Skip to main content

Making a CLI Application by wrapping

Project description

wrapp: Making a CLI Application by wrapping

wrapp helps you to make a CLI application without boilerplate of logging & argparse.

INSTALL

pip install wrapp

USAGE

TL;DR

  1. Create your Python script under a few rules. To do so, start with wrapp.new.

    wrapp.new > YOURS.py
    
  2. Edit YOURS.py as you like.

  3. Then you can run your Python script as an CLI app.

    wrapp YOURS.py
    

That's it. Let's enjoy !

Create your Python script under a few rules

By using wrapp.new,

wrapp.new > YOURS.py

you can get a template Python file named YOURS.py.

wrapp.new outputs template code at stdout.

$ cat YOURS.py
#!/usr/bin/env python3
from logging import getLogger


LOG = getLogger(__name__)


def set_logger(parent_name):
    global LOG
    LOG = getLogger(f'{parent_name}.{__name__}')


def add_arguments(parser):
    group = parser.add_argument_group(__name__)
    ...


def main(args):
    ...


# code below is an option.
# if you want to run it as an normal Python script
# (`python THIS_SCRIPT.py`), uncomment it.
# if __name__ == '__main__':
#     import wrapp
#     wrapp.main(add_arguments, main, set_logger)

This template has 1 global variable and 3 functions; LOG, set_logger(), add_arguments(), main().

LOG is the logger in this file. It may be changed in set_logger().

set_logger() is assumed to be called by wrapp. Python logging module has an implicit rule that when a logger name is delimited by ., corresponding loggers form parent-child relationship. e.g. wrapp.module named logger is child of the logger named wrapp. And when you configure the wrapp logger, wrapp.module logger is also configured as same as the wrapp logger.

If your script import some other your modules and you want to configure these modules' loggers as the same settings as this, you can call recursively set_logger().

import your_module2

def set_logger(parent_name):
    global LOG
    LOG = getLogger(f'{parent_name}.{__name__}')
    your_module2.set_logger(parent_name)

See also tests/template.py and tests/sub/template2.py.

add_arguments() is the function to set program arguments. add_arguments() takes 1 argument; parser which is assumed as an instance of argparse.ArgumentParser. However, I don't use a type hint now because of redusing import statements.

add_arguments() is also available recusive calling. See also tests/template.py and tests/sub/template2.py.

main() function is the entry point.

When you run wrapp YOURS.py,

  1. [Optional] wrapp call set_logger() and also set its logger.

  2. The program arguments are parsed as defined in add_arguments() and stored in the variable named args.

  3. All program arguments and options are output to console.

  4. Finally, the main(args) is called.

As shown above, wrapp assumes your Python file contains add_arguments(parser) and main(args). set_logger() is optional.

Run your Python script as an CLI app

Assume your Python script is YOURS.py.

wrapp YOURS.py --your-options ...

That is, just replace python to wrapp. Then you can keep your script simple:

  • if __name__ == '__main__': is not needed.
  • Also you don't need any noisy modules such as argparse, from argparse import ..., from logging import ....

"I want to run it as python my_script.py"

If you want to run your code as an usual Python way, just uncomment last 3 lines of the template.

Then you can run like

python my_script.py --some-option ...

Debugging

When you want to debug your script, run the code like this.

python3 -m pdb -m wrapp YOURS.py --your-options ...

Then the debugging mode (pdb) will be started.

FEATURES

  • No dependencies. wrapp only depends on Python standard libraries.

  • One file. If you don't need to install other packages at all, just copy src/wrapp/wrapp.py.

    $ cp PATH/TO/wrapp_repo/src/wrapp/wrapp.py ./wrapp
    $ chmod u+x wrapp
    $ ./wrapp.new > YOURS.py
    $ ./wrapp YOURS.py
    
  • It's like python-fire. But for wrapp, you don't need to import any other module in your Python code.

  • It's trivial but you also run your script without the extension; wrapp YOURS.

LICENSE

MIT License.

