Problem

The current load_bulk() method calls add_child() or add_root() in a loop, which means each node insertion triggers an individual database save operation. For large tree structures (thousands of nodes), this becomes a significant performance bottleneck.

Background: I'm using django-treebeard in a production enterprise application with high tree write throughput. We frequently need to load large tree structures (10,000-100,000+ nodes) from external sources, and the current load_bulk() implementation was unusable for these use cases. I've implemented a custom mixin using treebeard's utility methods to batch the inserts, which reduced our load time by 100x+. I'd love to contribute an implementation back to the community.

Current Behavior

# In load_bulk()
if parent:
    node_obj = parent.add_child(**node_data) 
else:
    node_obj = cls.add_root(**node_data)  

Each iteration performs a separate database round trip, making it infeasible to load large trees efficiently.

Proposed Solution

Add a true bulk loading method that:

1. Collects all node data in memory (or in batches)
2. Calculates tree structure metadata (path/depth/etc.) upfront
3. Uses Django's bulk_create() for efficient batch inserts

This could be:

• A new method like bulk_create_tree() or load_bulk_optimized()
• A parameter on existing load_bulk(): load_bulk(..., use_bulk_insert=True)

Questions

• Would the maintainers be interested in this feature?
• Any preference on API design (new method vs. parameter)?
• I'm happy to implement this if it aligns with the project's direction