feat: Consolidate FEMBase.jl into JuliaFEM (Phase 1 complete)

MAJOR MILESTONE: FEMBase + FEMBasis fully consolidated, JuliaFEM loads!

Consolidated files:
- src/elements/ (3 files): elements.jl, elements_lagrange.jl, integrate.jl
- src/fields/ (1 file): fields.jl (DCTI, DVTI, DCTV, DVTV, etc.)
- src/sparse/ (1 file): sparse.jl (SparseMatrixCOO, SparseVectorCOO)
- src/assembly/ (2 files): problems.jl, assembly.jl
- src/solvers/ (1 file): solvers_base.jl
- src/analysis.jl, src/core_types.jl (Node, IP, IntegrationPoint)

Changes to JuliaFEM.jl:
- Added dependencies: Tensors, Calculus
- Removed @reexport using FEMBase (now consolidated)
- Added 20+ include statements for consolidated files
- Include order: fields → core_types → fembase_compat → sparse → elements

Compatibility layer:
- Created fembase_compat.jl: Minimal FEMBase submodule for vendor packages
- Temporarily disabled vendor-specific Mortar2D functions in solvers_modal.jl

Bug fixes:
- Changed i == 1 → isequal(i, 1) in integrate.jl (== operator overridden by fields)
- Resolved all FEMBasis. namespace references throughout codebase

Result:
- ✅ JuliaFEM loads successfully on Julia 1.12.1
- ✅ 134 exported symbols (was 171 with separate FEMBase)
- ✅ Core types accessible: Seg2, Quad4, Problem, AbstractProblem, etc.
- ⚠️  Vendor packages show FEMBase cache warnings (expected, harmless)

TODO:
- Re-enable Mortar2D functions after vendor consolidation
- Field system == operator override needs redesign (Phase 4)
- Continue Phase 2: Consolidate remaining vendor packages

src/solvers/solvers_base.jl

@@ -0,0 +1,53 @@
+# This file is a part of JuliaFEM.
+# License is MIT: see https://github.com/JuliaFEM/FEMBase.jl/blob/master/LICENSE
+
+mutable struct LinearSystem{Tv, Ti<:Integer}
+    M :: SparseMatrixCSC{Tv, Ti}
+    K :: SparseMatrixCSC{Tv, Ti}
+    Kg :: SparseMatrixCSC{Tv, Ti}
+    C1 :: SparseMatrixCSC{Tv, Ti}
+    C2 :: SparseMatrixCSC{Tv, Ti}
+    D :: SparseMatrixCSC{Tv, Ti}
+    f :: SparseVector{Tv, Ti}
+    fg :: SparseVector{Tv, Ti}
+    g :: SparseVector{Tv, Ti}
+    u :: SparseVector{Tv, Ti}
+    la :: SparseVector{Tv, Ti}
+    dim :: Int
+end 
+
+function LinearSystem(dim::Int)
+    return LinearSystem(spzeros(dim, dim), spzeros(dim, dim),
+                        spzeros(dim, dim), spzeros(dim, dim),
+                        spzeros(dim, dim), spzeros(dim, dim),
+                        spzeros(dim), spzeros(dim), spzeros(dim),
+                        spzeros(dim), spzeros(dim), dim)
+end
+
+abstract type AbstractLinearSystemSolver end
+
+function solve!(::LinearSystem, ::Solver) where Solver<:AbstractLinearSystemSolver
+    @info("This is a placeholder function for solving linear systems. To solve " *
+          "linear systems, you must define a function " *
+          "solve!(system::LinearSystem, solver::$Solver)")
+end
+
+function can_solve(::LinearSystem, ::Solver) where Solver<:AbstractLinearSystemSolver
+    return (true, "OK")
+end
+
+function solve!(ls::LinearSystem, solvers::Vector{S}) where S<:AbstractLinearSystemSolver
+    for solver in solvers
+        Solver = typeof(solver)
+        cansolve, msg = can_solve(ls, solver)
+        if !cansolve
+            @info("Solver $Solver cannot solve linear system: $msg")
+            continue
+        end
+        timeit("solve linear system using solver $Solver") do
+            solve!(ls, solver)
+        end
+        return
+    end
+    error("Failed to solve linear system.")
+end