Minimal annot invalidation workaround

[tecosaur]

After discussion with Cody, this is an attempt at a more minimal version of #56194.

The intent is to work around the invalidation issue introduced by the split design with AnnotatedStrings, resolving the headache for 1.12.

Paired with JuliaLang/StyledStrings.jl#115

Many thanks to @topolarity for working out this approach, the discussion on balancing the short/long term fixes for this issue.

---

From my understanding of the problem, this should fix the write/print/show invalidations, but this needs to be checked.

[topolarity]

Just to clarify the approach here:

This does not resolve any actual type-piracy. Instead, it inserts an invalidation barrier around the existing StyledStrings type-piracy , similar to what we do in Logging. This should help with latency, but it does not resolve issues like JuliaLang/StyledStrings.jl#91

My personal preference would be one of the other pathways we've sketched that resolve type-piracy fully, but @tecosaur wanted an interim solution more quickly

[KristofferC]

Some timings w.r.t #57997 (comment) could perhaps be done to check if this approach fixes the symptom of the piracy.

[KristofferC]

This seems to work ok to fix the #57997 latency:

julia> using Pkg


julia> Pkg.UPDATED_REGISTRY_THIS_SESSION[] = true
true

julia> ENV["JULIA_PKG_PRECOMPILE_AUTO"] = 0
0

julia> @time @eval Pkg.add("Example")
   Resolving package versions...
No packages added to or removed from `~/.julia/environments/v1.12/Project.toml`
No packages added to or removed from `~/.julia/environments/v1.12/Manifest.toml`
  0.207511 seconds (776.73 k allocations: 62.073 MiB, 8.57% gc time, 67.03% compilation time: <1% of which was recompilation)

Also better with e.g.

# branch
julia> @time @eval Base.Precompilation.precompilepkgs()
  0.049761 seconds (171.77 k allocations: 9.485 MiB, 96.97% compilation time: 86% of which was recompilation)

# 1.12-nightly
julia> @time @eval Base.Precompilation.precompilepkgs()
  0.235922 seconds (1.17 M allocations: 60.883 MiB, 17.97% gc time, 97.54% compilation time: 97% of which was recompilation)

[topolarity]

CI is looking good. Does the StyledStrings PR need to merge before we merge this one?

[tecosaur]

Nope, I can (will) fast-forward merge, putting the referenced commit on main without changing the hash — if CI hadn't spotted any issues and it looks good to you, I think we're good to merge this 🙂

[topolarity]

LGTM.

Thanks for taking action on this issue @tecosaur, this should relieve invalidation pains significantly on 1.12