Free Space test cases + Nodal Interpolation + Small Fix in FunctionSpace: Bug with cell function in 2D for 3D Function Space + Vertices of GroupOfElement

DDM Tests: Rewritting SystemAbstract::getSolution + FunctionSpace::getUnorderedKeys() + FunctionSpace::getAllDofs with set + FunctionSpace::getKeys(set)

Small simplifications + Preparation for Formulation Independent FS + Preparation for multiple Formulations using same FS

 -- Using PView
 -- Removing Writer for FEMSolution and ElementSolution (layer for PView)
 -- Adapting System
 -- Interpolator simplified (left until PView can handel arbitrary basis)

** Mesh: new method
 -- Can access all the coordinates of the vertices
 -- Used by Interpolator

** Simulations: Huge Cleaning
 -- Intializing by SmallFem for all simulations
 -- Merging all Projections into on code
 -- Use of FEMSolutions and ElementSolutions
 -- Rewriting ShowFunctionSpace to use only FEMSolution
 -- Convergence Tests done again (see further)

** Assembler: Removing SystemShowFunctionSpace
 -- Done in Simulation

** FunctionSpace:
 -- Changing interface in BasisLagrange
 -- Cyclic permutations reactivated in ReferenceSpace (see further)

** WARNING: Without cyclic permutation TetEdgeBasis does NOT converge !!
 -- Works in 2D
 -- Works with TetNodalBasis
 -- Fails with TetEdgeBasis
 -- ShowFunctionSpace shows weird thing for some Tets
 -- Implementation or math error ????

** Convergence Tests WITH cyclic permutation
 -- Works everywhere (line, tri, quad, tet)

Lagrange Basis uses ReferenceSpace -- WARNING Orientation still misses + REworked line basis + ReferenceSpace takes only connectivity into account: seems enough

Overcomming OpenBLAS nested threads issue -- Jacobians no longer use BLAS -- Not a real problem, since those matrices are 3 by 3

Hex and Tri ReferenceSpace: load a serialized version if available - if not build a new reference space and serialize it for next time

HexReferenceSpace seems to Work: other test passed -- ReferenceSpace and PermutationTree: need to compress tag if they do not form a continous sequence -- HexNodeBasis: reworking with new framework -- PermutationTree: start serialization

ReferenceSpace for Hex: Seems OK \o/ -- ReferenceSpace uses now connectivity check instead of parallel face -- Need check with HexBasis

Cleaner implementation of face permutation matching + Cleaner implementation of high order element in FunctionSpace

    -- Add Point Mapping from ABC space to XYZ space
    -- Handle Elements with no Face

** LineBasis:
    -- Works again

** Jacobian and GroupOfJacobian:
    -- Computes Good Jacobian from Basis ReferenceSpace

** Formulation:
    -- Addapted for good Jacobian

** TermProjection:
    -- Addapted to compute function to project at good point (see ReferenceSpace ABCtoXYZ mapping)

** Laplace, Poisson, SteadyWave:
    -- Passed with all basis

** WARNING:
    -- Tet Basis NOT Retested (but seems to work)
    -- FunctionSpace::getDof doesn't work if wrong element type (because of basis implication)
    -- SystemAbstract::fixCoef doesn't work anymore (see above)
    -- Problem with cureved Element (getTypeForMSH gives HO element for FunctionSpace::getDof)

ReferenceSpace computes Jacobians: Seems OK -- TriEdgeBasis seems to be OK -- FunctionSpaceScalar and FunctionSpaceVector adapted with new ReferenceSpace::getJacobian

  ++ Now Dofs are numbered correctly
  ++ Tri, Quad and Tet Nodal Basis looks ok
  -- Need to compute Jacobians

Basis: Interface modification

     -- Minimal number of reference space (at least for tet, tri and quad)
     -- Automatic generation of these reference space
     -- New interface
     -- Modification of TriNodeBasis accordingly

** Need to compute Dof permutations automaticaly

** All Basis are commented (exept for TriNodeBasis)

** WARNING: Memory leak somewhere

Reworking ReferenceSpace -- Trying to have minimum reference space -- find permutation and reference space for 2d and tet

Working on Dof Reordering -- New GroupOfDof Interface -- Cleaning some tabs (replaced by spaces) -- BasisLagrange Implementation of Function and Grad

make bezierBasis independant from declaration of element tag in GmshDefines.
move static methods getTag(..) & **FromTag(..) from MElement to ElementType.
completely revert to double for monomials.