A new type of shell
Go to file
Jonathan Turner 2d8b558ac0
Merge pull request #559 from jonathandturner/zombie_killing
Another attempt to fix the zombie processes
2019-09-02 05:22:01 +12:00
.azure Attempt to speed up CI 2019-09-01 19:33:13 +12:00
.cargo A real parser (lalrpop) 2019-05-26 00:17:35 -07:00
.circleci adding missing quay.io registry references 2019-08-29 21:48:30 -04:00
assets Add a more complete syntax file (from bat) 2019-07-22 04:03:54 +12:00
docker so close, needed to have quay.io as from image base (not docker hub) 2019-08-29 20:56:48 -04:00
docs Added philosophy.md 2019-06-06 10:58:39 -07:00
images New animated gif 2019-08-23 17:19:02 +12:00
src Another attempt to fix the zombie processes 2019-09-02 04:45:30 +12:00
tests This commit introduces a basic help feature. We can go to it 2019-08-31 19:06:11 -05:00
.editorconfig Add editorconfig 2019-05-11 00:03:11 -07:00
.gitignore Tests pass! 2019-07-23 15:22:11 -07:00
Cargo.lock Random fixes 2019-09-01 09:19:59 +12:00
Cargo.toml Random fixes 2019-09-01 09:19:59 +12:00
CODE_OF_CONDUCT.md Create CODE_OF_CONDUCT.md 2019-08-25 19:19:47 -07:00
LICENSE Create LICENSE 2019-06-03 04:47:52 +12:00
Makefile.toml Add --loglevel and --develop 2019-06-01 10:00:42 -07:00
README.md This commit introduces a basic help feature. We can go to it 2019-08-31 19:06:11 -05:00
rust-toolchain Basic support for decimal numbers 2019-08-30 21:05:32 -07:00
rustfmt.toml Data flows across commands via streams now 2019-05-23 00:23:06 -07:00

Build Status Discord

Nu Shell

A modern shell for the GitHub era

Example of nushell

Status

This project has reached a minimum-viable product level of quality. While contributors dogfood it as their daily driver, it may be instable for some commands. Future releases will work fill out missing features and improve stability. Its design is also subject to change as it matures.

Nu comes with a set of built-in commands (listed below). If a command is unknown, the command will shell-out and execute it (using cmd on Windows or bash on Linux and MacOS), correctly passing through stdin, stdout and stderr, so things like your daily git workflows and even vim will work just fine.

There is also a book about Nu, currently in progress.

Installation

Local

Up-to-date installation instructions can be found in the installation chapter of the book.

To build Nu, you will need to use the nightly version of the compiler.

Required dependencies:

  • libssl (only needed on Linux)
    • on Debian/Ubuntu: apt install libssl-dev

Optional dependencies:

  • To use Nu with all possible optional features enabled, you'll also need the following:
    • on Linux (on Debian/Ubuntu): apt install libxcb-composite0-dev libx11-dev

To install Nu via cargo:

cargo +nightly install nu

You can also install Nu with all the bells and whistles:

cargo +nightly install nu --features raw-key,clipboard

The following optional features are currently supported:

  • raw-key - direct keyboard input, which creates a smoother experience in viewing text and binaries
  • clipboard - integration with the native clipboard via the clip command

Docker

If you want to pull a pre-built container, you can browse tags for the nushell organization on Quay.io. Pulling a container would come down to:

$ docker pull quay.io/nushell/nu
$ docker pull quay.io/nushell/nu-base

Both "nu-base" and "nu" provide the nu binary, however nu-base also includes the source code at /code in the container and all dependencies.

Optionally, you can also build the containers locally using the dockerfiles provided: To build the base image:

$ docker build -f docker/Dockerfile.nu-base -t nushell/nu-base .

And then to build the smaller container (using a Multistage build):

$ docker build -f docker/Dockerfile -t nushell/nu .

Either way, you can run either container as follows:

$ docker run -it nushell/nu-base
$ docker run -it nushell/nu
/> exit

The second container is a bit smaller, if size is important to you.

Philosophy

Nu draws inspiration from projects like PowerShell, functional programming languages, and modern cli tools. Rather than thinking of files and services as raw streams of text, Nu looks at each input as something with structure. For example, when you list the contents of a directory, what you get back is a list of objects, where each object represents an item in that directory. These values can be piped through a series of steps, in a series of commands called a 'pipeline'.

