Better error message when gradient is required but not available

[LilithHafner]

When running an optimizer that requires a gradient on a problem without a specified gradient, the error message is not as helpful as it could be:

julia> using Optimization

julia> f(u,p) = u[1]^2+p[1]
f (generic function with 1 method)

julia> optf = OptimizationFunction(f)
(::OptimizationFunction{true, SciMLBase.NoAD, typeof(f), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}) (generic function with 1 method)

julia> u0 = [1.0]
1-element Vector{Float64}:
 1.0

julia> p = [1.0]
1-element Vector{Float64}:
 1.0

julia> prob = OptimizationProblem(optf, u0, p)
OptimizationProblem. In-place: true
u0: 1-element Vector{Float64}:
 1.0

julia> sol = solve(prob, Optimization.LBFGS())
ERROR: MethodError: objects of type Nothing are not callable
The object of type `Nothing` exists, but no method is defined for this combination of argument types when trying to treat it as a callable object.
Stacktrace:
 [1] (::LBFGSB.L_BFGS_B)(func::Optimization.var"#16#28"{…}, grad!::Nothing, x0::Vector{…}, bounds::Matrix{…}; m::Int64, factr::Float64, pgtol::Float64, iprint::Int64, maxfun::Int64, maxiter::Int64)
   @ LBFGSB ~/.julia/packages/LBFGSB/UZibA/src/wrapper.jl:62
 [2] __solve(cache::OptimizationCache{…})
   @ Optimization ~/.julia/packages/Optimization/cfp9i/src/lbfgsb.jl:240
 [3] solve!
   @ ~/.julia/packages/SciMLBase/0ZQSg/src/solve.jl:186 [inlined]
 [4] #solve#718
   @ ~/.julia/packages/SciMLBase/0ZQSg/src/solve.jl:94 [inlined]
 [5] solve(::OptimizationProblem{true, OptimizationFunction{…}, Vector{…}, Vector{…}, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, @Kwargs{}}, ::Optimization.LBFGS)
   @ SciMLBase ~/.julia/packages/SciMLBase/0ZQSg/src/solve.jl:91
 [6] top-level scope
   @ REPL[9]:1
Some type information was truncated. Use `show(err)` to see complete types.

It would be nice to give a specific "gradient provider required" error with a hint on how to fix it.