Tab-Completion¶

Overview¶

As with many other shells, xonsh ships with the ability to complete partially-specified arguments upon hitting the “tab” key.

In Python-mode, pressing the “tab” key will complete based on the variable names in the current builtins, globals, and locals, as well as xonsh language keywords & operators, files & directories, and environment variable names. In subprocess-mode, xonsh additionally completes based on the names of any executable files on your $PATH, alias keys, and full Bash completion for the commands themselves.

xonsh also provides a mechanism by which the results of a tab completion can be customized (i.e., new completions can be generated, or a subset of the built-in completions can be ignored).

This page details the internal structure of xonsh’s completion system and includes instructions for implementing new tab completion functions.

Structure¶

xonsh’s built-in completers live in the xonsh.completers package, and they are managed through an instance of OrderedDict (__xonsh__.completers) that maps unique identifiers to completion functions.

The completers are divided to exclusive completers and non-exclusive completers. Non-exclusive completers are used for completions that are relevant but don’t cover the whole completions needed (e.g. completions for the built-in commands and/or).

When the “tab” key is pressed, xonsh loops over the completion functions in order, calling each one in turn and collecting its output until it reaches an exclusive one that returns a non-empty set of completions for the current line. The collected completions are then displayed to the user.

Listing Active Completers¶

A list of the active completers can be viewed by running the completer list command. This command will display names and descriptions of the currently-active completers, in the order in which they will be checked.

Writing a New Completer¶

Completers are implemented as Python functions that take a Completion Context object. Examples for the context object:

# ls /tmp/<TAB>
CompletionContext(
    command=CommandContext(
        args=(CommandArg(value='ls'),),
        arg_index=1, prefix='/tmp/',
        ),
    python=PythonContext(multiline_code="ls /tmp/", cursor_index=8, ctx={...})
)

# ls $(whic<TAB> "python") -l
CompletionContext(
    command=CommandContext(
        args=(CommandArg(value='python', opening_quote='"', closing_quote='"'),),
        arg_index=0, prefix='whic', subcmd_opening='$(',
    ),
    python=None
)

# echo @(sys.exe<TAB>)
CompletionContext(
    command=None,
    python=PythonContext(
        multiline_code="sys.exe", cursor_index=7,
        is_sub_expression=True, ctx={...},
    )
)

Note

Xonsh still supports legacy completers - see Legacy Completers Support. For backwards-compatibility, contextual completers need to be marked (as seen in the examples).

This function should return a python set of possible completions for command.prefix in the current context. If the completer should not be used in this case, it should return None or an empty set, which will cause xonsh to move on and try to use the next completer.

Occasionally, completers will need to return a match that does not actually start with prefix. In this case, a completer should instead return a tuple (completions, prefixlength), where completions is the set of appropriate completions, and prefixlength is the number of characters in line that should be treated as part of the completion.

Note

Further completion customizations can be made using the RichCompletion object - see Advanced Completions.

The docstring of a completer should contain a brief description of its functionality, which will be displayed by completer list.

Some simple examples follow. For real-world examples, see xompletions.git (subprocess-based, uses --git-completion-helper) and xonsh.completers.man (man page parsing with disk cache).

@ from xonsh.completers.tools import *

@ @contextual_completer
  def dummy_completer(context):
      '''
      Completes everything with options "lou" and "carcolh",
      regardless of the value of prefix.
      '''
      return {"lou", "carcolh"}

@ completer add dummy dummy_completer
@ xzecze<TAB>   # → lou, carcolh

@ @non_exclusive_completer
  @contextual_completer
  def nx_dummy_completer(context):
      '''
      Like dummy_completer but its results are ADDED to the other completions.
      '''
      return {"lou", "carcolh"}

@ completer add nx_dummy nx_dummy_completer

@contextual_completer
def python_context_completer(context):
    '''
    Completes based on the names in the current Python environment
    '''
    if context.python:
        last_name = context.python.prefix.split()[-1]
        return {i for i in context.python.ctx if i.startswith(last_name)}

