jswolf19

I don't see a problem in using IL instead of source as that gives a more accurate result in my mind. I realize that the [source](https://modess.io/npath-complexity-cyclomatic-complexity-explained/) provided in the feature...

@ddur An else, implicit or otherwise, does not increase the number of paths through the code because the jump before the else is unconditional. Similarly, a goto should not affect...

I was actually comparing the branch information with the IL from a debug build. If necessary I could fall back to the IL, but it would probably be better to...

@ddur Unfortunately, the original paper describing the algorithm is behind a pay wall, so I'm just going by the definition I cited above (the number of acyclic execution paths through...

@ddur, yes, I already provided that link [above](#issuecomment-303702102). As I said, the example provided there is simplistic and happens to be 22 because all paths through the function encounter both...

@ddur No worries. I'm sorry that it seems that we got off on the wrong foot. I'm sure I could have been better with how I created the issue, as...

There's already information on branching with the BranchPoints. It would likely need info on unconditional jumps (br and br.s), and for best analysis, info on early exiting via ret, throw,...

I may not have spent enough time looking, but it looked like the offsetend in the report told you where the branch would jump to. If that's the case, one...

Yeah, that's basically what I was imagining: a depth-first search from the entry point on a directed graph until you hit an exit(count one path) or a node already on...

> The proposed change would make one of the other providers which does support NPM to be the default. Cdnjs will continue to be supported. I'm currently using npm to...