Error in control variables in sub_6FCF77E0 (SkillAss.cpp)

[Araksson]

In the file SkillAss.cpp, When doing iterations by progressive skills, there is a second loop that uses "i" again as a control variable, it should be "j"

First control var

D2MOO/source/D2Game/src/SKILLS/SkillAss.cpp

Line 1041 in e430609

for (int32_t i = 0; i < sgptDataTables->nProgressiveStates; ++i)

Second control var

D2MOO/source/D2Game/src/SKILLS/SkillAss.cpp

Line 1076 in e430609

for (int32_t i = 0; i < nValue - 1; ++i)

[Necrolis]

This is technically a poor choice of naming, although not a bug due to C++ variable name shadowing rules (so the loops function correctly), see this short example:
https://godbolt.org/z/j3PPjdEaf

[Araksson]

In several places in D2Moo there are those errors of double nested 'for' with double control variable of the same name (I already reported them a while ago). In my case the compiler gives me a warning (because I have the option to show all warnings enabled) and when I run it it is actually "adding" invalid data to i within that loop and clearly the largest loop always starts again from 0-1-2 (depending on the value of the last progressive processed). And clearly it is an error because that variable is technically initialized twice.

[Necrolis]

And clearly it is an error because that variable is technically initialized twice.

its not if its declared twice at different scopes, please see the C++ block scoping rules here.
Which is why you'll see the following when /Wall is enabled:

D2MOO\source\D2Game\src\SKILLS\SkillAss.cpp(1079,42): warning C4456: declaration of 'i' hides previous local declaration

If you are experiencing bugs its likely as a result of the second declaration being missing.

I already reported them a while ago

If they were not created as tickets, please do list all the refs here, since they should be changed, as variable name shadowing is a "bad code smell".

@Lectem Alternately this ticket can be changed to a broad fix-all for warning C4456 outputs under /Wall (which IMO makes more sense).

[Araksson]

I reported the other errors similar to this one to Lecten quite a while ago, I think they fixed them.
I don't remember exactly where those errors are now to check them one by one, but they were in StatList, DRLG, ITEMS and PATH. But as I said, those codes were rewritten several times in D2Moo and the latest versions work well, so they must be corrected.

[Lectem]

We can add such warnings to clang-tidy, I wonder how many name shadowing errors would pop.