Traverse Nested Dict Paths Depth-First in Python

Recursively walk a nested dictionary depth-first and yield each full path from root to leaf as lists.

Medium Python 3.8+ Aug 9, 2026 Dictionaries & sets 13 views 0 copies

recursion generators nested-dicts depth-first traversal

Python code

32 lines

Python 3.8+

def depth_first_paths(node, path=None):
    if path is None:
        path = []
    
    if not isinstance(node, dict):
        yield path + [node]
        return
    
    for key, value in node.items():
        new_path = path + [key]
        if isinstance(value, dict):
            yield from depth_first_paths(value, new_path)
        else:
            yield new_path + [value]


if __name__ == "__main__":
    tree = {
        "root": {
            "left": {
                "leaf1": 1,
                "leaf2": 2
            },
            "right": {
                "leaf3": 3,
                "leaf4": 4
            }
        }
    }
    
    for path in depth_first_paths(tree):
        print(path)

Output

stdout

['root', 'left', 'leaf1', 1]
['root', 'left', 'leaf2', 2]
['root', 'right', 'leaf3', 3]
['root', 'right', 'leaf4', 4]

How it works

The function checks if the current node is a dict: if not, it yields the accumulated path plus the scalar value and returns. For dicts, it loops over items, extends the path with the key, and either recurses (yield from) into sub-dicts or emits the complete path with the leaf value. Passing path=None and creating a fresh list avoids the classic mutable default-argument bug, so each recursive call gets its own list. Using a generator means the traversal is lazy and memory-friendly for huge trees.

Common mistakes

Using a mutable default argument `path=[]` instead of `path=None`, causing paths to share state across calls
Forgetting `yield from` for nested dicts, so nested paths never get emitted
Treating lists inside the dict as leaves rather than continuing deeper

Variations

Return list of paths (all at once) instead of yielding — call `list(depth_first_paths(tree))`
Use `path.copy()` instead of `path + [key]` when accumulating paths manually

Real-world use cases

Exporting every leaf value from a nested JSON config into a flat key list for validation or migration.
Building a full map of API response fields to locate and update specific nested values.
Flattening deeply nested YAML/JSON data for logging, dashboards, or data pipelines.

Traverse Nested Dict Paths Depth-First in Python

Python code

Output

How it works

Common mistakes

Variations

Real-world use cases

More from Dictionaries & sets

Tutorials

Quizzes

Python code

Output

How it works

Common mistakes

Variations

Real-world use cases

More from Dictionaries & sets

Keep learning

Tutorials

Quizzes