CLI doesn't have a caching strategy

[GalenReich]

At the moment the CLI recomputes the grid of timezones (the most expensive part of the computation) every time it is executed. Which makes the process very slow if a user calls shadowfinder multiple times.

Instead it would be good if a user could provide a path to a timezone json file.

It would also be good if there was a subcommand to generate the timezone file.

The end result would be something like:

shadowfinder generate_timezone_grid # creates timezone_grid.json in the working directory (or the given path?)
shadowfinder find 10 5 2024-02-29 13:59:59 --time_format=utc --grid=timezone_grid.json

[borisnezlobin]

I'm new here, but I can work on this. Does this entail just adding a CLI command to do essentially what is in the README (and already written in Python, just not the CLI)?

[GalenReich]

There are a few parts to this:

• Add a new CLI subcommand called generate_timezone_grid that calls finder.generate_timezone_grid() and finder.save_timezone_grid() as in the Notebook implementation
• Modify the existing find CLI subcommand to take an optional argument grid that loads the generated file before running the search
• Modify the quick_find function (or refactor it away)

  ShadowFinder/src/shadowfinder/shadowfinder.py

  Lines 96 to 99 in e5f636b

  	def quick_find(self):
  	self.generate_timezone_grid()
  	self.find_shadows()
  	fig = self.plot_shadows()

  to avoid generating the grid if it has been loaded by the new grid argument

Hope that helps! Let me know if you run into any issues when getting set up.

[borisnezlobin]

Thanks Galen! I've done these things (correctly, I hope) and made a PR (#21). Could you look over it and let me know if it needs changes?