Created on 2021-06-22 13:42:41 by Didier Vibert. % Done: 66

after #6343 some more cleaning is necessary:

• renaming of attribute names in hdf5 (possibly rename also the ResultStore class members)
• scanning of all members in ResultStore classes, to track missing stuff to be added in HDF5 (at least in debug mode)
• add first pass results from linemodel
• implement debug mode filtering for writing HDF5
• remove saving CPdfCandidateszResult<T> from ResultStore since it is now aggregated by inheritance in CExtremaResult<T> which is saved in ResultStore
• break TCandidateZ inheritance from COperatorResult (feasible ?)
• go deeper in python introspection, by enabling swig downcasting (casting a derived type from base class pointer) see: https://stackoverflow.com/questions/27392602/swig-downcasting-from-base-to-derived
• more classes to differentiate COperatorResults derived classes
• modify TLineModelResult(const CContinuumModelSolution& cms) constructor (and thus remove casscaded default empty ctor in the class hierarchy TLineModelResult->TExtremaResult()->TCandidateZ())
• suppression du membre m_savedModelContinuumFittingResults de la classe CExtremaResult<TExtremaResult>
  (redondant avec le templating de CExtremaResult)
• class CModelFittingResult, ctor use move semantic instead of copying the linemodelsolution
• CModelFittingResult, redondancy (eg ::LineModelSolution.Fluxs & ::FitttedRaysFlux)
• CModelFittingResult, add direct integ fluxes
• COperatorLineModelExtremaResult::m_savedModelFittingResults,m_savedModelRulesResults, m_savedModelContinuumSpectrumResults, m_savedModelSpectrumResults, m_savedModelContinuumFittingResults seems unused ?