babashka.fs

File system utilities for babashka. This library isn't part of babashka itself yet, but will be. Also can be used from the JVM.

Status

Working towards an initial release. The API might still change. Awaiting user feedback.

Why

Babashka is a scripting utility. It's convenient to have cross platform file system utilities available for scripting. The namespace clojure.java.io already offers a bunch of useful features, but isn't up to date with java.nio.

The main inspiration for this library is clj-commons/fs but some of its functions can be optimized by leveraging java.nio and others might need revision. We do not guarantee any compatibility with the clj-commons library.

API docs

API docs are available at babashka.org/fs. Most functions take a string, java.io.File or java.nio.file.Path as input and return a java.nio.file.Path. Coercion into a File or Path can be done using fs/file and fs/path.

Usage

Until this library is included in babashka, you can use this library as a git lib:

#!/usr/bin/env bb

(require '[babashka.deps :as deps])
(deps/add-deps '{:deps {babashka/fs
                        {:git/url "https://github.com/babashka/fs"
                         :sha "b008b2d6ea64e49bf76066b7f057ae4638534e35"}}})

or use any later SHA.

On the JVM, add it to deps.edn.

(require '[babashka.fs :as fs])

Examples

The glob function takes a root path and a pattern. The pattern is interpreted as documented here.

(map str (fs/glob "." "**{.clj,cljc}"))

Output:

("project.clj" "test/babashka/fs_test.clj" "src/babashka/fs.cljc")

The function exec-path returns directories from the PATH. To search all these directories for an executable, e.g. java, you can combine it with list-dirs which searches files directly in the directories using an (optional) glob pattern:

(str (first (filter fs/executable? (fs/list-dirs (filter fs/exists? (fs/exec-path)) "java"))))
"/Users/borkdude/.jenv/versions/11.0/bin/java"

For convenience, the above use case is also supported using the which function:

(str (fs/which "java"))
"/Users/borkdude/.jenv/versions/11.0/bin/java"

Test

$ clojure -M:test

Prior art

• nate/fs
• hara.io
• clj-commons/fs
• Python pathlib

Codox

Static files including compiled JS are hosted on Github. This is set up like described here:

All the commands below assume that you already have a git project initialized and that you are in its root folder.

# Create an orphan branch named gh-pages
git checkout --orphan gh-pages
# Remove all files from staging
git rm -rf .
# Create an empty commit so that you will be able to push on the branch next
git commit --allow-empty -m "Init empty branch"
# Push the branch
git push origin gh-pages

Now that the branch is created and pushed to origin, let’s configure the worktree correctly:

# Come back to master
git checkout master
# Add gh-pages to .gitignore
echo "gh-pages/" >> .gitignore
git worktree add gh-pages gh-pages

After cloning this repo to a new dir:

git fetch origin gh-pages
git worktree add gh-pages gh-pages

To deploy to Github Pages:

script/release

License

Copyright © 2020-2021 Michiel Borkent

Distributed under the EPL License, same as Clojure. See LICENSE.