change case defaults

[burrbull]

Closes #803

cc @jonathanpallant, @Emilgardis

[Emilgardis]

I think it would be worth with this to add #[doc(alias = "<peripheral name in svd>")]

[burrbull]

Looks like starting _/digit sometimes incompatible with PascalCase.

[burrbull]

/ci diff pr

[burrbull]

I don't have ideas what to do with Freescale.

[Emilgardis]

ugh, why does the /ci triggers fail here, I have no issues with https://github.com/cross-rs/cross/blob/main/.github/workflows/try.yml ...

/ci diff pr

[github-actions]

Diff for comment

• pr

[Emilgardis]

also, sometimes the github action log messes up whitespace with colors

but, diff looks good, except

     impl RegisterBlock {
         ///0x04..0x14 - SRAM/NOR-Flash chip-select timing register %s
         #[inline(always)]
-        pub const fn btr(&self, n: usize) -> &BTR {
+        pub const fn Btr(&self, n: usize) -> &Btr {
             #[allow(clippy::no_effect)] [(); 4][n];
             unsafe { &*(self as *const Self).cast::<u8>().add(4).add(8 * n).cast() }
         }

should be btr, not Btr

[burrbull]

ugh, why does the /ci triggers fail here, I have no issues with

Are you sure someone else except you can run this?

[Emilgardis]

There's nothing in it that would prohibit other members to run it, the only check is that the user making the comment is a member, which we both are

[burrbull]

/ci diff pr

[Emilgardis]

    {
      "id": "IC_kwDOBDHJwM5yw0Sf",
      "author": {
        "login": "burrbull"
      },
      "authorAssociation": "MEMBER",

I don't understand...

/ci diff pr

[github-actions]

Diff for comment

• pr

[burrbull]

@Emilgardis run ci yet one time, please.

[Emilgardis]

can you try it, this time put some text before

[Emilgardis]

/ci diff pr

[github-actions]

Diff for comment

• pr

[Emilgardis]

/ci diff --pager "git diff --no-index" pr

[github-actions]

Diff for comment

• --pager "git diff --no-index" pr

[burrbull]

I'm not sure myself if that is the best way to go about it, maybe we could have "themes" where there's a default theme (the new behaviour), and another theme that looks like how it used to look like. You can still set individual formats if wanted.
so maybe

ident_formats_theme = "legacy"

Done. Not sure about new theme name.

[Emilgardis]

Done. Not sure about new theme name.

We could also not have a new theme, e.g not expose it

[Emilgardis]

maybe default?

With this variant though (legacy and default/new), we get into a issue of versioning, where if we change the theme we'd probably have to introduce a new theme for the previous version. I think only having legacy makes sense for now. Maybe there's an argument for having ranged theme names, such that saying theme = "0.35.1" means you get the appropriate theme for that version, but sounds like hell to maintain and keep track of

[thejpster]

2021 and 2024? (Or whatever year makes sense)

[burrbull]

@Emilgardis Can you fix svd2rust-regress "unused" warnings?