Batch node creation to avoid oversized Bolt transactions

[berrazuriz1]

Summary

Mirror the edge-batching pattern on create_nodes: chunk nodeList in Python (batch_size=5000) and send one Bolt transaction per chunk, with progress logging.

Motivation

In production we observed the Neo4j driver hitting:

neo4j.io: <CONNECTION> error: Failed to read from defunct connection IPv4Address(('p-0b67...
neo4j.pool: Unable to retrieve routing information

while a graph-creation job was writing nodes for a large repo. Cause: create_nodes passes the entire nodeList in a single Bolt transaction. For large repos this is a multi-MB UNWIND payload that the server takes long enough to process that AuraDB / network middleboxes close the underlying connection — the next driver call surfaces as defunct connection and the routing table refresh fails.

create_edges already chunks server-side via apoc.periodic.iterate, but without a Python-level batch loop the wire payload is unbounded the same way. This PR fixes only create_nodes (the smaller patch that addresses the symptom we observed); the same treatment for create_edges is a sensible follow-up.

Behavior

• Before: one Bolt transaction with len(nodeList) items.
• After: ceil(len(nodeList) / 5000) transactions, each carrying ≤ 5000 nodes. The inner apoc.periodic.iterate batch size stays at 100 (unchanged), so server-side processing semantics are identical.
• Adds two log lines: a one-line summary (Creating N nodes in batches of M) and a per-batch progress line, matching the style already used in create_edges.

Note on prod rollout

The LSP server pins blarify to feat/frontend-prompt-templates in its pyproject.toml. This PR lands on dev, which has a different module layout (blarify/db_managers/neo4j_manager.py) than the pinned branch (blarify/repositories/graph_db_manager/neo4j_manager.py). To get this fix into prod we'll need either:

• merge dev → feat/frontend-prompt-templates, or
• a sibling commit on feat/frontend-prompt-templates applying the same idea to its newer Neo4jManager.create_nodes (which sends the whole nodeList in one execute_write and only chunks via APOC).

Test plan

• poetry run ruff check blarify/db_managers/neo4j_manager.py — clean (verified locally).
• poetry run pyright blarify/db_managers/neo4j_manager.py — no new errors in this file (8 pre-existing errors in __init__ are out of scope).
• Manual: run a create_graph against a large repo and watch for Creating N nodes in batches of 5000 / per-batch lines in logs; confirm no defunct connection / routing errors mid-write.