fix: use .get() in get_leaf_nodes to avoid KeyError on leaf nodes

[ChiragB254]

Summary

• get_leaf_nodes() crashed with KeyError: 'nodes' when called on a tree built by the standard pipeline
• list_to_tree() calls clean_node() which deletes the nodes key entirely from leaf nodes (rather than setting it to [])
• Accessing structure['nodes'] directly on those nodes raises KeyError
• Fix: replace structure['nodes'] with structure.get('nodes') — returns None (falsy) safely when the key is absent

Changes

pageindex/utils.py — one-line change in get_leaf_nodes:

# Before
if not structure['nodes']:   # KeyError on leaf nodes

# After
if not structure.get('nodes'):   # safe: returns None if key absent

Why this is consistent

Every other nodes access in the codebase already uses .get():

• format_structure → structure.get('nodes')
• is_leaf_node → node.get('nodes')
• print_tree → node.get('nodes')

Test

from pageindex.utils import get_leaf_nodes, list_to_tree

flat = [
    {'structure': '1', 'title': 'Chapter 1', 'start_index': 1, 'end_index': 5},
    {'structure': '2', 'title': 'Chapter 2', 'start_index': 6, 'end_index': 10},
]
tree = list_to_tree(flat)
leaves = get_leaf_nodes(tree)   # previously raised KeyError, now returns both nodes
assert len(leaves) == 2

Fixes #330

[KylinMountain]

Thanks @ChiragB254, nice catch — consistent with the rest of the codebase. Merging! 🙏