Dev install

[Thumuss]

This PR addresses install command and workflow improvements.

Changes

Workflow Updates:

• Updated Rust toolchain from 1.85 to 1.89 in CI workflows
• Re-enabled clippy, rustfmt, and doc generation checks

Command Restructuring:

• Renamed workspace command to project (with prj alias)
• Enhanced sync command functionality

Code Quality:

• Replaced f.len() > 0 with !f.is_empty() for better readability
• Improved error handling and validation logic
• Removed unused imports and cleaned up dependencies

Documentation:

• Updated README to reflect command changes (workspace → project)
• Updated feature descriptions and examples

Build Configuration:

• Simplified Cargo.toml feature flags
• Removed justfile configuration

Tests:

• Added unit-testing

Breaking Changes

• The workspace command is now project (prj for short)
• Removed add, delete and bulk_delete commands

[Thumuss]

next regexs but tmrw

[SillyFreak]

Apart from the comments on individual code pieces, this is what stood out to me while looking through the changes:

commit messages: the term "refactor" only applies for changes that don't affect functionality, not to e.g. "install working", "remove bulk_delete"

in general, is it useful to have features for every individual command? If we want them: i think the code for individual features should be better isolated. E.g. the regexes in utils.rs all have their own #[cfg(feature = "...")]. if a regex is really command specific, I think it's better to declare it in the command's module.

In general, prefer functions over macros. using two different functions if there are different arguments is preferred in Rust.

• good macro uses: utpm_bail (does control flow), utpm_log (useful variadics for formatting)
• not so good: write_manifest/read_manifest (could be two functions each, where the one delegates to the other by specifying the default value, e.g. try_find and try_fin_in)
• format_package: it makes sense that there are multiple options, but I feel this could be better represented by implementing a trait on PackageSpec and VersionlessPackageSpec, as well as maybe on a new struct Namespace(pub String);

none of these things are meant as things to be fixed before this PR can be merged, just as things to consider in the future. (And considering the comments doesn't mean they need to be followed in every situation either, of course.)

[Thumuss]

• format_package: it makes sense that there are multiple options, but I feel this could be better represented by implementing a trait on PackageSpec and VersionlessPackageSpec, as well as maybe on a new struct Namespace(pub String);

Looks like a bit complicated to impl... I use an other way to impl this on other command. I think I'll stick to that for now since I can't find a way I like

[Thumuss]

I have removed format_package! but I didn't create a trait, as I didn't find a real use-case other than one command.
I will have that in mind tho

[SillyFreak]

Assuming you have tested this, the latest changes look good. in state.rs there are many empty lines introduced; I commented on one example

[SillyFreak]

Looks like the #[cfg] line should have been removed, not blanked (many other occurrences in this file)

[Thumuss]

Assuming you have tested this [...]

I didn't! I'll make unit-tests to make things easier to test

[Thumuss]

I'll try to fix all bugs mentioned before merging too
Testing is on his way too