REFACTOR: clean the root of the repo (#9231)

# Description
i've almost always wanted to clean up the root of the repo, so here is
my take at it, with some important advice given by @fdncred 😌

- `README.release.txt` is now gone and directly inline in the
`release-pkg` script used in the `release` *workflow*
- `build.rs` has been moved to `scripts/` and its path has been changed
in
[`Cargo.toml`](https://github.com/amtoine/nushell/blob/refactor/clean-root/Cargo.toml#L3)
according to the [*Build Scripts*
section](https://doc.rust-lang.org/cargo/reference/build-scripts.html#build-scripts)
of *The Cargo Book*
- i've merged `images/` into `assets/` and fix the only mention to the
GIF in the README
- i've moved the `docs/README.md` inside the main `README.md` as a new
[*Configuration*
section](https://github.com/amtoine/nushell/tree/refactor/clean-root#configuration)
- the very deprecated `pkg_mgrs/` has been removed
- all the `.nu`, `.sh`, `.ps1` and `.cmd` scripts have been moved to
`scripts/`

### things i've left as-is
- all the other `.md` documents
- the configuration files
- all the Rust and core stuff
- `docker/`
- `toolkit.nu`
- the `wix/` diretory which appears to be important for `winget`

# User-Facing Changes
scripts that used to rely on the paths to some of the scripts should now
call the scripts inside `scripts/` => i think this for the greater good,
it was not pretty nor scalable to have a bunch of scripts in the root of
our main `nushell` 😱

*i even think we might want to move these scripts outside the main
`nushell` repo*
maybe to `nu_scripts` or some other tool 👍 

# Tests + Formatting
- 🟢 `toolkit fmt`
- 🟢 `toolkit clippy`
-  `toolkit test`
-  `toolkit test stdlib`

# After Submitting
```
$nothing
```
This commit is contained in:
Antoine Stevan 2023-05-20 14:57:51 +02:00 committed by GitHub
parent 4b9f479e5c
commit 8eece32a8d
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
20 changed files with 180 additions and 131 deletions

View file

@ -122,7 +122,9 @@ print $'(char nl)All executable files:'; hr-line
print (ls -f $executable); sleep 1sec
print $'(char nl)Copying release files...'; hr-line
cp -v README.release.txt $'($dist)/README.txt'
"To use Nu plugins, use the register command to tell Nu where to find the plugin. For example:
> register ./nu_plugin_query" | save $'($dist)/README.txt'
[LICENSE $executable] | each {|it| cp -rv $it $dist } | flatten
# Sleep a few seconds to make sure the cp process finished successfully
sleep 3sec

View file

@ -1,5 +1,6 @@
[package]
authors = ["The Nushell Project Developers"]
build = "scripts/build.rs"
default-run = "nu"
description = "A new type of shell"
documentation = "https://www.nushell.sh/book/"

View file

@ -10,13 +10,14 @@
A new type of shell.
![Example of nushell](images/nushell-autocomplete6.gif "Example of nushell")
![Example of nushell](assets/nushell-autocomplete6.gif "Example of nushell")
## Table of Contents <!-- omit in toc -->
- [Status](#status)
- [Learning About Nu](#learning-about-nu)
- [Installation](#installation)
- [Configuration](#configuration)
- [Philosophy](#philosophy)
- [Pipelines](#pipelines)
- [Opening files](#opening-files)
@ -55,6 +56,22 @@ Detailed installation instructions can be found in the [installation chapter of
[![Packaging status](https://repology.org/badge/vertical-allrepos/nushell.svg)](https://repology.org/project/nushell/versions)
## Configuration
The default configurations can be found at [sample_config](crates/nu-utils/src/sample_config)
which are the configuration files one gets when they startup Nushell for the first time.
It sets all of the default configuration to run Nushell. From here one can
then customize this file for their specific needs.
To see where *config.nu* is located on your system simply type this command.
```rust
$nu.config-path
```
Please see our [book](https://www.nushell.sh) for all of the Nushell documentation.
## Philosophy

View file

@ -1,3 +0,0 @@
To use Nu plugins, use the register command to tell Nu where to find the plugin. For example:
> register ./nu_plugin_query

