instance and wrap a class with decorators the it exposes

cli = CLI()

@cli.Program()
class MyClass:

    @cli.Operation()
    def show(self):
        return "I'm a CLI Program!"

Run the main method of the CLI

if __name__ == "__main__":
    MyClass().CLI.main()

When running the script without arguments, It will open the CLI for user input:

MyClass> show
I'm a CLI Program!

MyClass> |

---

These are the available decorators

@Program

Program(name=None, version=None, description=None, log=None, style=None, verbosity=logging.INFO)

    a class decorator that defines the CLI program.
    Instantiation of the wrapped user class can be used as normal python code, accessing all it's attributes.
    It also exposes the CLI interface with an added attribute named 'CLI'

    * Only a single Program can be declared for every CLI instance

    @name           The name of the CLI program.                                                (Default is the class name)
    @version        The version of the CLI program.                                             (Default is a CLI without versioning)
    @description    The description of the CLI program.                                         (Default is the class documentation)
    @log            A path to a log file.                                                       (Default is no log file)
    @style          A dictionary that overrides the styling of the CLI for the given keys       (Keys: CLI.STYLE)
    @verbosity      Determines the verbosity level on the logger. Use None to silence STDOUT    (Keys: logging module)

@Operation

Operation()

    a method decorator that defines the execution code of an operation in the CLI

@Setting

Setting(initial_value, updates_value=True)

    a method decorator that creates a setting value for the CLI with name equals to the method name.
    It defines the execution code for setting the value into the created setting.

    @initial_value      An initial value that the setting will hold after class initialization
    @updates_value      Whether or not calling this method automatically updates the inner setting value

@Delegate

Delegate(reuse=True)

    a method decorator that delegates control to another CLI Object instance by exposing an access point. 
    The wrapped method does not accept any arguments and returns the CLI Object instance to delegate control to.
    * This allows for easy integration and combination of CLI Objects into a single CLI Program

    @reuse      Whether or not the returned CLI Object should be reused for later calls

@Validation

Validation()

    A method decorator that defines a validation to be performed on an execution (Operation / Setting / Delegation)
    Holds the same signature as the execution it is validating and raises an exception for invalid arguments.
    * A single execution can have multiple Validations

---

API Example

In this example, we are wrapping a class, that holds a Setting named value, and exposes a method called show that prints it to the screen.

from class_cli import CLI

cli = CLI()
@cli.Program()
class MyClass:

    @cli.Operation()
    def show(self): 
        if self.CLI.name is None:
            return "I'm a CLI Program!"
        else:
            return "My name is {}!".format(self.CLI.name)

    @cli.Setting()
    def name(self, str):
        return str

if __name__ == "__main__":
    MyClass().CLI.main()

In the main method we called MyClass().CLI.main() which checks for sys.argv for input and executes the commands using the instance.

$> python3 MyClass.py show
I'm a CLI Program!

If we execute the script without arguments, it will open the CLI for user input:

$> python3 MyClass.py
MyClass
    To exit, enter one of the following: 'q' 'quit' 'exit'
    To read commands from a file, enter one of the following: '.r' '.read'
    To access the program settings, enter one of the following: '.set' '.setting'\

    At any time, add '-h' flag to the command for help.

MyClass> show
I'm a CLI Program!

MyClass> .setting name John
=John

MyClass> show
My name is John!

MyClass> |

You can also open the CLI directly by calling run(*args) instead of main

 if __name__ == "__main__":
     MyClass().CLI.run()

Other commands:

run_line(line) which executes a single line of String

execute(*args) which executes the same as run, but does not print out log outputs.

Instance API

You can call the methods of a CLI Object directly without opening a CLI session:

>>> instance = MyClass()

>>> instance.show()
'I'm a CLI Program!'

>>> instance.name("Bob")
'Bob'

>> instance.show()
'My name is Bob!'

---

---

Delegation

The library treats Class CLIs as building blocks that can integrate into a larger program. This encourages for code separation between the components of a program.

