Skip to main content

Restore IP Addresses - Solution & Explanation

MediumStringBacktracking16 min readAsked at: Amazon, Microsoft, Wells Fargo +17
Practice this problem

Problem Statement

A valid IP address consists of exactly four integers separated by single dots. Each integer is between 0 and 255 (inclusive) and cannot have leading zeros.

  • For example, "0.1.2.201" and "192.168.1.1" are valid IP addresses, but "0.011.255.245", "192.168.1.312" and "192.168@1.1" are invalid IP addresses.

Given a string s containing only digits, return all possible valid IP addresses that can be formed by inserting dots into s. You are not allowed to reorder or remove any digits in s. You may return the valid IP addresses in any order.

 

Example 1:

Input: s = "25525511135"
Output: ["255.255.11.135","255.255.111.35"]

Example 2:

Input: s = "0000"
Output: ["0.0.0.0"]

Example 3:

Input: s = "101023"
Output: ["1.0.10.23","1.0.102.3","10.1.0.23","10.10.2.3","101.0.2.3"]

 

Constraints:

  • 1 <= s.length <= 20
  • s consists of digits only.

Approach Overview

Problem Overview: Given a string containing only digits, generate all possible valid IPv4 addresses by inserting three dots. Each address must contain exactly four segments, and each segment must be between 0 and 255 without leading zeros (except the number 0 itself).

Approach 1: Backtracking (O(n^3) time, O(1) extra space)

This approach builds the IP address segment by segment using backtracking. Start from the first character and try to take 1, 2, or 3 digits as the next segment. After selecting a segment, validate it: the integer must be ≤255 and must not contain leading zeros like "01". If the segment is valid, recursively continue building the next part until four segments are formed.

The recursion stops when either four segments are created or the string is fully consumed. If both conditions match simultaneously, join the segments using dots and add the result to the output. Because each segment can only be length 1–3 and there are exactly four segments, the search space is small. This method systematically explores combinations while pruning invalid branches early.

This solution heavily relies on string slicing and validation checks. Interviewers expect candidates to reason about constraints such as segment length, leading zeros, and numeric bounds. The recursion depth is at most four, so auxiliary space remains constant aside from the output list.

Approach 2: Iterative Split Enumeration (O(n^3) time, O(1) extra space)

An alternative approach avoids recursion and instead enumerates all possible ways to place three dots using nested loops. Use three indices i, j, and k to represent split points in the string. These indices define four segments: s[0:i], s[i:j], s[j:k], and s[k:n]. Each segment must be length 1–3, so the loops only explore valid boundaries.

For every split combination, validate the four segments individually. Convert each substring to an integer and check the constraints: value ≤255 and no leading zeros unless the segment is exactly "0". If all segments pass validation, combine them into a dotted string and add it to the result list.

This method performs explicit enumeration instead of recursion, which some developers find easier to reason about during debugging. It still relies on careful string manipulation and boundary checks.

Recommended for interviews: The backtracking solution is the expected answer. It demonstrates control over recursive state exploration and pruning invalid candidates early. The iterative split method also works and shows strong reasoning about segment boundaries, but backtracking better reflects the pattern interviewers associate with constrained combinatorial generation problems.

Approach 1: Backtracking Approach

This approach utilizes backtracking to explore all possible placements of the three dots in the string. We attempt to generate valid IP segments recursively, ensuring each generated segment maintains the restrictions of a valid IP address. By backtracking, we can explore each possible path of placing dots and ensure we cover all potential valid configurations.

This Python function defines a helper function backtrack, which takes the current starting index and a list of constructed IP parts. We recursively attempt to build valid segments, ensuring they adhere to IP rules (no leading zeros, within [0, 255] range), and append valid results to the result list. We explore segments of lengths 1 to 3, backtracking appropriately.

Code

Python

Java

Complexity

Time Complexity: O(2^n), due to the exponential number of ways to split the string, though practical time for small strings like ours is quite feasible.
Space Complexity: O(n), for the recursion stack and to store the IP parts temporarily.

Try this approach in the editor →

Approach 2: Iterative Approach

In this approach, we iterate through potential splits for the three dots required to form four IP sections. For each potential set of positions, we validate and construct the IP address. Although typically not as elegant as the recursive backtracking solution, it's systematic and allows us to explore and validate possible IP addresses through combinatorial checks.