View file

Before

Width:  |  Height:  |  Size: 1.2 MiB

After

Width:  |  Height:  |  Size: 1.2 MiB

View file

@ -1,25 +0,0 @@
echo '-------------------------------------------------------------------'
echo 'Building nushell (nu) with dataframes and all the plugins'
echo '-------------------------------------------------------------------'
echo $'(char nl)Building nushell'
echo '----------------------------'
cargo build --features=dataframe
let plugins = [
nu_plugin_inc,
nu_plugin_gstat,
nu_plugin_query,
nu_plugin_example,
nu_plugin_custom_values,
nu_plugin_formats,
]
for plugin in $plugins {
$'(char nl)Building ($plugin)'
'----------------------------'
cd $'crates/($plugin)'
cargo build
cd ../../
ignore
}

View file

@ -1,54 +0,0 @@
#!/usr/bin/env nu
let start = (date now)
# Script to generate coverage locally
#
# Output: `lcov.info` file
#
# Relies on `cargo-llvm-cov`. Install via `cargo install cargo-llvm-cov`
# https://github.com/taiki-e/cargo-llvm-cov
# You probably have to run `cargo llvm-cov clean` once manually,
# as you have to confirm to install additional tooling for your rustup toolchain.
# Else the script might stall waiting for your `y<ENTER>`
# Some of the internal tests rely on the exact cargo profile
# (This is somewhat criminal itself)
# but we have to signal to the tests that we use the `ci` `--profile`
let-env NUSHELL_CARGO_TARGET = "ci"
# Manual gathering of coverage to catch invocation of the `nu` binary.
# This is relevant for tests using the `nu!` macro from `nu-test-support`
# see: https://github.com/taiki-e/cargo-llvm-cov#get-coverage-of-external-tests
print "Setting up environment variables for coverage"
# Enable LLVM coverage tracking through environment variables
# show env outputs .ini/.toml style description of the variables
# In order to use from toml, we need to make sure our string literals are single quoted
# This is especially important when running on Windows since "C:\blah" is treated as an escape
cargo llvm-cov show-env | str replace (char dq) (char sq) -a | from toml | load-env
print "Cleaning up coverage data"
cargo llvm-cov clean --workspace
print "Building with workspace and profile=ci"
# Apparently we need to explicitly build the necessary parts
# using the `--profile=ci` is basically `debug` build with unnecessary symbols stripped
# leads to smaller binaries and potential savings when compiling and running
cargo build --workspace --profile=ci
print "Running tests with --workspace and profile=ci"
cargo test --workspace --profile=ci
# You need to provide the used profile to find the raw data
print "Generating coverage report as lcov.info"
cargo llvm-cov report --lcov --output-path lcov.info --profile=ci
let end = (date now)
$"Coverage generation took ($end - $start)."
# To display the coverage in your editor see:
#
# - https://marketplace.visualstudio.com/items?itemName=ryanluker.vscode-coverage-gutters
# - https://github.com/umaumax/vim-lcov
# - https://github.com/andythigpen/nvim-coverage (probably needs some additional config)

View file

@ -1,15 +0,0 @@
# Documentation
The default configurations can be found at [sample_config](../crates/nu-utils/src/sample_config)
which are the configuration files one gets when they startup Nushell for the first time.
It sets all of the default configuration to run Nushell. From here one can
then customize this file for their specific needs.
To see where *config.nu* is located on your system simply type this command.
```rust
$nu.config-path
```
Please see our [book](https://www.nushell.sh) for all of the Nushell documentation.

