• portability - write as a command line tool

This script is the beginning of a more responsible approach to using precious SSD space. It aims to make staying neat and space-efficient easy by showing users the directories that are actively used versus stale versus ready for archiving.

One of the principles of responsible resource management is conservation. The greenest building is the one that is never built. Similarly, disk space is something that should be conserved (I think).

• TODO: Make this even easier. This is not very user friendly installation. Not user friendly usage either.
• TODO: Go to PyPA to learn how to package this up for PyPI, see sample repo and slack-cleaner for inspiration

clone the repo: git clone
create a virtual environment python -m venv venv
activate it . venv/bin/activate
install the module into it pip install --editable
run the tests pytest tests
run it our_du .

• TODO: Make it actually give results that are readable and sorted by age and size.

Set up a virtual environment using python -m venv virtual_env. Make sure to have wheel installed per the [Python Packaging Authority][PyPA] - not sure why.

Get required python package dependencies using pip install -r requirements.txt.

We use setuptools develop mode. pip install --editable . This workflow involves installing our module into the local virtualenv.

• pytest - testing framework

I activate the virtual environment to run tests using pytest tests -v. I suck at testing so I just chose something off-the-shelf.

In the early discovery stages of the project I wrote wrappers around GNU stat and GNU find, and did quite a bit of learning about those command line tools/how to use them to get file size and access time data.

If you are interested for academic purposes you can see how things looked then:

git checkout -b discover discovery

The Python [subprocess][subprocess] library as proposed in PEP is a great way to make Python a replacement language for over-complicated shell scripts.

One of the main ways to gain insight about disk usage is the GNU coreutil [du][du].

With [subprocess][subprocess], we can take the output of [du][du] and manipulate it with Python.

This script relies heavily on GNU coreutils. In essence, we are trying to a reasonable front-end to core system tools that will help us efficiently manage disk space.

[stat][stat] can return the birth date, most recent access date, and the file size of any file.

[du][du] can return file sizes for all files within a file hierarchy.

[fts][fts] helps us traverse the file hierarchy.

[find][find] traverses the file hierachy while running [stat][stat] along the way. It provides options for filtering the file hierarchy as you go so you can find what you are interested in. It is part of the GNU [findutils][findutils].