Updating ECLGraph::haveConnData so that 1D and 2D cases are handled c…

[vkip]

…orrectly (in these cases missing dimension arrays are not present in the data file).

[atgeirr]

Sounds like a useful and important fix, thanks! Assigning @bska to check and merge this (if OK). There are some annoying tab-vs-spaces issues (we try to go with spaces exclusively), but we can fix that.

[bska]

This will return true if any of the (I,J,K) vectors are present. We need it to return true only if all the (relevant) vectors are present.

[atgeirr]

I am a little confused now: if you are running a 1d case, it will actually be represented as a 3d case, but with a string of cells in only one direction (for example a 1 x 1 x 10 grid), correct? If so, would not the simulator still output fluxes in the I+ and J+ directions (that would be zero)?

Do you have a case that shows this, that causes the current code to fail?

[atgeirr]

Is this something we should resurrect, or just close @vkip?