How do I "pretend" to install a package using pip?

Question

This feels like such a simple question, but I can't find any reference in the pip documentation and the only question that seemed relevant mentions a flag that has apparently been deprecated since version 1.5 (version 8.1 is current at the time of this writing).

How do I "pretend" to install a package or list of packages using pip, without actually installing them? I have two separate use cases for this:

• I need to see what packages out of a long (~70 line) requirements.txt are missing, without actually installing them; seeing what requirements are already satisfied without installing the missing requirements would satisfy this for me.
• Finding the dependencies for a package that I have not yet installed on my computer, without using something like Portage or Aptitude.

Answer 1

There is also the pretty useful pip-tools package that provides a pip-sync tool which you can execute in a "dry run" mode against your requirements file(s):

$ mkvirtualenv test_so
New python executable in test_so/bin/python
Installing setuptools, pip, wheel...done.
...
(test_so) $ pip install pip-tools
...
Installing collected packages: six, click, first, pip-tools
(test_so) $ echo "Django==1.6.11" > requirements.txt
(test_so) $ pip-sync --dry-run requirements.txt 
Would install:
  Django==1.6.11

Also, here is a partially relevant thread: Check if requirements are up to date.

Answer 2

Pip (v.23.2.1) has the --dry-run option. From the documentation:

Don’t actually install anything, just print what would be. Can be used in combination with --ignore-installed to ‘resolve’ the requirements.

You can it use it as

pip install poetry --dry-run

This will print normal output without actually installing anything.

Answer 3

Per the pip documentation, the proper way to generate the requirements.txt file is via pip freeze > requirements.txt. Hopefully this is what you wanted.