Show where a Python package is used

June 12, 2015

A while back I was doing some spring cleaning of my Python packages. I noticed that there were a bunch of packages that I couldn’t recall installing. I wanted to know if I could safely remove them, so I wrote a simple bash script to tell me called pip-uses.


Rather than post the source here and let it get more out of date every time I change it, you can find the source on GitHub. It’s in my bin repository, where I keep my notable helper scripts; feel free to poke around.


I was primarily influenced by Homebrew’s brew uses command. It does a nice job of giving you exactly the information you want, and I think the way the command is named makes sense.

Homebrew: brew uses
$ brew uses --installed pango

pip-uses gives you basically the experience:

Pip: pip-uses
$ pip-uses stevedore

In this example, the Python package virtualenvwrapper uses stevedore, just as imagemagick uses pango. Both commands can save you from accidentally removing a crucial dependency and answer the burning question, “How in the world did this thing get installed?”

Wish List

I’m not doing much Python development these days, but if I had some spare time I’d love for the script to also have these features:

  • Recursive enumeration of dependencies
    • It’d be nice if pip-uses kept recursively searching until it found no more dependencies. This way, it’d be easy to see if you could safely uninstall a whole slew of packages that you’re no longer using.
  • Operate on more than one package
    • I didn’t need it at the time, so I didn’t implement it, but it’d be nice if the command took a variable amount of arguments and ran the same logic on all supplied packages.
  • Integrate with pip
    • Programs like brew and pip allow developers to add “external commands” by adding commands to the PATH that look like brew-xyz or git-xyz. I couldn’t find if there was a special way to add external commands to pip.

If you find this script useful and end up implementing one of these feature or more on top of pip-uses, be sure to send me a Pull Request!

Testing, Types, & Correctness

It’s important to both have strong testing practices and languages with disciplined type systems. The hardest part of writing quality software is ensuring that it runs without bugs. This is why testing and type systems are complementary—they’re distinct tools to help us write better code. Continue reading

If, Else, & Purity

Published on September 01, 2017

Haskell Development with Neovim

Published on July 16, 2017