Re-building directory tree from a list of random paths using pygtrie

trie_traversal.py

import pygtrie

# 0 denotes a directory and 1 denotes a file
path_entries = [
    ("./pcache/pcache.egg-info/dependency_links.txt", 1),
    ("./pcache/LICENSE", 1),
    ("./pcache/build/lib/pcache/__init__.py", 1),
    ("./pcache/pcache.egg-info/top_level.txt", 1),
    ("./pcache/pcache/pcache.py", 1),
    ("./pcache/pcache/__init__.py", 1),
    ("./pcache", 0),
    ("./pcache/pcache.egg-info/PKG-INFO", 1),
    ("./pcache/README.md", 1),
    ("./pcache/setup.py", 1),
    ("./pcache/.travis.yml", 1),
    ("./pcache/pcache", 0),
    ("./pcache/build/lib", 0),
    ("./pcache/.gitignore", 1),
    ("./pcache/tests.py", 1),
    ("./pcache/dist/pcache-0.0.3-py3-none-any.whl", 1),
    ("./pcache/build/lib/pcache", 0),
    ("./pcache/build/lib/pcache/pcache.py", 1),
    ("./pcache/pcache.egg-info/SOURCES.txt", 1),
    ("./pcache/pcache.egg-info", 0),
    ("./pcache/build", 0),
    ("./pcache/build/bdist.macosx-10.9-x86_64", 0),
    ("./pcache/dist", 0),
    ("./pcache/dist/pcache-0.0.3.tar.gz", 1),
]


def callback(path_conv, path, children, is_file=False):
    """node_factory for the traverse method of the trie object

    Read more about it here:
    https://pygtrie.readthedocs.io/en/latest/#pygtrie.Trie.traverse
    """
    if is_file:
        print(path_conv(path))
        return 0
    if path:
        print(path_conv(path))
    return sum(children)


if __name__ == "__main__":
    # Initialize the trie with the path separator
    trie = pygtrie.StringTrie(separator="/")
    # Add keys to the trie. The path forms the key and the entry type
    # decides whether it is a leaf or not. It means that files will
    # form the leaves whereas directories will ideally be the
    # intermediate nodes
    for key, value in path_entries:
        trie[key] = bool(value)
    # Finally, traverse the trie
    trie.traverse(callback)