@contextual_completer
def unbeliever_completer(context):
    '''
    Replaces "lou carcolh" with "snail" if tab is pressed after at least
    typing the "lou " part.
    '''
    if (
        # We're completing a command
        context.command and
        # We're completing the second argument
        context.command.arg_index == 1 and
        # The first argument is 'lou'
        context.command.args[0].value == 'lou' and
        # The prefix startswith 'carcolh' (may be empty)
        'carcolh'.startswith(context.command.prefix)
    ):
        return {'snail'}, len('lou ') + len(context.command.prefix)

# Save boilerplate with this helper decorator:

@contextual_command_completer_for("lou")
def better_unbeliever_completer(command):
    """Like unbeliever_completer but with less boilerplate"""
    if command.arg_index == 1 and 'carcolh'.startswith(command.prefix):
        return {'snail'}, len('lou ') + len(command.prefix)

To understand how xonsh uses completers and their return values try to set $XONSH_COMPLETER_TRACE to True:

@ $XONSH_COMPLETER_TRACE = True
@ pip c<TAB>
TRACE COMPLETIONS: Getting completions with context:
CompletionContext(command=CommandContext(args=(CommandArg(value='pip', opening_quote='', closing_quote=''),), arg_index=1, prefix='c', suffix='', opening_quote='', closing_quote='', is_after_closing_quote=False, subcmd_opening=''), python=PythonContext('pip c', 5, is_sub_expression=False))
TRACE COMPLETIONS: Got 3 from exclusive 'pip' for 'c':
'cache': src=pip, type=exclusive, prefix_len=1, append_space=True
'check': src=pip, type=exclusive, prefix_len=1, append_space=True
'config': src=pip, type=exclusive, prefix_len=1, append_space=True

The header reads as Got <N> from <type> '<name>' for '<prefix>' — showing the number of results, the completer’s exclusivity, its name, and the prefix the completer tried to match. Each body line shows the completion value, the src (completer that produced it), the completer type (exclusive or non-exclusive), and any non-default RichCompletion attributes (prefix_len, display, description, append_space, close_quote — short for append_closing_quote — style, pvd — short for provider). The format is grep-friendly, so you can filter by source or type, e.g. ... | grep 'src=pip' or ... | grep 'type=non-exclusive'.

Fine-grained origin via `provider`¶

A completer may set a provider tag on each RichCompletion to identify the sub-source inside the completer. This is trace-only metadata and does not affect the completion UX.

For example, the base completer runs a union of Python names, $PATH executables, aliases, and file paths. When you type the first argument and hit <TAB>, the trace distinguishes them:

@ aliases['qwe-xonsh'] = 'echo'
@ xonsh<TAB>
TRACE COMPLETIONS: Got 2 from exclusive 'base' for 'xonsh':
'qwe-xonsh ': src=base, pvd='alias', type=exclusive, prefix_len=5, append_space=True
'xonsh ': src=base, pvd='command', type=exclusive, prefix_len=5, append_space=True

Completers that are invoked but return no usable matches are also reported, so you can see the full decision path:

TRACE COMPLETIONS: Got 0 from non-exclusive 'environment_vars' for 'xonsh'.
TRACE COMPLETIONS: Got 0 from exclusive 'bash' for 'xonsh'.
TRACE COMPLETIONS: Got 3 from exclusive 'path' for './do':
...

This way you can see immediately that qwe-xonsh comes from an alias while xonsh is a real executable on $PATH. Built-in providers: alias, command, python, path, plus the xompletion module name (e.g. pip, gh) for completions produced by the xompleter. Custom completers may set any string they like.

Setting `provider` in your own completer¶

There are two ways to tag your completions, depending on whether you want per-item control or want to label the whole result at once.