View file

@ -1,16 +0,0 @@
Id: Nushell.Nushell
Publisher: Nushell
Name: Nushell
Version: 0.17.0
License: MIT
LicenseUrl: https://github.com/nushell/nushell/blob/main/LICENSE
AppMoniker: nushell
Tags: shell, nu, nushell, functional, data, analysis
Description: Nushell. A new type of shell.
Homepage: https://www.nushell.sh/
InstallerType: msi
Installers:
- Arch: x64
Url: https://github.com/nushell/nushell/releases/download/0.17.0/nu_0_17_0_windows.msi
Sha256: 3EF3FCE4069510AD78577DC61E29F3DC02F4A384DE5E2CCD9D9862FC7E4D849E
ManifestVersion: 0.1.0

27
scripts/README.md Normal file
View file

@ -0,0 +1,27 @@
## run the scripts
> **Note**
> the following table must be read as follows:
> - an `x` means *it works*
> - a `?` means *no data available*
>
> `.nu` scripts must be run as `nu .../foo.nu`
> `.sh` scripts must be run as `./.../foo.sh`
> `.ps1` scripts must be run as `powershell .../foo.ps1`
>
> let's say a script is called `foo`
> - an `x` in the *`./scripts`* column means *`foo` can be run from `./scripts`*
> - an `x` in the *root* column means *`foo` can be run from the root of `nushell`*
> - an `x` in the *anywhere* column means *`foo` can be run from anywhere!*
| script | `./scripts/` | root | anywhere |
| ----------------------- | ------------ | ---- | -------- |
| `build-all-maclin.sh` | x | x | x |
| `build-all-windows.cmd` | ? | x | ? |
| `build-all.nu` | x | x | x |
| `coverage-local.nu` | x | x | x |
| `coverage-local.sh` | x | x | x |
| `install-all.ps1` | ? | x | ? |
| `install-all.sh` | x | x | x |
| `register-plugins.nu` | x | x | x |
| `uninstall-all.sh` | x | x | x |

View file

@ -1,6 +1,10 @@
#!/usr/bin/env bash
set -euo pipefail
DIR=$(readlink -f $(dirname "${BASH_SOURCE[0]}"))
REPO_ROOT=$(dirname $DIR)
echo "---------------------------------------------------------------"
echo "Building nushell (nu) with dataframes and all the plugins"
echo "---------------------------------------------------------------"
@ -15,11 +19,17 @@ NU_PLUGINS=(
)
echo "Building nushell"
cargo build --features=dataframe
(
cd $REPO_ROOT
cargo build --features=dataframe
)
for plugin in "${NU_PLUGINS[@]}"
do
echo '' && cd crates/"$plugin"
echo "Building $plugin..."
echo "-----------------------------"
cargo build && cd ../..
(
cd "$REPO_ROOT/crates/$plugin"
cargo build
)
done

36
scripts/build-all.nu Normal file
View file

@ -0,0 +1,36 @@
print '-------------------------------------------------------------------'
print 'Building nushell (nu) with dataframes and all the plugins'
print '-------------------------------------------------------------------'
let repo_root = ($env.CURRENT_FILE | path dirname -n 2)
def build-nushell [] {
print $'(char nl)Building nushell'
print '----------------------------'
cd $repo_root
cargo build --features=dataframe
}
def build-plugin [] {
let plugin = $in
print $'(char nl)Building ($plugin)'
print '----------------------------'
cd $'($repo_root)/crates/($plugin)'
cargo build
}
let plugins = [
nu_plugin_inc,
nu_plugin_gstat,
nu_plugin_query,
nu_plugin_example,
nu_plugin_custom_values,
nu_plugin_formats,
]
for plugin in $plugins {
$plugin | build-plugin
}

