Support broadcasting over multiple DenseAxisArrays

[odow]

Closes #2267

Here's the proof:

julia> using JuMP

julia> S = [:a, :b]
2-element Array{Symbol,1}:
 :a
 :b

julia> model = Model()
A JuMP Model
Feasibility problem with:
Variables: 0
Model mode: AUTOMATIC
CachingOptimizer state: NO_OPTIMIZER
Solver name: No optimizer attached.

julia> @variable(model, x[S], start = 0.0)
1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
    Dimension 1, [:a, :b]
And data, a 2-element Array{VariableRef,1}:
 x[a]
 x[b]

julia> start_value.(x)
1-dimensional DenseAxisArray{Float64,1,...} with index sets:
    Dimension 1, [:a, :b]
And data, a 2-element Array{Float64,1}:
 0.0
 0.0

julia> set_start_value.(x, start_value.(x) .+ 1);

julia> start_value.(x)
1-dimensional DenseAxisArray{Float64,1,...} with index sets:
    Dimension 1, [:a, :b]
And data, a 2-element Array{Float64,1}:
 1.0
 1.0

[odow]

It looks like the error check was added a long time ago, and we never revisited the assumption: #1397

[codecov]

Codecov Report

Merging #2533 (1b56b46) into master (54c3c2c) will increase coverage by 0.18%.
The diff coverage is 94.73%.

@@            Coverage Diff             @@
##           master    #2533      +/-   ##
==========================================
+ Coverage   92.94%   93.13%   +0.18%     
==========================================
  Files          44       44              
  Lines        5457     5476      +19     
==========================================
+ Hits         5072     5100      +28     
+ Misses        385      376       -9     

Impacted Files	Coverage Δ
src/Containers/DenseAxisArray.jl	88.33% <94.73%> (+1.18%)	⬆️
src/parse_nlp.jl	95.50% <0.00%> (+4.29%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 54c3c2c...1b56b46. Read the comment docs.