Per-item: pass provider=... directly to each RichCompletion. This is most useful when a single completer has internal branches and you want to distinguish them in trace:

from xonsh.completers.tools import (
    RichCompletion,
    contextual_command_completer_for,
)

@contextual_command_completer_for("deploy")
def complete_deploy(command):
    """Complete ``deploy <env>`` with fast- and slow-paths tagged."""
    cached = {"prod", "staging"}       # e.g. loaded from a local cache
    remote = {"dev-42", "dev-43"}      # e.g. hit an HTTP endpoint

    for env in cached:
        if env.startswith(command.prefix):
            yield RichCompletion(env, provider="deploy:cache")
    for env in remote:
        if env.startswith(command.prefix):
            yield RichCompletion(env, provider="deploy:remote")

After completer add deploy complete_deploy the trace will show:

TRACE COMPLETIONS: Got 3 from exclusive 'deploy' for 'd':
'dev-42': src=deploy, pvd='deploy:remote', type=exclusive, prefix_len=1
'dev-43': src=deploy, pvd='deploy:remote', type=exclusive, prefix_len=1
'staging': src=deploy, pvd='deploy:cache', type=exclusive, prefix_len=1

Whole-result: wrap the completer’s return value with tag_provider. This preserves the return shape (None, iterable, or (comps, lprefix) tuple) and only tags completions that do not already carry their own provider:

from xonsh.completers.tools import (
    contextual_command_completer_for,
    tag_provider,
)

@contextual_command_completer_for("mycmd")
def complete_mycmd(command):
    """Completer whose entire output is tagged ``provider='mycmd'``."""
    results = {"start", "stop", "status"}
    return tag_provider(results, "mycmd")

tag_provider is also what the built-in xompleter uses under the hood to label each xompletion module’s output — so any RichCompletion that already specifies its own provider is kept intact and won’t be overwritten by the outer tag.

Registering a Completer¶

Once you have created a completion function, you can add it to the list of active completers via the completer add command or xonsh.completers.completer.add_one_completer function:

Usage:
    completer add NAME FUNC [POS]

NAME is a unique name to use in the listing

FUNC is the name of a completer function to use.

POS (optional) is a position into the list of completers at which the new completer should be added. It can be one of the following values:

"start" indicates that the completer should be added to the start of the list of completers (
it should be run before all other exclusive completers)
"end" indicates that the completer should be added to the end of the list of completers (it should be run after all others)
">KEY", where KEY is a pre-existing name, indicates that this should be added after the completer named KEY
"<KEY", where KEY is a pre-existing name, indicates that this should be added before the completer named KEY

If POS is not provided, it defaults to "start".

Note

It is also possible to manipulate __xonsh__.completers directly, but this is the preferred method.

Removing a Completer¶

To remove a completer from the list of active completers, run completer remove NAME, where NAME is the unique identifier associated with the completer you wish to remove.

Advanced Completions¶

To provide further control over the completion, a completer can return a RichCompletion object. Using this class, you can:

Provide a specific prefix length per completion (via prefix_len)
Control how the completion looks in prompt-toolkit (via display, description and style) -
use the jedi xontrib to see it in action.
Append a space after the completion (append_space=True)

Completing Closed String Literals¶

When the cursor is appending to a closed string literal (i.e. cursor at the end of ls "/usr/"), the following happens:

The closing quote will be appended to all completions.
I.e the completion /usr/bin will turn into /usr/bin". To prevent this behavior, a completer can return a RichCompletion with append_closing_quote=False.
If not specified, lprefix will cover the closing prefix.
I.e for ls "/usr/", the default lprefix will be 6 to include the closing quote. To prevent this behavior, a completer can return a different lprefix or specify it inside RichCompletion.

