Skip to main content

Automatically generate shell tab completion scripts for python CLI apps

Project description

  • What: Automatically generate shell tab completion scripts for python CLI apps

  • Why: Speed & correctness. Alternatives like argcomplete and pyzshcomplete are slow and have side-effects

  • How: shtab processes an argparse.ArgumentParser object to generate a tab completion script for your shell

Features

  • Outputs completion for

    • bash

    • zsh

  • Supports

  • Supports arguments, options and subparsers

  • Supports path completion

Installation

  • pip install shtab

  • conda install -c conda-forge shtab

Usage

The only requirement is that external CLI applications provide an importable argparse.ArgumentParser object (or alternatively an importable function which returns a parser object). This may require a trivial code change.

Once that’s done, simply add eval "$(shtab --shell=bash your_cli_app.your_parser_object)" to ~/.bash_completion (assuming bash).

Below are various examples of enabling shtab’s own tab completion scripts.

bash

# Install locally
echo 'eval "$(shtab --shell=bash shtab.main.get_main_parser)"' \
  >> ~/.bash_completion

# Install locally (lazy load for bash-completion>=2.8)
echo 'eval "$(shtab --shell=bash shtab.main.get_main_parser)"' \
  > "${BASH_COMPLETION_USER_DIR:-${XDG_DATA_HOME:-$HOME/.local/share}/bash-completion}/completions/shtab"

# Install system-wide
echo 'eval "$(shtab --shell=bash shtab.main.get_main_parser)"' \
  | sudo tee "$(pkg-config --variable=completionsdir bash-completion)"/shtab

# Install system-wide (legacy)
echo 'eval "$(shtab --shell=bash shtab.main.get_main_parser)"' \
  | sudo tee "$BASH_COMPLETION_COMPAT_DIR"/shtab

# Install once (will have to re-run if the target's CLI API changes,
# but doesn't need target to always be in $PYTHONPATH)
shtab --shell=bash shtab.main.get_main_parser --error-unimportable \
  | sudo tee "$BASH_COMPLETION_COMPAT_DIR"/shtab

zsh

Note that zsh requires completion script files to be named _{EXECUTABLE} (with an underscore prefix).

# Install once (will have to re-run if the target's CLI API changes,
# but doesn't need target to always be in $PYTHONPATH)
shtab --shell=zsh shtab.main.get_main_parser --error-unimportable \
  | sudo tee /usr/local/share/zsh/site-functions/_shtab

To be more eager, place the generated script somewhere in $fpath. For example, add these lines to the top of ~/.zshrc:

fpath=($fpath ~/.local)
shtab --shell=zsh shtab.main.get_main_parser > ~/.local/_shtab

Examples

See the examples/ folder for more.

Any existing argparse-based scripts should be supported with minimal effort. For example, starting with this existing code:

#!/usr/bin/env python
import argparse

def get_main_parser():
    parser = argparse.ArgumentParser(prog="<MY_PROG>", ...)
    parser.add_argument(...)
    parser.add_subparsers(...)
    ...
    return parser

if __name__ == "__main__":
    parser = get_main_parser()
    args = parser.parse_args()
    ...

Assuming this code example is installed in MY_PROG.command.main, simply run:

# bash
echo 'eval "$(shtab --shell=bash MY_PROG.command.main.get_main_parser)"' \
  >> ~/.bash_completion

# zsh
shtab --shell=zsh -u MY_PROG.command.main.get_main_parser \
  | sudo tee /usr/local/share/zsh/site-functions/_MY_PROG

FAQs

Not working? Make sure that shtab and the application you’re trying to complete are both accessible from your environment.

“Eager” installation (completions are re-generated upon login/terminal start) is recommended. Naturally, shtab and the CLI application to complete should be accessible/importable from the login environment. If installing shtab in a different virtual environment, you’d have to add a line somewhere appropriate (e.g. $CONDA_PREFIX/etc/conda/activate.d/env_vars.sh).

By default, shtab will silently do nothing if it cannot import the requested application. Use --error-unimportable to noisily complain.

Advanced Configuration

See the examples/ folder for more.

Complex projects with subparsers and custom completions for paths matching certain patterns (e.g. --file=*.txt) are fully supported (see iterative/dvc:command/completion.py for example).

