Use of uninitialized values in GenericMoltresMaterial

[arfc]

Use of uninitialized values in GenericMoltresMaterial

[katyhuff]

@gridley Perhaps it should be the length of the temperature vector initialized above?

[gridley]

Yep! Although there's no stuff in the temperature vector at that point. I'm really confused how this could have got there. Probably some accidental line deletion which didn't get caught in units tests, since this is in a code section which supports not using CHI_D, a previously missing feature in the code. If CHI_D is supplied, this shouldn't get called. Anyways... yes. I'm going to fix this now.

[gridley]

It may appear to not be initialized, but always should be by the time that for (j=0; j<xsec_names.size;++j) loop gets to it. That's because xsec_names is hard-coded, and CHI_D always comes past the first element. As long as the first element has been passed, tempLength should be defined.

I think it makes sense to write the code this way. This part of code only gets turned on if CHI_D is missing, and you have to assume that the list of temperatures given for CHI_D matches the length of what was given in previous XSEC files. The best way to remove the "possibility" (someone editing the xsec_names variable) would be assume that cross sections can be interpolated in disparate ways, e.g. cubic splines on CHI, monotone cubics on CHI_D. Then CHI_D could be given a constant value of 1 in group 1. Since that would be a bit of effort and of marginal benefit, I propose we leave the code as-is. I hope this explanation doesn't suck.

[lindsayad]

You do assignment in splineConstruct but you don't in the methods I cited. In all three cases, you might as well zero initialize tempLength upon declaration to silence the compiler warning. Otherwise you are going to create concerned users when they see compiler warnings

[katyhuff]

Closed with #76