So if you want to change/remove the quotes from a string, the following completer can be written:

@contextual_command_completer
def remove_quotes(command):
    """
    Return a completer that will remove the quotes, i.e:
    which "python"<TAB> -> which python
    echo "hi<TAB> -> echo hi
    ls "file with spaces"<TAB> -> ls file with spaces
    """
    raw_prefix_len = len(command.raw_prefix)  # this includes the closing quote if it exists
    return {RichCompletion(command.prefix, prefix_len=raw_prefix_len, append_closing_quote=False)}

Completing Aliases¶

You can attach a custom completer to a function alias using the @aliases.completer decorator:

def _complete_hello(command, alias):
    return {'world', 'there', 'xonsh'}

@aliases.register
@aliases.completer(_complete_hello)
def _hello(args):
    echo @(args)

Now hello <TAB> will suggest world, there, and xonsh.

You can also set the xonsh_complete attribute manually:

def _hello(args):
    echo @(args)

_hello.xonsh_complete = lambda *a, **kw: {'world', 'there', 'xonsh'}
aliases['hello'] = _hello

The completer function receives two keyword arguments:

command: the CommandContext for the current completion
alias: the resolved alias object

Command Completers (xompletions)¶

xonsh includes a package called xompletions that provides tab-completions for specific commands like pip, gh, cd, etc. Each command gets its own Python module inside the xompletions/ directory.

How it works:

When the user presses TAB, the xompleter completer (registered as complete_xompletions) extracts the command name from args[0].
It looks for a matching module in xompletions/ — first by exact name, then by regex patterns.
If found, it calls the module’s xonsh_complete(ctx) function.
The function returns completions or None (to let the next completer handle it).

Creating a command completer¶

To create a completer for a command, place a Python file named after the command in any directory listed in $XONSH_COMPLETER_DIRS. The file must contain a xonsh_complete function:

# ~/.config/xonsh/completers/mycmd.py
from xonsh.parsers.completion_context import CommandContext

def xonsh_complete(ctx: CommandContext):
    """Completes mycmd subcommands."""
    if ctx.arg_index == 1:
        return {'start', 'stop', 'status'}

$XONSH_COMPLETER_DIRS = ["~/.config/xonsh/completers"]

Now mycmd <TAB> will suggest start, stop, and status.

xonsh also ships built-in completers in the xompletions/ package (for pip, gh, cd, etc.).

Handling command name variants with `wrap`¶

The file name must match the command name exactly (gh.py for gh). On Windows, extensions like .exe are stripped automatically via $PATHEXT, so gh.exe will find gh.py.

However, if a command has other name variants (e.g. pip3.11, python3.12), the exact file name won’t match. For these cases, you can register regex patterns from your xonshrc or xontrib:

from xonsh.completers.commands import complete_xompletions as xmp
xmp.wrap(r"\bmycmd(?:\d)*$", "mycmd")

This maps mycmd, mycmd2, mycmd3 etc. to the mycmd completer module.

xonsh ships with built-in patterns for pip (covers xpip, pip3.11, pip.exe) and python (covers python3, python3.12, python.exe).

Completing `python -m <module>`¶

When an alias resolves to python -m <module> (e.g. xpip → python -m pip), xonsh uses the xompletions/python.py completer to delegate to the module’s completer.

The mapping is stored in PYTHON_MODULE_COMPLETERS and can be extended from xonshrc:

from xompletions.python import PYTHON_MODULE_COMPLETERS

# Simple completer with static options
def _complete_mytool(ctx, module_arg_index):
    return {'start', 'stop', 'status'}

PYTHON_MODULE_COMPLETERS['mytool'] = _complete_mytool

Now python -m mytool <TAB> will suggest start, stop, and status. This also works through aliases:

aliases['mt'] = ['python', '-m', 'mytool']
mt <TAB>  # completes with start, stop, status

For modules that use the argcomplete protocol, a ready-made helper is available:

from xompletions.python import PYTHON_MODULE_COMPLETERS, _complete_argcomplete

PYTHON_MODULE_COMPLETERS['my_argcomplete_tool'] = _complete_argcomplete

Emoji & Symbols¶

Need a 🐈 in your commit message? xonsh has a built-in emoji completer. It is disabled by default. To enable, set the trigger prefixes:

@ $XONSH_COMPLETER_EMOJI_PREFIX = '::'
@ $XONSH_COMPLETER_SYMBOLS_PREFIX = ':::'

Then type :: followed by a keyword and press TAB to search for colorful emoji:

echo "great job ::fire<TAB>"   →  echo "great job 🔥"
echo "::cat<TAB>"              →  echo "🐈"

For classic unicode symbols (arrows, math, stars), use ::::

echo ":::arrow<TAB>"  →  echo "→"
echo ":::star<TAB>"   →  echo "★"

Set $XONSH_COMPLETER_EMOJI_PREFIX or $XONSH_COMPLETER_SYMBOLS_PREFIX to None to disable the corresponding completer.

Selecting a tab completion result¶

In the prompt_toolkit shell, you can cycle through possible tab-completion results using the TAB key and use ENTER to select the completion you want. By default, ENTER will also execute the current line. If you would prefer to not automatically execute the line (say, if you’re constructing a long pathname), you can set

$COMPLETIONS_CONFIRM = True

in your xonsh RC.

By default, TABs cycle through the full list. Set $COMPLETION_MODE = "menu-complete" to instead insert the first whole completion on the first TAB and cycle through the rest on further TABs. This is only one of many knobs; see Tab Completion Behavior for the full list (display style, menu rows, threading, trace output, and more).

Man Page Completer¶

When no dedicated completer exists for a command, xonsh falls back to parsing the command’s man page to extract option names (-v, --verbose, etc.). This works automatically for most CLI tools.

For commands that use per-subcommand man pages (docker-run, cargo-build, systemctl-start, etc.), xonsh tries the hyphenated form man <cmd>-<subcmd> first and falls back to man <cmd>.

Parsed options are cached on disk under $XONSH_DATA_DIR/generated_completions/man/. The cache is invalidated automatically when the man page file is updated (e.g. after a package upgrade). To force a refresh — for example, after upgrading xonsh itself with an improved parser — clear the cache:

@ rm $XONSH_DATA_DIR/generated_completions/man/*

Installing man pages for tools¶

Some tools (Docker, Podman, etc.) ship man pages in a separate package. If docker run -<TAB> returns no completions, the man page may not be installed. On macOS with Homebrew:

$ brew install docker

On Linux, man pages are usually part of the main package, but some distributions split them out (e.g. docker-doc on Debian/Ubuntu).

You can verify whether a man page is available:

$ man -w docker-run

If this prints a path, completions will work. If it prints an error, install the corresponding package.

Legacy Completers Support¶

Before completion context was introduced, xonsh had a different readline-like completion API. While this legacy API is not recommended, xonsh still supports it.

Warning

The legacy completers are less robust than the contextual system in many situations, for example:

ls $(which<TAB> completes with the prefix $(which
ls 'a file<TAB> completes with the prefix file (instead of a file)

See Completion Context PR for more information.

Legacy completers are python functions that aren’t marked by @contextual_completer and receive the following arguments:

prefix: the string to be matched (the last whitespace-separated token in the current line)
line: a string representing the entire current line
begidx: the index at which prefix starts in line
endidx: the length of the prefix in line
ctx: the current Python environment, as a dictionary mapping names to values

Their return value can be any of the variations of the contextual completers’.

Tab-Completion¶

Overview¶

Structure¶

Listing Active Completers¶

Writing a New Completer¶

Fine-grained origin via provider¶

Setting provider in your own completer¶

Registering a Completer¶

Removing a Completer¶

Advanced Completions¶

Completing Closed String Literals¶

Completing Aliases¶

Command Completers (xompletions)¶

Creating a command completer¶

Handling command name variants with wrap¶

Completing python -m <module>¶

Emoji & Symbols¶

Selecting a tab completion result¶

Man Page Completer¶

Installing man pages for tools¶

Legacy Completers Support¶

Fine-grained origin via `provider`¶

Setting `provider` in your own completer¶

Handling command name variants with `wrap`¶

Completing `python -m <module>`¶