The metaform.template generator does not choose the most complex substructure to traverse, and does not combine the keys at deeper levels, rather, chooses the last element in the lists traversed as the one to be included in the template, leaving out others:
Video: https://wiki.mindey.com/shared/shots/37935f7d674e69ed8b9a59ee6-metaform-template-generation-challenge.mp4
A fix would be to create a function, that estimates the traversal complexity (not completely traversing, but up to the level, to be able to compare elements in the lists), and choose to prioritize the more complex ones.
Another fix should be found through reviewing the process of combining the elements at the same levels within different substructures in the lists.
The
metaform.templategenerator does not choose the most complex substructure to traverse, and does not combine the keys at deeper levels, rather, chooses the last element in the lists traversed as the one to be included in the template, leaving out others:Video: https://wiki.mindey.com/shared/shots/37935f7d674e69ed8b9a59ee6-metaform-template-generation-challenge.mp4
A fix would be to create a function, that estimates the traversal complexity (not completely traversing, but up to the level, to be able to compare elements in the lists), and choose to prioritize the more complex ones.
Another fix should be found through reviewing the process of combining the elements at the same levels within different substructures in the lists.