Serialize and Deserialize N-ary Tree

Serialize and Deserialize N-ary Tree - Problem

Serialization is the process of converting a complex data structure into a format that can be stored or transmitted and later reconstructed perfectly. Think of it like packing a complex 3D puzzle into a flat box - you need a systematic way to flatten it and rebuild it exactly as it was!

Your challenge is to design an algorithm to serialize (convert to string) and deserialize (reconstruct from string) an N-ary tree, where each node can have any number of children (not just 2 like binary trees).

Goal: Create two functions:

serialize(root) → converts tree to string
deserialize(data) → reconstructs tree from string

Example: A tree with root 1, children [3,2,4], and node 3 having children [5,6] could be serialized as "1[3[5,6],2,4]" or "1,null,3,2,4,5,6,null,null,null,null" - the format is entirely up to you!

Key requirement: deserialize(serialize(tree)) === tree ✨

Input & Output

example_1.py — Basic N-ary Tree

$ Input: root = [1,null,3,2,4,5,6] Tree structure: 1 /|\ 3 2 4 /| 5 6

› Output: serialize(root) = "1,3,3,2,5,0,6,0,2,0,4,0" deserialize("1,3,3,2,5,0,6,0,2,0,4,0") = original tree

💡 Note: The root node 1 has 3 children. Node 3 has 2 children (5,6). Nodes 2,4,5,6 are leaves with 0 children. The preorder traversal captures the structure perfectly.

example_2.py — Single Node Tree

$ Input: root = [1] Tree structure: 1

› Output: serialize(root) = "1,0" deserialize("1,0") = Node(1) with empty children list

💡 Note: A single node tree serializes to just the value and children count (0). This is the simplest case - one node, no children.

example_3.py — Deep Linear Tree

$ Input: root represents: 1 | 2 | 3 | 4

› Output: serialize(root) = "1,1,2,1,3,1,4,0" deserialize("1,1,2,1,3,1,4,0") = original linear tree

💡 Note: Each non-leaf node has exactly 1 child, creating a linear chain. The serialization shows: 1→1 child→2→1 child→3→1 child→4→0 children.

Constraints

The number of nodes in the tree is in the range [0, 10⁴]
0 ≤ Node.val ≤ 10⁴
The height of the n-ary tree is less than or equal to 1000
No restriction on serialization format - be creative!
Must ensure deserialize(serialize(root)) == root

Visualization

Tap to expand

Asked in

G Google 35 a Amazon 28 f Facebook 22 ⊞ Microsoft 18 Apple 12

The optimal solution uses preorder traversal with children count encoding. During serialization, we visit each node and record its value followed by the number of children. During deserialization, we use the children count to know exactly how many subtrees to recursively build. This elegant approach achieves O(n) time and O(h) space complexity while producing compact serializations.

Common Approaches

✓ Brute Force (Level-by-Level with Positions)

⏱️ Time: O(n²) Space: O(n²)

This approach stores extensive metadata for each node including its level, position among siblings, and parent information. While it works, it creates very verbose serializations with redundant information.

Preorder with Markers (Optimal)

⏱️ Time: O(n) Space: O(h)

This optimal approach uses a preorder traversal (root, then children) with special markers to indicate the structure. We mark the end of each node's children list, allowing perfect reconstruction in a single pass.

Brute Force (Level-by-Level with Positions) — Algorithm Steps

For serialize: traverse tree level by level, storing each node's value with its complete position path
Include level number, sibling index, and parent reference for each node
Create a verbose string with all this metadata
For deserialize: parse all the metadata to reconstruct parent-child relationships
Build the tree by processing each level and connecting nodes to their parents

Visualization

Tap to expand

Step-by-Step Walkthrough

Level-by-Level Traversal

Visit nodes level by level, storing complete position information

Metadata Collection

For each node, collect value, level, parent path, and sibling index

Verbose Encoding

Create detailed string with all metadata for later reconstruction

Code -

solution.c — C

#include <stdio.h>
#include <stdlib.h>
#include <string.h>

struct Node {
    int val;
    int numChildren;
    struct Node** children;
};

struct Node* createNode(int val) {
    struct Node* node = (struct Node*)malloc(sizeof(struct Node));
    node->val = val;
    node->numChildren = 0;
    node->children = NULL;
    return node;
}

static char serializeResult[100000];
static int serializePos;

void serializeDFS(struct Node* node) {
    if (!node) return;
    
    char temp[50];
    
    // Add comma if not first element
    if (serializePos > 0) {
        serializeResult[serializePos++] = ',';
    }
    
    // Add node value
    sprintf(temp, "%d", node->val);
    strcpy(serializeResult + serializePos, temp);
    serializePos += strlen(temp);
    
    // Add comma and child count
    serializeResult[serializePos++] = ',';
    sprintf(temp, "%d", node->numChildren);
    strcpy(serializeResult + serializePos, temp);
    serializePos += strlen(temp);
    
