allow vcat on single ChainedVector

[stuartthomas25]

Formerly, vcat(a::ChainedVector) failed with an error. This PR makes one-input vcat return an unaliased ChainedVector, matching the typical Julia behavior that vcat(a) == a but vcat(a) !== a.

It also adds wrapper-preserving unaliasing for nested ChainedVector and SentinelArray storage, with regression tests for identity, mutation isolation, missing and sentinel values, heterogeneous chunks, nesting, and inference.

Co-authored by Codex

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 95.64%. Comparing base (fa840f9) to head (0e73ea6).
⚠️ Report is 18 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #96      +/-   ##
==========================================
+ Coverage   94.82%   95.64%   +0.82%     
==========================================
  Files           5        5              
  Lines        1043     1056      +13     
==========================================
+ Hits          989     1010      +21     
+ Misses         54       46       -8     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[JoaoAparicio]

One observation: The typical Julia behavior is that vcat copies its inputs:

x = [1,2,3]
vcat(x) === x  # false

whereas here

x = ChainedVector([[1,2,3],[4,5]])
vcat(x) === x  # true

[ararslan]

I'd suggest folding this into the same method, either with a fast path like

Suggested change

	Base.vcat(A::ChainedVector{T, AT}) where {T, AT <: AbstractVector{T}} = A
	function Base.vcat(A::ChainedVector{T, AT}, arrays::ChainedVector{T, AT}...) where {T, AT <: AbstractVector{T}}
	function Base.vcat(A::ChainedVector{T, AT}, arrays::ChainedVector{T, AT}...) where {T, AT <: AbstractVector{T}}
	isempty(arrays) && return Base.unaliascopy(A)

or by fixing the source of the error directly by passing init=0 to sum a few lines down. (The latter would require bumping the minimum Julia compat to at least 1.6, but IMO that's fine.)