Add direct support to scripts for a little more configurability:

#!/usr/bin/env python
import argparse
import shtab  # for completion magic

def get_main_parser():
    parser = argparse.ArgumentParser(prog="pathcomplete")
    parser.add_argument(
        "-s", "--print-completion-shell", choices=["bash", "zsh"]
    )
    parser.add_argument(
        "--file",
        choices=shtab.Optional.FILE,  # file tab completion, can be blank
    )
    parser.add_argument(
        "--dir",
        choices=shtab.Required.DIRECTORY,  # directory tab completion
        default=".",
    )
    return parser

if __name__ == "__main__":
    parser = get_main_parser()
    args = parser.parse_args()
    print("received --file='%s' --dir='%s'" % (args.file, args.dir))

    # completion magic
    shell = args.print_completion_shell
    if shell:
        print(shtab.complete(parser, shell=shell))

docopt

Simply use argopt to create a parser object from docopt syntax:

#!/usr/bin/env python
"""Greetings and partings.

Usage:
  greeter [options] [<you>] [<me>]

Options:
  -g, --goodbye  : Say "goodbye" (instead of "hello")
  -b, --print-bash-completion  : Output a bash tab-completion script
  -z, --print-zsh-completion  : Output a zsh tab-completion script

Arguments:
  <you>  : Your name [default: Anon]
  <me>  : My name [default: Casper]
"""
import sys, argopt, shtab  # NOQA

parser = argopt.argopt(__doc__)
if __name__ == "__main__":
    args = parser.parse_args()
    if args.print_bash_completion:
        print(shtab.complete(parser, shell="bash"))
        sys.exit(0)
    if args.print_zsh_completion:
        print(shtab.complete(parser, shell="zsh"))
        sys.exit(0)

    msg = "k thx bai!" if args.goodbye else "hai!"
    print("{} says '{}' to {}".format(args.me, msg, args.you))

Alternatives

  • argcomplete

    • executes the underlying script every time <TAB> is pressed (slow and has side-effects)

    • only provides bash completion

  • pyzshcomplete

    • executes the underlying script every time <TAB> is pressed (slow and has side-effects)

    • only provides zsh completion

  • click

    • different framework completely replacing argparse

    • solves multiple problems (rather than POSIX-style “do one thing well”)

Contributions

Please do open issues & pull requests! Some ideas:

  • support fish

  • support powershell

  • support tcsh

Download files

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

Source Distribution

shtab-1.0.1.tar.gz (15.4 kB view details)

Uploaded Source

Built Distribution

shtab-1.0.1-py2.py3-none-any.whl (10.6 kB view details)

Uploaded Python 2 Python 3

File details

Details for the file shtab-1.0.1.tar.gz.

File metadata

  • Download URL: shtab-1.0.1.tar.gz
  • Upload date:
  • Size: 15.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.1.1 pkginfo/1.5.0.1 requests/2.24.0 setuptools/47.1.1 requests-toolbelt/0.9.1 tqdm/4.46.1 CPython/3.7.7

File hashes

Hashes for shtab-1.0.1.tar.gz
Algorithm Hash digest
SHA256 d0714b3ca297cce12195a76a034201291e25248d9522ac62909c92933fccc5ca
MD5 f074354042abbc0eba43cfa05476780f
BLAKE2b-256 05caa06be05f429e3d57cfc0970ed6c6b0a7e4fa5c49e5e7aeac3cdc3e124a31

See more details on using hashes here.

Provenance

File details

Details for the file shtab-1.0.1-py2.py3-none-any.whl.

File metadata

  • Download URL: shtab-1.0.1-py2.py3-none-any.whl
  • Upload date:
  • Size: 10.6 kB
  • Tags: Python 2, Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.1.1 pkginfo/1.5.0.1 requests/2.24.0 setuptools/47.1.1 requests-toolbelt/0.9.1 tqdm/4.46.1 CPython/3.7.7

File hashes

Hashes for shtab-1.0.1-py2.py3-none-any.whl
Algorithm Hash digest
SHA256 cc7741385b5e75d916bdd2e335ce9199ee213e467694b42b8dd73ab171c68b75
MD5 e0d5980648688a22db3693bed8b49814
BLAKE2b-256 1f5938c4308a5888cf4612df75fe94f5b9ec7bb6a1df94f98ae14b8794ee0e62

See more details on using hashes here.

Provenance

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