Solution 1: Print the inorder and preoder traversals of both the trees and see if the traversals of
one tree are substrings in other tree's traversals.
However, this solutions requires O(m+n) space where m and n are nodes in respective trees.
Solution 2: We can check the trees node by node also if there is some logic for nodes that do not match.
boolean isSubtree (TreeNode t1, TreeNode t2)
if (t2 == null) // An empty subtree can always be found in any tree
if (t1 == null) // If no more tree is left to search, return false
// If there is a match, check the left and right nodes to match with
// left and right subtrees
if (t1.cdata == t2.cdata)
return isSubtree(t1.left, t2.left) && isSubtree(t1.right, t2.right);
// If there is no match, check left and right subtrees for a
// match with current tree
return isSubtree(t1.left, t2) || isSubtree(t1.right, t2);
Note that the above is an O(mn) implementation where m and n are tree sizes.
An O(m+n) implementation can be done if extra space is used to store inorder and preorder traversals of each tree.
If inorder and preorder traversal of the shorter tree are substrings in larger tree, then it is fully contained in larger tree.
Got a thought to share or found a bug in the code? We'd love to hear from you: