Transform relations

[elalish]

Related #282

This adds transformation feedback to to the MeshGL output, as well as taking these transforms and original IDs as input. The ReserveIDs function allows you to make your own IDs to pass in for when your input mesh has multiple materials, for instance.

I think this is moving the API in the right direction, because the tests are getting cleaner while checking more things. I still need to add feedback about if the reference is inverted to complete the normal transforms for #282.

[codecov]

Codecov Report

Base: 91.36% // Head: 91.48% // Increases project coverage by +0.11% 🎉

Coverage data is based on head (dfe77db) compared to base (bbb1fab).
Patch coverage: 94.20% of modified lines in pull request are covered.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #315      +/-   ##
==========================================
+ Coverage   91.36%   91.48%   +0.11%     
==========================================
  Files          33       33              
  Lines        3626     3677      +51     
==========================================
+ Hits         3313     3364      +51     
  Misses        313      313              

Impacted Files	Coverage Δ
samples/src/bracelet.cpp	100.00% <ø> (ø)
src/manifold/include/manifold.h	100.00% <ø> (ø)
src/manifold/src/impl.h	70.00% <ø> (ø)
src/utilities/include/hashtable.h	81.57% <0.00%> (-2.21%)	⬇️
src/manifold/src/csg_tree.cpp	92.82% <71.42%> (-0.90%)	⬇️
src/manifold/src/constructors.cpp	96.02% <90.00%> (+0.11%)	⬆️
src/manifold/src/impl.cpp	95.08% <95.12%> (+0.65%)	⬆️
src/manifold/src/boolean_result.cpp	97.40% <100.00%> (+0.03%)	⬆️
src/manifold/src/edge_op.cpp	98.95% <100.00%> (+<0.01%)	⬆️
src/manifold/src/manifold.cpp	88.10% <100.00%> (+1.53%)	⬆️
... and 7 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[elalish]

@geoffder I updated cbind enough it would build, but there are some API updates here, which I presume is causing the CI failure.

[elalish]

Possibly related to #302: I noticed the ubuntu OMP builds have started to flake out sometimes, appearing to infinite-loop on Samples.Bracelet, but only about 50% of the time. Let's keep an eye on this. I've made the OMP builds not required for now.

[geoffder]

@elalish The CI error

 [ RUN      ] CBIND.level_set
unknown file: Failure
C++ exception with description "Error in file: /__w/manifold/manifold/test/meshIO/src/meshIO.cpp (207): 'validChannels' is false: Invalid colorChannels." thrown in the test body.
[  FAILED  ] CBIND.level_set (1849 ms)

doesn't reproduce on my machine compiled with the current master. My first naive inclination was to try using the MeshGL extracted from a Manifold constructed from the level set mesh, rather than directly exporting it, thinking that the process may fill in whatever was missing (I have not looked into it) but it doesn't appear to hit the same error for me. Not sure what would be different for me vs the CI given that the exception is coming from within Manifold code (not something external like assimp which may have version mismatch).

On the topic of parallel backends, I found that testing hung out on Samples.Knot42 when built with OMP, and Samples.Bracelet when built with TBB.

[elalish]

Strange - I wonder how it could succeed on your local machine. Seems like a simple enough problem once it repros...

[elalish]

@geoffder I ran into this in #317 and solved it there. Turns out it was because I changed the API to return a uint_32t instead of an int - apparently C++ prefers to convert int to uint during a compare rather than the other way around. :/

[geoffder]

@elalish Ahh interesting, good catch! I wonder if that behaviour changed between compiler versions then since I didn't hit the bug.