PATH splitting and events

[camillemonchicourt]

I inserted a new PATH (in red) in l_t_troncon (using QGIS) :

The existing path (blue) was well splitted :

But one trek that was using the existing path has changed. I dont think it is because of QGIS but because of splitting and events calculation triggers. Departure and arrival are now moved were the existing path was splitted :

When editing the TREK, it is even more broken :

After correcting it :

It is OK now :

[camillemonchicourt]

Just reproduced it on another trek cutting.

[leplatrem]

Splitting a return-topology with its extremity on a loop. A simple test should be able to confirm this bug...