https://github.com/jjlin/docker-image-extract
Minimal-dependency shell script to pull and extract all files from an image in Docker Hub
https://github.com/jjlin/docker-image-extract
docker docker-image shell shell-script
Last synced: 5 months ago
JSON representation
Minimal-dependency shell script to pull and extract all files from an image in Docker Hub
- Host: GitHub
- URL: https://github.com/jjlin/docker-image-extract
- Owner: jjlin
- License: mit
- Created: 2020-05-03T03:51:45.000Z (about 6 years ago)
- Default Branch: main
- Last Pushed: 2023-08-09T05:46:00.000Z (almost 3 years ago)
- Last Synced: 2025-08-02T21:51:36.938Z (11 months ago)
- Topics: docker, docker-image, shell, shell-script
- Language: Shell
- Homepage:
- Size: 40 KB
- Stars: 103
- Watchers: 7
- Forks: 21
- Open Issues: 3
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
The `docker-image-extract` script pulls and extracts all files from an image
in [Docker Hub](https://hub.docker.com/). For multi-platform images, you can
choose the platform-specific image to pull.
`docker-image-extract` has minimal dependencies that should exist in pretty
much any Linux environment, with the possible exception of `curl`/`wget`:
* `curl` or `wget`
* `find`
* `grep`
* `gzip`
* `mkdir`
* `rm`
* `sed`
* `sh`
* `tar`
If you are using [BusyBox](https://busybox.net/), version 1.34.0 or later is
required, as the `wget` implementation in earlier versions does not recognize
the HTTP 307/308 redirects returned by Docker Hub.
## Usage
```
./docker-image-extract [OPTIONS...] IMAGE[:REF]
IMAGE can be a community user image (like 'some-user/some-image') or a
Docker official image (like 'hello-world', which contains no '/').
REF is either a tag name or a full SHA-256 image digest (with a 'sha256:' prefix).
The default ref is the 'latest' tag.
Options:
-p PLATFORM Pull image for the specified platform (default: linux/amd64)
For a given image on Docker Hub, the 'Tags' tab lists the
platforms supported for that image.
-o OUT_DIR Extract image to the specified output dir (default: ./output)
-h Show help with usage examples
Examples:
# Pull and extract all files in the 'hello-world' image tagged 'latest'.
$ ./docker-image-extract hello-world:latest
# Same as above; ref defaults to the 'latest' tag.
$ ./docker-image-extract hello-world
# Pull the 'hello-world' image for the 'linux/arm64/v8' platform.
$ ./docker-image-extract -p linux/arm64/v8 hello-world
# Pull an image by digest.
$ ./docker-image-extract hello-world:sha256:90659bf80b44ce6be8234e6ff90a1ac34acbeb826903b02cfa0da11c82cbc042
```
## Sample output
```
$ ./docker-image-extract busybox:musl
Getting API token...
Getting image manifest for library/busybox:musl...
Fetching and extracting layer e4df39e4f349fc83694349a5a94afc8d9857fa66f7421427f3d2463a58d876d3...
Image contents extracted into ./output.
$ ls output
bin dev etc home root tmp usr var
$ ls -l output/bin/{busybox,ls}
-rwx------ 393 jjlin users 1148648 Apr 7 13:18 output/bin/busybox
-rwx------ 393 jjlin users 1148648 Apr 7 13:18 output/bin/ls
```
## See also
* https://rmannibucau.metawerx.net/post/docker-extracts-fileystem-with-bash
The logic in this script is based on this blog post, but pretty much
everything has been rewritten and improved.
* https://github.com/moby/moby/blob/master/contrib/download-frozen-image-v2.sh
This script downloads the layers of a Docker image in a form suitable for use
with [`docker load`](https://docs.docker.com/engine/reference/commandline/load/).
It doesn't actually extract the layers, though.