This is done by wrapping a class method with the @Delegate decorator.

cli = CLI()
@cli.Program()
class Main:
    @cli.Operation()
    def show(self): return "Main CLI"

    @cli.Delegate()
    def inner(self):
        inner_cli = CLI()
        @cli.Program()
        class Inner:
            @cli.Operation()
            def show(self): return "Inner CLI"
        return Inner()

if __name__ == "__main__":
    Main().CLI.main()

You can access the inner CLI by calling the inner method:

>>> Main()
<__main__.Main object at 0x03DBF5D0>

>>> Main().inner()
<__main__.Main.inner.<locals>.Inner object at 0x03B6D750>

From the CLI, you can pass commands to it by prefixing with the inner method, or open it directly for commands

Main> show
Main CLI

Main> inner show
Inner CLI

Main> inner

Main\Inner> show
Inner CLI

Main\Inner> q

Main> |

---

---

Logging

To log messages, the cli holds a Logger instance. to access it, use the CLI.log keyword

def method(self):
    self.CLI.log.info("This is an information Line")
    self.CLI.log.warning("This is a Warning")
    self.CLI.log.error("This is an Error")
    self.CLI.log.debug("This line is shown only when Program decorator is called with 'debug=True'"")

MyClass> method
This is an information Line
[WARNING] This is a Warning
2019-10-10 17:51:21,807
[ERROR] This is an Error
2019-10-10 17:51:21,808
[DEBUG][test:12] This line is shown only when Program decorator is called with 'debug=True'

MyClass> |

You can see documentation for the Logger object here

---

---

Help Messages

The CLI makes use of user code documentation in order to provide help messages to the user providing the --help/-h flag in the input. Using the --help flag will display a message relevant to the user input. Using it alone will display a usage message for the entire program while adding it after an Operation or Setting name will display usage information for that Operation or Setting.

@cli.Operation()
def method(self):
    """
    A description of the method
    """
    pass

CLI> method -h
usage: CLI method [-h]

=========================
A description of the method

=========================
optional arguments:
-h, --help show this help message and exit\

CLI>|

Any Validations for the method will be added to the description

@cli.Validation()
def method(self, arg):
    """
    cannot perform operation when disabled
    """
    if self.disabled:
        raise Exception("Cannot perform operations when disabled")

CLI> method -h
usage: CLI method [-h]

=========================
A description of the method
* cannot perform operation when disabled
=========================

optional arguments:
-h, --help show this help message and exit\

CLI>|

If there are arguments for the method, you can add descriptions for them inside the method documentation with a @ prefix as demonstrated below.

@cli.Operation()
def method(self, arg):
    """
    A description of the method

    Accepts:
        @arg    arg description
        @arg    another line of the description
    """
    pass

CLI> method -h
usage: CLI method [-h] arg

=========================
A description of the method

Accepts:
  @arg arg description
  @arg another line of the description

  * cannot perform operation when disabled

=========================
positional arguments:
 arg    =========================
       arg description
       another line of the description
       Type: <class 'str'>
       =========================
optional arguments:
-h, --help show this help message and exit

CLI>|

---

---

Type Annotations

Pythons type annotations are used by the CLI in order to use type validation and conversion.

Type Conversion

Since the user input is a String, by default the arguments passed to a method are String

Default Behavior:

@cli.Operation()
def method(self, arg):
    print(type(arg), ';', arg)

CLI> method value
<class 'str'> ; value

CLI>|

Basic Annotation: Using annotations, we can provide a type to tell the CLI what is expected, and use it to convert the String into its proper type. The CLI accepts annotations which are a Python Callable that receive a String and returns it converted to the desired type.

@cli.Operation()
def method(self, arg:int):
    print(type(arg), ';', arg)

CLI> method 17
<class 'int'> ; 17

CLI>|

Iterable Annotation: You can also use an Iterable as an annotation to specify a set of options.

