Changes to hardware initialization

[dbgen1]

Hardware initialization occurs in pysquared.py, where all the devices are initialized. Many of these devices the same behavior when it comes to their initialization, so it is beneficial to write functions that can handle initialization behavior. This PR does a couple things:

1. Add a init_general_hardware function which is able to handle the hardware initialization for any device that simply needs a firmware constructor. To keep the current code behavior unchanged, this function also optionally takes a callable which modifies the initialization behavior if self.orpheus is true. A detailed docstring in this function describes its behavior in more detail.
2. Moves the initialization of hardware devices into their own functions, which are called in the constructor of Satellite. Using separate single-use functions for implementation makes the code self-documenting and allows for easier testability as per (3).
3. Add a @safe_init decorator to the Satellite class which wraps all the initialization functions in the standard try catch used for every hardware instance with a defined error severity if needed, and can print detailed debug information if self.debug is enabled.

[Mikefly123]

Hey @dbgen1! The proposed changes sound like they could be pretty great. Particularly, the creation of individual functions for hardware initialization seems like a good abstraction to increase code readability and also probably allow us to individually deinit and reinit individual pieces of hardware more easily.

I had a separate question though about this PR, it looks like there are a bunch of files that are "changed" but are identical to the origin files. Not super familiar with why this happens sometimes with GitHub. Is it a file formatting issue?

[dbgen1]

I'm not sure why it happens. I noticed it as well and used a diff tool to verify that nothing important was actually changed. It might have to do with the way I cloned the repo to my machine?

[Mikefly123]

@nateinaction have you seen this happen before? I think it happened once or twice before to me on Mac. @dbgen1 are you using cloud syncing by any chance. I think this was happening to be when I was cloning my repos to a folder that was being backed up by my iCloud Drive and it went away once I started cloning to a folder that was not backed up by the iCloud.

I think I fixed the issue by removing and then recloning the repository.

[nateinaction]

Hmmm. This is pretty strange and I would think that the pre-commit hook would have caught this. It looks like an extra invisible char was added to every line by your editor. I suspect some delving into your editor config might solve this problem but you also might need to reapply your changes onto a new branch:

git diff origin/main --word-diff-regex=.
diff --git a/.github/workflows/ci.yaml b/.github/workflows/ci.yaml
index db70d31..c1b8104 100644
--- a/.github/workflows/ci.yaml
+++ b/.github/workflows/ci.yaml
@@ -1,53 +1,53 @@
name: ci{+^M+}
{+^M+}
on:{+^M+}
  pull_request:{+^M+}
  push:{+^M+}
    branches:{+^M+}
      - main{+^M+}
{+^M+}
jobs:{+^M+}
  lint:{+^M+}
...

Even though I've seen the code in office hours and like the idea, the PR isn't easily reviewable in this state. Blocking.

[dbgen1]

The issue was caused by the editor changing the whitespace at the end of each line from a Unix style (LF) line ending to a windows style (CRLF) line ending. I fixed this by configuring git to automatically convert the line ending style to the correct one used in this repository (LF). The latest commit contains all the actual changes and shows that there were no other modifications.

[nateinaction]

Much better! Looks like there was an issue with resolving a merge conflict that removed #87. Once that's fixed I'll test this on my board and give you a 👍

[dbgen1]

Going to do one more commit to appease the linter and this should be good to go!

[nateinaction]

It looks like this stops the hardware from being initialized but does not stop the hardware from being set to an initialized state in the hardware dictionary.

Device  | Status
|I2C0   | True |

[dbgen1]

I fixed this by moving the dictionary setting to each individual init function rather than the decorator, which is a method that makes more logical sense anyway.