Fix crashing on an empty MeshGL with one runIndex element #1261
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #1261 +/- ##
==========================================
+ Coverage 92.18% 92.20% +0.01%
==========================================
Files 31 31
Lines 5896 5898 +2
==========================================
+ Hits 5435 5438 +3
+ Misses 461 460 -1 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
|
Thanks. Will have a look later today. |
|
An empty MeshGL with one runIndex is now valid and no longer crashes. I'm not sure about test naming though. |
|
Seems fine to me. Btw, I'm thinking if we should do more input validation, i.e., to make sure the indices are correct and don't segfault when they are incorrect. |
|
I agree. It may be better to make another PR instead of this one If possible, I'll investigate more types of exotic input later to see if any of them cause crashes or segfaults. |
|
Sounds good, I will let @elalish decide for the tests. I don't have a strong opinion about this. |
Fixes #1260
runIndex == 1to prevent out-of-bounds access; otherwise, it failed and raisedRunIndexWrongLength.runIndexAnyway, I just read the description of
runIndex. It appears that its length can be equal to the size ofrunOriginalIDorrunOriginalID + 1. Therefore, an emptyrunOriginalID(length = 0) followed by a single elementrunIndexmay actually be valid. Given this, I think it shouldn't drop the errorRunIndexWrongLengthand let the input pass. (like my workaround)Any ideas ?
CC @elalish