@cli.Operation()
def method(self, value:[1, "value", (0,0)]):
    print(type(value), ';', value)

CLI> method 1
<class 'int'> ; 1

CLI> method value
<class 'str'> ; value

CLI> method '(0, 0)'
<class 'tuple'> ; (0, 0)

CLI>|

• Since the string representation of the items in the List are used to select the value from it, The String representations of the Items need to be unique.

Type Validation

During typing of input, The CLI will tip the user for the expected inputs, as well as block the user from entering invalid types into a command.

@cli.Operation()
def method(self, arg1, arg2:int=0, arg3:[1, -1]=1, **extras):
    pass

CLI> method value number|


invalid literal for int() with base 10: 'number'
arg1 arg2[=0] arg3[=1] {*extras} : {<class 'int'>}
CLI> method value 0 2|


'2' is not a valid option
arg1 arg2[=0] arg3[=1] {*extras} : {<value from: 1, -1>}
CLI> method value 0 -1 arg|


'arg' must be in the format [key]=[value]
arg1 arg2[=0] arg3[=1] {**extras} : {<class 'dict'> [items=0]}
CLI> method value 0 -1 arg=val|



arg1 arg2[=0] arg3[=1] {**extras} : {<class 'dict'> [items=1]}

---

Auto Completion

The CLI can auto-complete command names, file paths and argument values.

@cli.Operation()
def method(self, arg1, arg2:int=0, arg3:[1, -1]=None, *extras):
    pass

@cli.Setting()
def name(self, arg1):
    return arg1

Pressing 'TAB' will open up all available suggestions

CLI> |
   | method |
   | .setting  |
   |  .read   |

CLI> .setting |
      | name |

When browsing suggestions, you can see the expected arguments at the bottom:

CLI> method|
   | method |
   |  .setting  |
   |  .read   |

arg1 arg2[=0] arg3[=None] {*extras}

When the expected argument is an Iterable, the CLI will suggest all the Iterable items

CLI> method value 0 |
          |  1 |
          | -1 |

arg1 arg2[=0] arg3[=None] {*extras} : {1, -1}

If you are providing your own Callable for annotation, you can make use of the auto complete mechanism by implementing _complete_(self, keyword)

---

Implementing a Custom Annotation

You can use any callable(str) as a type annotation for the CLI.

Representation

The CLI displays the callable to the user as its String representation. To set the description, override the _str_(self) method

def __str__(self):
    return "A Custom Type"

Callable

The annotation must be a callable that accepts a string and returns a list of strings. The callable returns all the suggestions relevant to the given keyword

def __call__(self, key):
    if key in self.options_dict:
        return self.options_dict[key]

Validation

If the given string is not a valid option, you can throw an exception inside the _call_ method

def __call__(self, str):
    if key in self.options_dict:
        return self.options_dict[key]
    throw Exception("'{}'' is not a valid option".format(str))

This will block the user from entering a non existing key

Autocomplete Suggestion

To make use of the auto complete mechanism, implement the _complete_(self, keyword) method

def __complete__(self, str):
    return [s for s in self.suggestions if s.startswith(str)]

Dynamic Iterable Annotation

We've seen that you can use an iterable as an annotation for a variable

@cli.Program()
class MyClass:
    @cli.Operation()
    def oper(self, value:[True, False]):
        print("You have selected", value)

In this example, we are using a list to specify that only boolean values are accepted.

What if we want to use a list that is dynamic and can change throughout the run of the program. To do that, we need to define the list outside the class scope in order to access it in method signature

values = [1, 2, 3]

@cli.Program()
class MyClass:
    @cli.Operation()
    def oper(self, value:values):
        print("You have selected", value)

    @cli.Operation()
    def add_value(self, value:int):
        values.append(value)

The oper operation will now accept values in [1, 2, 3]. We can also use the add_value operation in order to add more options to be selected

For better code structure, we can use the with statement with the CLI object

with CLI() as cli:
    values = [1, 2, 3]

    @cli.Program()
    class MyClass:
        @cli.Operation()
        def oper(self, value:values):
            print("You have selected", value)

        @cli.Operation()
        def add_value(self, value:int):
            values.append(value)

This will make it clear that the values variable is associated with the CLI.