pytreeds

A dependency-free, beginner-focused Python tree library suitable for Pyodide. One Tree variable holds all nodes, and navigation/traversal uses names (strings) consistently. Attach optional Python data to individual nodes.

Version: The current source release is 0.2.3. See the PyPI project page for the latest published version.

Quick start

from pytreeds import Tree

org = Tree()
org.add_node("CEO", data={"level": 1})
org.add_node("CTO", parent="CEO", data=[90, 85])
org.add_node("CFO", parent="CEO")
org.add_node("COO", parent="CEO")
org.add_node("Engineering", parent="CTO")
org.add_node("DevOps", parent="CTO")
print(org)

print(org.get_node_data("CTO"))  # [90, 85]
org.set_node_data("CTO", {"team": "Technology"})
print(org.get_node_data("CTO"))

print(list(org.dfs()))
print(list(org.bfs()))

org.move_node("CTO", parent="COO")  # The entire CTO subtree moves
org.delete_node("CFO")             # Works on leaf or subtree
print(org)

# Three layouts for the same tree
print(org.get_diagram(style="branch"))
print(org.get_diagram(style="horizontal"))
print(org.get_diagram(style="vertical"))

# Include node data or display a smaller subtree
print(org.get_diagram("CTO", max_depth=1, show_data=True))
# Data is shown beneath each name, using double quotes for strings

Omitting parent in add_node() creates a root only when the tree is empty. Every later node requires a parent. One Tree contains exactly one root.

Install

Regular Python: pip install pytreeds

Pyodide (asynchronous package loader):

import micropip
await micropip.install("pytreeds")

API summary

Build: Tree(), add_node(name, parent=None, data=None).

Edit: delete_node(), move_node(), rename_node(), copy(). Removing a node removes all its descendants; moving one moves its entire subtree.

Inspect: get_root(), get_size(), get_height(), get_depth(), get_leaf_count(), is_leaf(), contains().

Relationships: get_parent(), get_children(), get_siblings(), get_ancestors(), get_descendants(), get_leaves().

Data: get_node_data(name), set_node_data(name, data). Any Python object can be attached, including lists and dictionaries. Values are returned by reference; copy() deep-copies data, so objects incompatible with deepcopy() may cause an error when copying.

Traversal: dfs(), bfs(), and ordinary for name in tree. The dfs() method uses left-to-right preorder (visits parents before their children).

Diagrams: print(tree) for display, or get_diagram(name=None, style='branch', max_depth=None, show_data=False) to retrieve a diagram as a string.

See the illustrated terminology guide and the Detailed Reference, with a dedicated page for every method.

Local tests

python -m unittest discover -s tests -v

Releases are gated by automated unit tests and a strict Sphinx documentation build.