Apply linter suggestions for cleaner code

[Lingepumpe]

This PR is on top of the FitSDK 21.141.00 branch.

In this PR additionally to the SDK update, I fixed issues that the linter (clippy) was pointing out. This also involved fixes in autogenerated profile/decode.rs and profile/field_types.rs files - for this I adapted the code that generates them. Additionally I added cargo clippy --fix and cargo fmt to the update_profile.sh shell script, as for profile/field_types.rs this leads to a nice collapsing of huge match "|" cascades into simple ranges, which is much more readable (and maybe even faster, I am not sure what the compiler does).

[stadelmanma]

Thanks for looking at this, I'm fine with the clippy fixes in the main "human written" code base, those are good. However, I don't think I want to apply them to the auto-generated code at the tail end of the shell script. Clippy re-writes things in a nontrivial way at times, and that will make it harder to track down bugs since the code we see getting written in the generation step doesn't exist 1:1 in the code being run anymore.

That being said, using clippy to suggest ways we can improve code generation does seem like a very good idea, especially if they would be trivial to implement like where it changed .map().flatten() to an .and_then()

Also I call "rustfmt" via code in the main script,

fitparse-rs/generate-fit-profile/src/main.rs

Line 31 in 1fdba05

Command::new("rustfmt")

so the extra call in the shell script isn't needed. But maybe it was the application of clippy that required that to be done again.

[Lingepumpe]

Yes, exactly, the final "cargo fmt" was needed due to the auto-fixes. If you prefer I will just remove the automatic fixing after autogeneration entirely, but I really do like the changes it makes (it is only one change, fixing the clippy::manual_range_patterns lint). So for now I have adjusted the cargo clippy call to be restricted to only fixing manual_range_patterns.
The fix it applies always looks like this:

But I am also fine dropping the auto fixing part of my PR if you prefer, let me know.

[stadelmanma]

I think for now I'd like to drop any automatic code changes (other than the programmatically called rustfmt), so more or less removing the changes to bin/update_profile.sh. The profile code is impossible to fully validate at this time so I'd like to keep it's creation as straight forward as possible instead of adding another layer of complexity.

I love the other changes, and wouldn't be opposed to updating the codegen script to create the desired output for matches! on large ranges directly. Same for anything else suggested by clippy that makes sense, although I'm not sure how far down the rabbit hole we should go, since those files aren't really meant to be "read" by humans.

[Lingepumpe]

As requested, I have dropped the automatic code changes (clippy fix + cargo fmt) from update_profile.sh. Instead, I changed the code that generates profile/field_types.rs to detect ranges and generate the first..=last syntax in this case.

Also, I rebased the branch on master, as the FitSDK 21.141.00 branch was merged.

[stadelmanma]

This looks great now, I'll do some local testing probably tomorrow and get it merged in!

[stadelmanma]

Thanks!