Pipelines

In Unix, it's common to pipe between commands to split up a sophisticated command over multiple steps. Nu takes this a step further and builds heavily on the idea of pipelines. Just as the Unix philosophy, Nu allows commands to output from stdout and read from stdin. Additionally, commands can output structured data (you can think of this as a third kind of stream). Commands that work in the pipeline fit into one of three categories

  • Commands that produce a stream (eg, ls)
  • Commands that filter a stream (eg, where type == "Directory")
  • Commands that consumes the output of the pipeline (eg, autoview)

Commands are separated by the pipe symbol (|) to denote a pipeline flowing left to right.

/home/jonathan/Source/nushell(master)> ls | where type == "Directory" | autoview
--------+-----------+----------+--------+--------------+----------------
 name   | type      | readonly | size   | accessed     | modified
--------+-----------+----------+--------+--------------+----------------
 target | Directory |          | 4.1 KB | 19 hours ago | 19 hours ago
 images | Directory |          | 4.1 KB | 2 weeks ago  | a week ago
 tests  | Directory |          | 4.1 KB | 2 weeks ago  | 18 minutes ago
 docs   | Directory |          | 4.1 KB | a week ago   | a week ago
 .git   | Directory |          | 4.1 KB | 2 weeks ago  | 25 minutes ago
 src    | Directory |          | 4.1 KB | 2 weeks ago  | 25 minutes ago
 .cargo | Directory |          | 4.1 KB | 2 weeks ago  | 2 weeks ago
--------+-----------+----------+--------+--------------+----------------

Because most of the time you'll want to see the output of a pipeline, autoview is assumed. We could have also written the above:

/home/jonathan/Source/nushell(master)> ls | where type == Directory

Being able to use the same commands and compose them differently is an important philosophy in Nu. For example, we could use the built-in ps command as well to get a list of the running processes, using the same where as above.

C:\Code\nushell(master)> ps | where cpu > 0
------------------ +-----+-------+-------+----------
 name              | cmd | cpu   | pid   | status
------------------ +-----+-------+-------+----------
 msedge.exe        |  -  | 0.77  | 26472 | Runnable
 nu.exe            |  -  | 7.83  | 15473 | Runnable
 SearchIndexer.exe |  -  | 82.17 | 23476 | Runnable
 BlueJeans.exe     |  -  | 4.54  | 10000 | Runnable
-------------------+-----+-------+-------+----------

Opening files

Nu can load file and URL contents as raw text or as structured data (if it recognizes the format). For example, you can load a .toml file as structured data and explore it:

/home/jonathan/Source/nushell(master)> open Cargo.toml
-----------------+------------------+-----------------
 dependencies    | dev-dependencies | package
-----------------+------------------+-----------------
 [object Object] | [object Object]  | [object Object]
-----------------+------------------+-----------------

We can pipeline this into a command that gets the contents of one of the columns:

/home/jonathan/Source/nushell(master)> open Cargo.toml | get package
-------------+----------------------------+---------+---------+------+---------
 authors     | description                | edition | license | name | version
-------------+----------------------------+---------+---------+------+---------
 [list List] | A shell for the GitHub era | 2018    | MIT     | nu   | 0.2.0
-------------+----------------------------+---------+---------+------+---------

Finally, we can use commands outside of Nu once we have the data we want:

/home/jonathan/Source/nushell(master)> open Cargo.toml | get package.version | echo $it
0.2.0

Here we use the variable $it to refer to the value being piped to the external command.

Shells

By default, Nu will work inside of a single directory and allow you to navigate around your filesystem. Sometimes, you'll want to work in multiple directories at the same time. For this, Nu offers a way of adding additional working directories that you can jump between.

To do so, use the enter command, which will allow you create a new "shell" and enter it at the specified path. You can toggle between this new shell and the original shell with the p (for previous) and n (for next), allowing you to navigate around a ring buffer of shells. Once you're done with a shell, you can exit it and remove it from the ring buffer.

Finally, to get a list of all the current shells, you can use the shells command.

Plugins

Nu supports plugins that offer additional functionality to the shell and follow the same object model that built-in commands use. This allows you to extend nu for your needs.