BACKGROUNDS

As I wrote tons of Python CLI applications, I noticed that,

  • argparse is the best practice to add my program command options.
  • logging is not bad if I modify something (format, ...).
  • But I noticed that there are many similar lines in my applications. And they make my code more dirty.

Here is my application code pattern. Please note that there is nothing infomative.

#!/usr/bin/env python3
from argparse import ArgumentParser, ArgumentDefaultsHelpFormatter
from logging import getLogger
from pathlib import Path
import logging.config


_LOG = getLogger(__name__)


def add_arguments(parser):
    parser.add_argument(
            'in_file', type=Path,
            help='An input file.')
    parser.add_argument(
            '--out-dir', '-d', type=Path, default=None,
            help='A directory.')


def _main(args):
    _LOG.debug('debug')
    _LOG.info('info')
    _LOG.warning('warning')
    _LOG.error('error')
    _LOG.critical('critical')
    ...


def _parse_args():
    parser = ArgumentParser(formatter_class=ArgumentDefaultsHelpFormatter)
    parser.add_argument('-i', '--input-files', nargs='*', help='input files.')
    args = parser.parse_args()
    logging.config.fileConfig('logging.conf')
    for k,v in vars(args).items():
        _LOG.info('{}= {}'.format(k, v))
    return args


def _print_args(args):
    for k, v in vars(args).items():
        _LOG.info(f'{k}= {v}')


if __name__ == '__main__':
    parser = ArgumentParser(formatter_class=ArgumentDefaultsHelpFormatter)
    add_arguments(parser)
    args = parser.parse_args()
    logging.config.fileConfig('logging.conf')
    _print_args(args)
    _main(args)

So I decided to separate it to 2 files; one is the contents only and the other is a wrappter to make any Python files an CLI app.

Finally, I can make the above code much more simple,

#!/usr/bin/env python3
from logging import getLogger
from pathlib import Path


_LOG = getLogger(__name__)


def add_arguments(parser):
    parser.add_argument(
            'in_file', type=Path,
            help='An input file.')
    parser.add_argument(
            '--out-dir', '-d', type=Path, default=None,
            help='A directory.')


def main(args):
    _LOG.debug('debug')
    _LOG.info('info')
    _LOG.warning('warning')
    _LOG.error('error')
    _LOG.critical('critical')
    ...

It's similar to python-fire.

But when I used the fire, I have to insert from fire import Fire and Fire(your_func). I'd like to remove even such a few code.

Then I'm completly free from noisy modules / code !

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

wrapp-0.5.1.tar.gz (4.7 kB view details)

Uploaded Source

Built Distribution

wrapp-0.5.1-py3-none-any.whl (5.8 kB view details)

Uploaded Python 3

File details

Details for the file wrapp-0.5.1.tar.gz.

File metadata

  • Download URL: wrapp-0.5.1.tar.gz
  • Upload date:
  • Size: 4.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.4.0 CPython/3.9.2 Linux/5.10.0-21-amd64

File hashes

Hashes for wrapp-0.5.1.tar.gz
Algorithm Hash digest
SHA256 263e1e97a150a554c2c83515e3444a495bb472e753ecc98f80ffa8917be0c40e
MD5 cdb92141add31a49651b23cae4db021c
BLAKE2b-256 265cf7dfe8f5c4cf8e76506953f75c9ccfeaaedc626506ddef5d0324432c8982

See more details on using hashes here.

File details

Details for the file wrapp-0.5.1-py3-none-any.whl.

File metadata

  • Download URL: wrapp-0.5.1-py3-none-any.whl
  • Upload date:
  • Size: 5.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.4.0 CPython/3.9.2 Linux/5.10.0-21-amd64

File hashes

Hashes for wrapp-0.5.1-py3-none-any.whl
Algorithm Hash digest
SHA256 79797b45cb982ba74dbbefb6c1847b68e04f2d92c33fab5fafd0dc8d0be15d83
MD5 94dab049e92a577d55602f9dbb42a099
BLAKE2b-256 fe2401683d482fcaffc321f15dfc3a11b42cea9728523e95bb76b46dc49c6d60

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page