Add ERst::dataSize()

[kriben]

ERst reports how many times a name occurs in a report step through occurrence_count(), but not how many elements those occurrences hold. A caller building a summary of the arrays of a restart file, for instance one that treats the global grid and the LGRs as a single result, therefore had to walk listOfRstArrays() for every grid and add the sizes up itself.

Add dataSize(), which returns the total number of elements stored under a name in a report step, counting every occurrence. An unknown name gives zero and a non existing report step throws, as in occurrence_count().

[bska]

Add dataSize(), which returns the total number of elements stored under a name in a report step, counting every occurrence.

That's a very weird thing to compute. Why do you need it? In a run with LGRs, the total data size per array is just sum of the number of active cells per grid. That's a constant.

[kriben]

Add dataSize(), which returns the total number of elements stored under a name in a report step, counting every occurrence.

That's a very weird thing to compute. Why do you need it? In a run with LGRs, the total data size per array is just sum of the number of active cells per grid. That's a constant.

It is not the always sum of active cells, I think. As a counter-example: in the LGR_TESTMOD file FIPGAS/FIPOIL/FIPWAT are written for the global grid only: 30 values when the reservoir has 350 active cells.

In ResInsight we are classifying the arrays as a "cell result for the porosity model" if they are valueCount % reservoirActiveCellCount == 0. That is our way of excluding well and drilling related ones (e.g. IWEL and XCON) which don't exist on all time steps. So we need to know the sizes from the file for that. But it does not have to be ERst that adds them up: we can add a small method in ResInsight to do it instead.