60
scripts/coverage-local.nu Executable file
View file

@ -0,0 +1,60 @@
#!/usr/bin/env nu
def compute-coverage [] {
cd ($env.CURRENT_FILE | path dirname -n 2)
print "Setting up environment variables for coverage"
# Enable LLVM coverage tracking through environment variables
# show env outputs .ini/.toml style description of the variables
# In order to use from toml, we need to make sure our string literals are single quoted
# This is especially important when running on Windows since "C:\blah" is treated as an escape
cargo llvm-cov show-env | str replace (char dq) (char sq) -a | from toml | load-env
print "Cleaning up coverage data"
cargo llvm-cov clean --workspace
print "Building with workspace and profile=ci"
# Apparently we need to explicitly build the necessary parts
# using the `--profile=ci` is basically `debug` build with unnecessary symbols stripped
# leads to smaller binaries and potential savings when compiling and running
cargo build --workspace --profile=ci
print "Running tests with --workspace and profile=ci"
cargo test --workspace --profile=ci
# You need to provide the used profile to find the raw data
print "Generating coverage report as lcov.info"
cargo llvm-cov report --lcov --output-path lcov.info --profile=ci
}
let start = (date now)
# Script to generate coverage locally
#
# Output: `lcov.info` file
#
# Relies on `cargo-llvm-cov`. Install via `cargo install cargo-llvm-cov`
# https://github.com/taiki-e/cargo-llvm-cov
# You probably have to run `cargo llvm-cov clean` once manually,
# as you have to confirm to install additional tooling for your rustup toolchain.
# Else the script might stall waiting for your `y<ENTER>`
# Some of the internal tests rely on the exact cargo profile
# (This is somewhat criminal itself)
# but we have to signal to the tests that we use the `ci` `--profile`
let-env NUSHELL_CARGO_TARGET = "ci"
# Manual gathering of coverage to catch invocation of the `nu` binary.
# This is relevant for tests using the `nu!` macro from `nu-test-support`
# see: https://github.com/taiki-e/cargo-llvm-cov#get-coverage-of-external-tests
compute-coverage
let end = (date now)
print $"Coverage generation took ($end - $start)."
# To display the coverage in your editor see:
#
# - https://marketplace.visualstudio.com/items?itemName=ryanluker.vscode-coverage-gutters
# - https://github.com/umaumax/vim-lcov
# - https://github.com/andythigpen/nvim-coverage (probably needs some additional config)

View file

@ -1,5 +1,8 @@
#!/usr/bin/env bash
DIR=$(readlink -f $(dirname "${BASH_SOURCE[0]}"))
REPO_ROOT=$(dirname $DIR)
# Script to generate coverage locally
#
# Output: `lcov.info` file
@ -20,16 +23,19 @@ export NUSHELL_CARGO_TARGET=ci
# This is relevant for tests using the `nu!` macro from `nu-test-support`
# see: https://github.com/taiki-e/cargo-llvm-cov#get-coverage-of-external-tests
# Enable LLVM coverage tracking through environment variables
source <(cargo llvm-cov show-env --export-prefix)
cargo llvm-cov clean --workspace
# Apparently we need to explicitly build the necessary parts
# using the `--profile=ci` is basically `debug` build with unnecessary symbols stripped
# leads to smaller binaries and potential savings when compiling and running
cargo build --workspace --profile=ci
cargo test --workspace --profile=ci
# You need to provide the used profile to find the raw data
cargo llvm-cov report --lcov --output-path lcov.info --profile=ci
(
cd $REPO_ROOT
# Enable LLVM coverage tracking through environment variables
source <(cargo llvm-cov show-env --export-prefix)
cargo llvm-cov clean --workspace
# Apparently we need to explicitly build the necessary parts
# using the `--profile=ci` is basically `debug` build with unnecessary symbols stripped
# leads to smaller binaries and potential savings when compiling and running
cargo build --workspace --profile=ci
cargo test --workspace --profile=ci
# You need to provide the used profile to find the raw data
cargo llvm-cov report --lcov --output-path lcov.info --profile=ci
)
# To display the coverage in your editor see:
#

View file

@ -1,7 +1,10 @@
#!/usr/bin/env bash
# Usage: Just run `./install-all.sh` in nushell root directory
set -euo pipefail
DIR=$(readlink -f $(dirname "${BASH_SOURCE[0]}"))
REPO_ROOT=$(dirname $DIR)
echo "-----------------------------------------------------------------"
echo "Installing nushell (nu) with dataframes and all the plugins"
echo "-----------------------------------------------------------------"
@ -9,7 +12,7 @@ echo ""
echo "Install nushell from local..."
echo "----------------------------------------------"
cargo install --force --path . --features=dataframe
cargo install --force --path "$REPO_ROOT" --features=dataframe
NU_PLUGINS=(
'nu_plugin_inc'
@ -26,5 +29,5 @@ do
echo "----------------------------------------------"
echo "Install plugin $plugin from local..."
echo "----------------------------------------------"
cd crates/"$plugin" && cargo install --force --path . && cd ../../
cargo install --force --path "$REPO_ROOT/crates/$plugin"
done