docker-ubuntu-with-pipenv

Dockerfile that generates a docker image with Ubuntu and a user motoko that has pyenv with python, pip, and pipenv pre-installed.

When a container is run using this image, the commands are run automatically as motoko user. However, motoko will be able to run any command with the sudo command without need of authenticating itself.

This is intended for simulating a real-case scenario, where a user wants to have pyenv installed at a user level, to avoid messes with root bins.

Example 1: interactive mode

docker container run -it francesco086/pipenv

motoko@...:~$ pyenv versions
* 3.7.2 (set by /home/motoko/.pyenv/version)

motoko@...:~$ mkdir myproject
motoko@...:~$ cd myproject
motoko@...:~$ pipenv install requests
motoko@...:~$ ls
Pipfile  Pipfile.lock

Example 2: build image with pipenv already installed

You have a Pipfile:

[[source]]
name = "pypi"
url = "https://pypi.org/simple"
verify_ssl = true

[dev-packages]

[packages]
pandas = "*"
numpy = "*"
sqlalchemy = "*"
scipy = "*"

[requires]
python_version = "3.7"

You can generate a docker image (say you/my_env) with a Dockerfile like this:

FROM francesco086/pipenv

COPY Pipfile /app/
WORKDIR /app

RUN pipenv install

Then you can do:

docker container run -it you/my_env

motoko@...:~$ pipenv shell
Launching subshell in virtual environment…
 . /home/motoko/.local/share/virtualenvs/app-4PlAip0Q/bin/activate
motoko@eac5a40ca71f:/app$  . /home/motoko/.local/share/virtualenvs/app-4PlAip0Q/bin/activate

motoko@...:~$ ipython
Python 3.7.2 (default, Mar 13 2019, 14:41:35) 
Type 'copyright', 'credits' or 'license' for more information
IPython 7.3.0 -- An enhanced Interactive Python. Type '?' for help.

In [1]:  import pandas as pd

In [2]: pd.Series([1,2,3]).apply(lambda x: x*2)                                                                                                                                               
Out[2]: 
0    2
1    4
2    6
dtype: int64

For the developers

In the root folder there is the Docker template, where one does not need to specify the Ubuntu tag to use (UBUNTU_TAG) and the python version (PYTHON_VERSION).

There will then be several folder with a name format ubuntu*-python* (e.g. ubuntu18.04-python3.7.2). The script propagate_dockerfile_to_subfolders.sh can be used to copy the Dockerfile template in each folder and set the proper UBUNTU_TAG and PYTHON_VERSION.

Thanks to this mechanism, it is possible to abstract from the specific versions of Ubuntu and python.