Add recursivefill! and recursivecopy! for ragged arrays

[JoshuaLampert]

The functions recursivefill! and recursivecopy! were missing for the RecursiveArrayToolsRaggedArrays.jl sublibrary. Previously, the following MWE did not work:

using RecursiveArrayToolsRaggedArrays
using OrdinaryDiffEqTsit5

function rhs!(du, u, p, t)
    du .= u
    return nothing
end

n = 2
left = ones(n)
right = left

u = VectorOfArray([copy(left), copy(right)])

tspan = (0.0, 1.0)
prob = ODEProblem(rhs!, u, tspan)

solver = Tsit5()
kwargs = (; dt=0.1, adaptive=false, save_everystep=false)

sol = solve(prob, solver; kwargs...)

This PR fixes it.

Checklist

• Appropriate tests were added
• Any code changes were done in a way that does not break public API
• All documentation related to code changes were updated
• The new code follows the
  contributor guidelines, in particular the SciML Style Guide and
  COLPRAC.
• Any new documentation only uses public API

Additional context

Add any other context about the problem here.

[JoshuaLampert]

I'm testing this PR in JoshuaLampert/SimpleDiscontinuousGalerkin.jl#74, where CI is still failing. I'm investigating.

[JoshuaLampert]

With this PR, the tests in https://github.com/JoshuaLampert/SimpleDiscontinuousGalerkin.jl/actions/runs/24992576972/job/73220281516 and the ragged array tests here are now passing. In addition to recursivefill! and recursivecopy! this also fixes mapreduce for nested RaggedVectorOfArray and adds an extension for DiffEqBase.jl, such that the correct ODE_DEFAULT_NORM is used for problems using RaggedVectorOfArray as underlying data structure.
I used Claude to help me write the code.

[ChrisRackauckas]

This is circular: these overloads need to go into DiffEqBase instead.

[JoshuaLampert]

Totally fine with me. Do you (or one of your bots) create a PR in DiffEqBase.jl? Please ping me there if there is one.

[JoshuaLampert]

I created SciML/OrdinaryDiffEq.jl#3572.

[ChrisRackauckas]

yeah sorry Mr. Bot is a bit slow right now with all the travel and the big stuff of the recent majors.

[JoshuaLampert]

Yeah, no worries. I removed the extension again as this will be handled in DiffEqBase.jl now. So I think this is ready from my side.