    // Recursively serialize children
    for (int i = 0; i < node->numChildren; i++) {
        serializeDFS(node->children[i]);
    }
}

char* serialize(struct Node* root) {
    if (!root) {
        char* result = (char*)malloc(1);
        result[0] = '\0';
        return result;
    }
    
    serializeResult[0] = '\0';
    serializePos = 0;
    serializeDFS(root);
    serializeResult[serializePos] = '\0';
    
    char* result = (char*)malloc(serializePos + 1);
    strcpy(result, serializeResult);
    return result;
}

static char** tokens;
static int tokenCount;
static int deserializeIndex;

struct Node* buildTree() {
    if (deserializeIndex >= tokenCount) {
        return NULL;
    }
    
    int val = atoi(tokens[deserializeIndex++]);
    
    if (deserializeIndex >= tokenCount) {
        return createNode(val);
    }
    
    int childCount = atoi(tokens[deserializeIndex++]);
    
    struct Node* node = createNode(val);
    if (childCount > 0) {
        node->children = (struct Node**)malloc(childCount * sizeof(struct Node*));
        node->numChildren = childCount;
        
        for (int i = 0; i < childCount; i++) {
            struct Node* child = buildTree();
            if (child) {
                node->children[i] = child;
            }
        }
    }
    
    return node;
}

struct Node* deserialize(char* data) {
    if (!data || strlen(data) == 0) {
        return NULL;
    }
    
    static char dataCopy[100000];
    static char* tokenArray[10000];
    
    strcpy(dataCopy, data);
    tokens = tokenArray;
    tokenCount = 0;
    deserializeIndex = 0;
    
    char* token = strtok(dataCopy, ",");
    while (token != NULL && tokenCount < 10000) {
        tokens[tokenCount++] = token;
        token = strtok(NULL, ",");
    }
    
    return buildTree();
}

struct Node* buildTreeFromArray(int* arr, int size) {
    if (size == 0) return NULL;
    
    struct Node* root = createNode(arr[0]);
    if (size == 1) return root;
    
    static struct Node* queue[10000];
    int front = 0, rear = 0;
    queue[rear++] = root;
    int i = 2;
    
    while (front < rear && i < size) {
        struct Node* parent = queue[front++];
        
        static struct Node* children[1000];
        int childCount = 0;
        
        while (i < size && arr[i] != -1) {
            struct Node* child = createNode(arr[i]);
            children[childCount] = child;
            queue[rear++] = child;
            childCount++;
            i++;
        }
        
        if (childCount > 0) {
            parent->children = (struct Node**)malloc(childCount * sizeof(struct Node*));
            parent->numChildren = childCount;
            for (int k = 0; k < childCount; k++) {
                parent->children[k] = children[k];
            }
        }
        i++; // Skip null
    }
    
    return root;
}

int treesEqual(struct Node* root1, struct Node* root2) {
    if (!root1 && !root2) return 1;
    if (!root1 || !root2) return 0;
    if (root1->val != root2->val) return 0;
    if (root1->numChildren != root2->numChildren) return 0;
    
    for (int i = 0; i < root1->numChildren; i++) {
        if (!treesEqual(root1->children[i], root2->children[i])) {
            return 0;
        }
    }
    
    return 1;
}

int main() {
    char inputLine[10000];
    fgets(inputLine, sizeof(inputLine), stdin);
    
    // Remove newline
    inputLine[strcspn(inputLine, "\n")] = 0;
    
    static int arr[10000];
    int size = 0;
    
    if (strcmp(inputLine, "[]") != 0) {
        // Parse input format like "1,,3,2,4,5,6"
        char* p = inputLine;
        char* start = p;
        
        while (1) {
            if (*p == ',' || *p == '\0') {
                // Found a token
                if (p == start) {
                    // Empty token (represents null)
                    arr[size++] = -1; // Use -1 as null marker
                } else {
                    // Non-empty token, parse as number
                    char temp = *p;
                    *p = '\0';
                    if (strcmp(start, "null") == 0) {
                        arr[size++] = -1;
                    } else {
                        arr[size++] = atoi(start);
                    }
                    *p = temp;
                }
                
                if (*p == '\0') break;
                
                p++;
                start = p;
            } else {
                p++;
            }
        }
    }
    
    // Build tree
    struct Node* root = buildTreeFromArray(arr, size);
    
    // Test serialize/deserialize
    char* serialized = serialize(root);
    struct Node* deserialized = deserialize(serialized);
    
    // Check if trees are equal
    int result = treesEqual(root, deserialized);
    printf("%s\n", result ? "true" : "false");
    
    free(serialized);
    return 0;
}

Time & Space Complexity

Time Complexity

⏱️

O(n²)

Each node requires processing its full path information, leading to quadratic time in the worst case for deep trees

⚠ Quadratic Growth

Space Complexity

O(n²)

Storing complete path information for each node can require quadratic space for deep trees

⚠ Quadratic Space

89.3K Views

Medium Frequency

~25 min Avg. Time

1.8K Likes

Ln 1, Col 1

Smart Actions

💡 Explanation

AI Ready

💡 Suggestion Tab to accept Esc to dismiss

// Output will appear here after running code

Code Editor Closed

Click the red button to reopen

Input & Output

Constraints

Visualization

Related Problems

Common Approaches

Brute Force (Level-by-Level with Positions) — Algorithm Steps

Visualization

Code -

Time & Space Complexity

Select Compiler