There are a few examples in the plugins directory.

Plugins are binaries that are available in your path and follow a "nu_plugin_*" naming convention. These binaries interact with nu via a simple JSON-RPC protocol where the command identifies itself and passes along its configuration, which then makes it available for use. If the plugin is a filter, data streams to it one element at a time, and it can stream data back in return via stdin/stdout. If the plugin is a sink, it is given the full vector of final data and is given free reign over stdin/stdout to use as it pleases.

Goals

Nu adheres closely to a set of goals that make up its design philosophy. As features are added, they are checked against these goals.

  • First and foremost, Nu is cross-platform. Commands and techniques should carry between platforms and offer first-class consistent support for Windows, macOS, and Linux.

  • Nu ensures direct compatibility with existing platform-specific executables that make up people's workflows.

  • Nu's workflow and tools should have the usability in day-to-day experience of using a shell in 2019 (and beyond).

  • Nu views data as both structured and unstructured. It is an object shell like PowerShell.

  • Finally, Nu views data functionally. Rather than using mutation, pipelines act as a means to load, change, and save data without mutable state.

Commands

Initial commands

command description
cd path Change to a new path
cp source path Copy files
ls (path) View the contents of the current or given path
mkdir path Make directories, creates intermediary directories as required.
mv source target Move files or directories.
date (--utc) Get the current datetime
ps View current processes
sys View information about the current system
which filename Finds a program file.
open {filename or url} Load a file into a cell, convert to table if possible (avoid by appending '--raw')
post url body (--user ) (--password ) Post content to a url and retrieve data as a table if possible
rm {file or directory} Remove a file, (for removing directory append '--recursive')
exit (--now) Exit the current shell (or all shells)
enter (path) Create a new shell and begin at this path
p Go to previous shell
n Go to next shell
shells Display the list of current shells
help Display help information about commands
version Display Nu version

Filters on tables (structured data)

command description
pick ...columns Down-select table to only these columns
reject ...columns Remove the given columns from the table
get column-or-column-path Open column and get data from the corresponding cells
sort-by ...columns Sort by the given columns
where condition Filter table to match the condition
inc (field) Increment a value or version. Optional use the field of a table
add field value Add a new field to the table
sum Sum a column of values
edit field value Edit an existing field to have a new value
reverse Reverses the table.
skip amount Skip a number of rows
skip-while condition Skips rows while the condition matches.
first amount Show only the first number of rows
last amount Show only the last number of rows
nth row-number Return only the selected row
str (field) Apply string function. Optional use the field of a table
tags Read the tags (metadata) for values
from-array Expand an array/list into rows
to-array Collapse rows into a single list
to-json Convert table into .json text
to-toml Convert table into .toml text
to-yaml Convert table into .yaml text
to-bson Convert table into .bson text
to-csv Convert table into .csv text
to-bson Convert table into .bson binary data
to-tsv Convert table into .tsv text
to-sqlite Convert table to sqlite .db binary data
reverse Reverse the rows of a table

Filters on text (unstructured data)

command description
from-bson Parse binary data as .bson and create table
from-csv Parse text as .csv and create table
from-ini Parse text as .ini and create table
from-json Parse text as .json and create table
from-sqlite Parse binary data as sqlite .db and create table
from-toml Parse text as .toml and create table
from-tsv Parse text as .tsv and create table
from-xml Parse text as .xml and create a table
from-yaml Parse text as a .yaml/.yml and create a table
lines Split single string into rows, one per line
size Gather word count statistics on the text
split-column sep ...fields Split row contents across multiple columns via the separator
split-row sep Split row contents over multiple rows via the separator
trim Trim leading and following whitespace from text data
{external-command} $it Run external command with given arguments, replacing $it with each row text

Consuming commands

command description
autoview View the contents of the pipeline as a table or list
binaryview Autoview of binary data
clip Copy the contents of the pipeline to the copy/paste buffer
save filename Save the contents of the pipeline to a file
table View the contents of the pipeline as a table
textview Autoview of text data
tree View the contents of the pipeline as a tree
vtable View the contents of the pipeline as a vertical (rotated) table

License

The project is made available under the MIT license. See "LICENSE" for more information.