[Feature] Extract compressed files bash function

[Mte90]

We can integrate https://github.com/xvoland/Extract/blob/master/extract.sh as extract for tons of formats (or rewrite in amber).

Or also this one https://github.com/Bash-it/bash-it/blob/master/plugins/available/extract.plugin.bash

[Ph0enixKM]

This is a great idea!

[b1ek]

it can be added to stdlib (but it will be needed to be rewritten in amber)

[Mte90]

@rbtylee shared to me his version of extract written in Amber:

// extract.ab: An amber extraction library function
//
// Copyright 2024 Robert Wiley <ylee@bodhilinux.com>
//
// This work is free. You can redistribute it and/or modify it under the
// terms of the Do What The Fuck You Want To Public License, Version 2,
// as published by Sam Hocevar. See http://www.wtfpl.net/ for more details.

// Amber version of Bash-it extract function (simplified somewhat)
// Added mkdir repects verbosity
//       Only process if file exist

// This is a work in progress, not everything may work right
// not well tested and no tests

import * from "std"

pub fun glob(path:Text, pattern:[Text]): Bool
{
    loop p in pattern {
        $[[ {path} =~ {p}\$ ]]$ failed {
            continue}
        return true
        
    }
    return false
}

pub fun extract(path: Text, verbose: Bool): Null{
    if file_exist(path) {
        let filename = unsafe $basename -- {path}$
        let dirname  = unsafe $dirname -- {path}$
        if verbose : let v = "v"
        else: let v = ""

        let target = unsafe $sed 's/\(\.tar\.bz2\$\|\.tbz\$\|\.tbz2\$\|\.tar\.gz\$\|\.tgz\$\|\.tar\$\|\.tar\.xz\$\|\.txz\$\|\.tar\.Z\$\|\.7z\$\|\.nupkg\$\|\.zip\$\|\.war\$\|\.jar\$\)//g' <<< {filename}$
        
        if {
            filename == target: target = ""
            verbose: unsafe  $mkdir -{v} {dirname}/{target}$
            else: unsafe  $mkdir {dirname}/{target}$
        }
        
        if {
            glob(path, [".tar.bz2", ".tbz", ".tbz2"]): unsafe $tar "x{v}jf" "{path}" -C "{dirname}/{target}"$
            glob(path, [".tar.gz", ".tgz"]): unsafe $tar "x{v}zf" {path} -C {dirname}/{target}$
            glob(path, [".tar.xz", ".txz"]): unsafe $tar "x{v}Jf" {path} -C {dirname}/{target}$
            glob(path, [".tar.Z"]): unsafe $tar "x{v}Zf" {path} -C {dirname}/{target}$
            glob(path, [".bz2"]): unsafe $bunzip2 {path}$
            glob(path, [".deb"]): unsafe $dpkg-deb -x{v} {path} "\$\{1:0:-4}"$
            glob(path, [".pax.gz"]): unsafe $gunzip {path}; set -- "\$@" "\$\{1:0:-3}"$
            glob(path, [".gz"]): unsafe $gunzip {path}$
            glob(path, [".pax"]): unsafe $pax -r -f {path}$
            glob(path, [".pkg"]): unsafe $pkgutil --expand {path} "\$\{1:0:-4}"$
            glob(path, [".rar"]): unsafe $unrar x {path}$
            glob(path, [".rpm"]): unsafe $rpm2cpio {path} | cpio -idm{v}$
            glob(path, [".tar"]): unsafe $tar "x{v}f" {path} -C {dirname}/{target}$
            glob(path, [".xz"]): unsafe $xz --decompress {path}$
            glob(path, [".zip", ".war", ".jar", ".nupkg"]): unsafe $unzip {path} -d {dirname}/{target}$
            glob(path, [".Z"]): unsafe $uncompress {path}$
            else: echo "Error: Unsupported file type"
        }    
    }
    else: echo "Error: File not found"
}

I think that is perfect, we need just to remove the .z/pax files as they are very uncommon and add a check with is_command for the various cases.

[b1ek]

this would create confusion with the runtime dependency checker feature, since it will check for every function from there

[Mte90]

That's true so maybe the stdlib should not this kind of check for command check.

[rbtylee]

Perhaps amber can include more than the std library and if you want extract include it in another library like util for example. Do not run the runtime dependency checker on libs found in util. (I am assuming the runtime dependency checker is something that lives in the rust code).

[Ph0enixKM]

@b1ek I think that we need to disable RDC for unsafe commands

[Mte90]

To proceed with this one I think that we need to define and implement the RDC as it is a very complex case.