Issue handling trailing empty segments about nifly HOT 7 CLOSED

@BadDogSkyrim As you can see based on the following commits, that part of the code is a bit prone to errors and regression.
eae5755
bb1f7a6

I don't feel comfortable touching that function with a 10-foot pole without doing extensive tests with different NIF files and combinations of segments. Unfortunately nifly doesn't have much FO4 or any segmentation tests currently.

If you have the time and interest, do you think you could possibly do some testing with different ways of filling segments?

• Shape without segments (or just one segment)
• Shape with empty segments before used segments
• Shape with used segments and empty trailing segments
• Shape with empty segments inbetween used segments (gaps)
• Resaving and editing various vanilla FO4 NIFs with different types of segmentation
• more...

from nifly.

Sure, I can do. I've got a set of regression tests in place already--cuz I'm too much of a klutz to do this stuff without a safety net--and I can make sure your scenarios are covered. I know heads do your #4 (empty segments between used segments) bc heads are seg 1 and necks are seg 3 in the head nifs (that's 0-based).

from nifly.

I ran full regression tests on that fix I posted, covering all your scenarios, and they all check out. So I'm as confident in it as I can be.

I can share the tests, but they call in through my DLL wrapper layer so they may not be that useful to you.

from nifly.

@BadDogSkyrim Thanks for testing it thoroughly. Feel free to make a PR anytime, I will do some of my own testing then as well.

from nifly.

@BadDogSkyrim @sts1skj Fixed this with #26

from nifly.

Cool, thx. I had pulled in the latest changes preparatory to a PR, but I'll pull down these changes instead and run them through my test suite.

from nifly.

The new updates check out on my end.

from nifly.