This C++ solution uses a three-level nested loop to estimate possible positions of dots for dividing the string into four parts. For each configuration, it calls a helper function isValid to check if each part is a valid IP section (not exceeding 255 and no leading zero unless the part is zero itself).

Code

C++

JavaScript

Complexity

Time Complexity: O(n^3), given the triple nested loop for positions.
Space Complexity: O(1), ignoring the output list space.

Try this approach in the editor →

Approach 3: DFS

We define a function dfs(i), which represents the list of IP addresses that can be formed starting from the ith position of string s.

The execution steps of function dfs(i) are as follows:

If i is greater than or equal to the length of string s, it means that we have completed the splicing of four segments of the IP address. At this point, we need to check whether it meets the requirements of the four segments of the IP address. If it does, add the current IP to the answer.

If i is less than the length of string s, it means that we still need to splice a segment of the IP address. At this point, we need to determine the value of this segment of the IP address. If the value is greater than 255, or the current position i is 0 and the value of several positions after i is greater than 0, it means that it does not meet the requirements, so we return directly. Otherwise, add it to the IP address list, and continue to search for the next segment of the IP address.

The time complexity is O(n times 3^4), and the space complexity is O(n). Here, n is the length of string s.

Code

Python

Java

C++

Go

TypeScript

C#

Try this approach in the editor →

Complexity Comparison

ApproachComplexity
Backtracking Approach

Time Complexity: O(2^n), due to the exponential number of ways to split the string, though practical time for small strings like ours is quite feasible.
Space Complexity: O(n), for the recursion stack and to store the IP parts temporarily.

Iterative Approach

Time Complexity: O(n^3), given the triple nested loop for positions.
Space Complexity: O(1), ignoring the output list space.

DFS

Detailed Complexity Analysis

ApproachTimeSpaceWhen to Use
BacktrackingO(n^3)O(1) auxiliary (depth ≤ 4)Best general solution. Cleanly explores segment combinations with early pruning.
Iterative Split EnumerationO(n^3)O(1)When recursion is avoided and you want explicit control over split indices.

Video Solution

Restore IP Addresses - Leetcode 93 - PythonNeetCode61,330 views views

Watch 9 more video solutions →

Frequently Asked Questions

Is Restore IP Addresses easy or hard?
Restore IP Addresses is classified as a Medium difficulty problem. The challenge comes from correctly handling constraints such as leading zeros, segment bounds (0–255), and ensuring exactly four segments while exploring combinations efficiently.
Restore IP Addresses Python/Java solution
Python and Java implementations usually follow the backtracking pattern. The function recursively chooses segment lengths from 1 to 3, validates each substring, and appends valid segments to a list. Once four segments are formed and the string is consumed, they are joined with dots to produce a valid IP address.
How to solve Restore IP Addresses in O(n)?
A true O(n) solution is not practical because the algorithm must check multiple combinations of segment boundaries. The closest practical complexity is O(n^3), achieved using backtracking or enumerating split indices. The constraints on segment length keep the actual runtime small in practice.
What is the best approach for Restore IP Addresses?
Backtracking is the most common and interview-friendly approach. It builds the IP address one segment at a time and prunes invalid segments early using constraints like value ≤255 and no leading zeros. Because an IP address has exactly four segments with maximum length three, the search space remains small and manageable.
Is Restore IP Addresses asked at Google/Amazon/Meta?
Restore IP Addresses is a common medium-level backtracking problem and appears frequently in interview preparation lists for companies like Amazon, Google, and Meta. It tests string manipulation, constraint validation, and recursive search techniques.
What data structure is used in Restore IP Addresses?
The core structure is a list or array used to store the current segments while building the IP address. The algorithm also relies on recursion (backtracking) or loop indices to explore possible splits, along with basic string slicing for segment validation.
What is the time complexity of Restore IP Addresses?
The typical complexity is O(n^3). Each segment can have length 1–3 and there are four segments, so the algorithm explores combinations of split positions while validating segments. The recursion depth never exceeds four, so the auxiliary space is O(1) excluding the output list.

Ready to solve this problem?

Practice Restore IP Addresses with our built-in code editor and test cases.

Practice on FleetCode