print_tree

This package helps you to print your tree structure in a pretty format.

Install

pip install print-tree2

I wish to use print_tree as the name but the package is already on pypi though it's not working. To verify your installation, try:

git clone https://github.com/liwt31/print_tree.git
cd print_tree/
pytest

Features

There is already a pretty print tree lib on GitHub pptree, why reinvent another wheel? The most important reason is:

• This package provides an inherit interface that does not require you to modify any of your code, while pptree by design requires your node class to have certain methods or properties.

Besides, this package:

• Does not change the order of your children (pptree does).
• Uses Unicode to achieve cross-platform -- pptree fails on Windows because of wider space on terminals.
• Supports colored terminal strings (Only for Linux currently) and full-width characters.

Still, pptree is a wonderful package. The implementation is really beautiful and I borrowed some ideas from the author.

Documentation

The print*.py files in the example directory provides several examples on how to use the package. And below can be regard as an explanation of these files. The files in example directory enabled coloring to demonstrate the ability of this package to print tree with colors. Because the effect can not be seen in this MD file, corresponding codes are ommited. Suppose we have the Node class:

class Node(object):

    def __init__(self, value, parent):
        self.value = value
        self.children = []
        if parent is not None:
            parent.children.append(self)

As an example, let's construct a tree as follows:

data_structure = Node('Data Stucture', None)

vector = Node('Vector', data_structure)
list_ = Node('List', data_structure)
tree = Node('Tree', data_structure)
graph = Node('Graph', data_structure)

dag = Node('DAG', graph)
avl = Node('AVL', tree)
splay = Node('Splay', tree)
b = Node('B', tree)
quad = Node('Quand', tree)
kd = Node('kd', tree)

To print the tree, we have to tell print_tree two things:

1. how to transverse the tree from the root node.
2. how to interpret every node as a string.

To achieve these goals, we inherent print_tree from the package then override get_children and get_node_str:

from print_tree import print_tree

class print_custom_tree(print_tree):

    def get_children(self, node):
        return node.children

    def get_node_str(self, node):
        return str(node.value)

get_children should accept a Node and return a list with element type Node or PlaceHolder (see below for more info on PlaceHolder), and get_node_str accept a Node and return a string. Then we can use print_custom_tree as if it's a function:

>>> print_custom_tree(data_structure)

             ┌Vector
             ├List
Data Stucture┤
             │    ┌AVL
             │    ├Splay
             ├Tree┼B
             │    ├Quand
             │    └kd
             └Graph─DAG

If you feel uncomfortable about the naming of the class, you can import PrintTree then use PrintTree instead.

Now let's move on to some more complex examples. In the example directory I have defined a primitive search tree with custom numbers of branch. For brevity only the __init__ function of the Node is shown here. If branch == 2 then it's a binary search tree.

class Node(object):

    def __init__(self, value, branch):
        self.values = [value]
        self.branch = branch
        self.children = [None] * branch

If we wish to emphasize on the binary structure, we can override get_children and get_node_str as follows:

class print_binary(print_tree):
    def get_children(self, node):
        l_child, r_child = node.children
        if r_child is None and l_child is None:
            return []
        else:
            r_child = r_child or PlaceHolder
            l_child = l_child or PlaceHolder
            return [r_child, l_child]

    def get_node_str(self, node):
        return str(node.values[0])

In this case it is possible that the return list of get_children contains PlaceHolder, which can be directly imported by from print_tree import PlaceHolder. If PlaceHolder is in the return list, print_tree will take it as a placeholder: nothing will be shown, but it takes blank space:

# Tree (bst) already initialized
>>> print_binary(bst.root)

        ┌19┐
        │  │  ┌18
        │  └17┤
        │     │  ┌16
        │     └15┘
        │       
        │       
     ┌14┤
     │  └13
  ┌12┘
  │ 
  │ 
11┤
  │     ┌10
  │   ┌9┤
  │   │ └8
  │ ┌7┤
  │ │ │   ┌6
  │ │ │ ┌5┤
  │ │ │ │ │ ┌4
  │ │ │ │ └3┘
  │ │ │ │   
  │ │ └2┤
  │ │   └1
  └0┘

Because the tree is randomly generated, the result is probably different from what you saw when you test your installation. However, in both cases, you can read the inorder transverse of the tree from bottom to top as list(range(20)) (0 to 19). The effect of PlaceHolder becomes prominent after we delete them (also note on the benefit of inheritance):

class print_binary_without_placeholder(print_binary):
    def get_children(self, node):
        l_child, r_child = node.children
        children = []
        if r_child is not None:
            children.append(r_child)
        if l_child is not None:
            children.append(l_child)
        return children

# initialize the tree(bst)
...

print_binary_without_placeholder(bst.root)

              ┌18
        ┌19─17┤
        │     └15─16
  ┌12─14┤
  │     └13
11┤
  │     ┌10
  │   ┌9┤
  │   │ └8
  └0─7┤
      │   ┌6
      │ ┌5┤
      │ │ └3─4
      └2┤
        └1