Trie vs Hash Map for Prefix Matching in DSA Rust - Complexity Comparison

Time Complexity: Trie vs Hash Map for Prefix Matching

O(m) for Trie, O(n * m) for Hash Map

Understanding Time Complexity

We want to understand how fast prefix matching works using two common data structures: Trie and Hash Map.

Which one handles searching prefixes more efficiently as the data grows?

Scenario Under Consideration

Analyze the time complexity of the following Rust code snippets for prefix matching.

// Trie prefix search
fn trie_search(root: &TrieNode, prefix: &str) -> bool {
    let mut node = root;
    for ch in prefix.chars() {
        if let Some(next) = node.children.get(&ch) {
            node = next;
        } else {
            return false;
        }
    }
    true
}

// Hash Map prefix search
fn hashmap_search(words: &Vec, prefix: &str) -> bool {
    for word in words {
        if word.starts_with(prefix) {
            return true;
        }
    }
    false
}
    

The first function searches a prefix in a Trie by following nodes. The second checks each word in a list using a Hash Map-like vector.

Identify Repeating Operations

Look at what repeats in each method.

• Trie search primary operation: Loop over each character in the prefix.
• Trie search how many times: Once per character in prefix (length m).
• Hash Map search primary operation: Loop over all words, then check prefix match.
• Hash Map search how many times: Up to n words, each prefix check up to m characters.

How Execution Grows With Input

How does work increase when input grows?

Input Size (n = words, m = prefix length)	Approx. Operations
n=10, m=3	Trie: ~3 steps; Hash Map: up to 10*3=30 steps
n=100, m=5	Trie: ~5 steps; Hash Map: up to 100*5=500 steps
n=1000, m=10	Trie: ~10 steps; Hash Map: up to 1000*10=10,000 steps

Pattern observation: Trie work grows only with prefix length, while Hash Map work grows with number of words times prefix length.

Final Time Complexity

Time Complexity: O(m) for Trie, O(n * m) for Hash Map

Trie searches only the prefix characters once, while Hash Map checks every word, making Trie faster for large data.

Common Mistake

[X] Wrong: "Hash Map prefix search is always faster because hash lookups are quick."

[OK] Correct: Hash Map here means checking all words, which takes longer as words grow. Trie directly follows prefix characters, so it scales better.

Interview Connect

Understanding these differences helps you choose the right tool for prefix search tasks, showing your grasp of efficient data handling.

Self-Check

What if we stored all prefixes in a Hash Map instead of full words? How would that affect the time complexity?

Practice

1. Which data structure is generally more efficient for prefix matching in a large set of strings?

easy

A. Linked List, because it allows easy insertion and deletion

B. Hash Map, because it provides constant time lookup for any key

C. Array, because it stores elements in contiguous memory

D. Trie, because it stores common prefixes once and allows fast prefix queries

Solution

1. Step 1: Understand prefix matching requirements

   Prefix matching needs to find all strings starting with a given prefix efficiently.

2. Step 2: Compare Trie and Hash Map for prefix queries

   Trie stores common prefixes once and allows traversal by characters, making prefix queries fast. Hash Map is good for exact key lookup but not for prefix searches.

3. Final Answer:

   Trie, because it stores common prefixes once and allows fast prefix queries -> Option D

4. Quick Check:

   Prefix matching = Trie [OK]

Hint: Prefix queries need shared paths, so Trie is best [OK]

Common Mistakes:

• Thinking Hash Map supports prefix queries efficiently
• Confusing exact lookup with prefix search
• Assuming arrays or lists are good for prefix matching

2. Which Rust code snippet correctly declares a HashMap to store strings as keys and integers as values?

easy

A. let map = HashMap::new();

B. let map: HashMap = HashMap::new();

C. let map: std::collections::HashMap = std::collections::HashMap::new();

D. let map = HashMap();

Solution

1. Step 1: Check correct HashMap declaration syntax in Rust

   Rust requires specifying types inside angle brackets and calling HashMap::new() without parameters.

2. Step 2: Validate each option

   let map: std::collections::HashMap = std::collections::HashMap::new(); correctly declares a HashMap with String keys and i32 values using full path and new().

3. Final Answer:

   let map: std::collections::HashMap = std::collections::HashMap::new(); -> Option C

4. Quick Check:

   Correct Rust HashMap syntax = let map: std::collections::HashMap = std::collections::HashMap::new(); [OK]

Hint: Use HashMap::new() with type annotations for keys and values [OK]

Common Mistakes:

• Using parentheses instead of ::new()
• Swapping key and value types
• Omitting type annotations when needed

3. Given the following Rust code using a Trie for prefix matching, what will be the output?

struct TrieNode {
    children: std::collections::HashMap,
    is_end: bool,
}

impl TrieNode {
    fn new() -> Self {
        TrieNode { children: std::collections::HashMap::new(), is_end: false }
    }
}

fn main() {
    let mut root = TrieNode::new();
    // Insert "cat"
    let mut node = &mut root;
    for c in "cat".chars() {
        node = node.children.entry(c).or_insert(TrieNode::new());
    }
    node.is_end = true;

    // Check prefix "ca"
    let mut node = &root;
    let mut found = true;
    for c in "ca".chars() {
        if let Some(next) = node.children.get(&c) {
            node = next;
        } else {
            found = false;
            break;
        }
    }
    println!("{}", found);
}

medium

A. false

B. Compilation error

C. true

D. Runtime panic

Solution

1. Step 1: Analyze TrieNode struct definition

   TrieNode { children: HashMap, ... } creates a recursive type without indirection (e.g., Box<TrieNode>). Rust cannot compute the size of TrieNode, resulting in a compilation error.

2. Step 2: Confirm error prevents execution

   Code fails to compile, so no output or runtime behavior occurs.

3. Final Answer:

   Compilation error -> Option B

4. Quick Check:

   Recursive TrieNode without Box = compile error [OK]

Hint: Rust recursive structs require Box or Rc for children [OK]

Common Mistakes:

• Assuming prefix must be a complete word
• Confusing is_end with prefix existence
• Ignoring Rust recursive type size requirements

4. The following Rust code attempts to use a HashMap for prefix matching but has a bug. What is the bug?

use std::collections::HashMap;

fn main() {
    let mut map = HashMap::new();
    map.insert("apple", 1);
    map.insert("app", 2);

    let prefix = "ap";
    let mut found = false;
    for key in map.keys() {
        if key.starts_with(prefix) {
            found = true;
            break;
        }
    }
    println!("{}", found);
}

medium

A. Inefficient prefix search due to linear scan over keys

B. Using &str keys without lifetime causes compilation error

C. HashMap keys are unordered, so starts_with check is invalid

D. No bug; code works correctly for prefix matching

Solution

1. Step 1: Analyze prefix matching approach

   The code scans all keys linearly and checks if they start with the prefix, which works but is inefficient for large data.

2. Step 2: Identify the bug type

   There is no syntax or runtime error, but the approach is inefficient compared to Trie which supports fast prefix queries.

3. Final Answer:

   Inefficient prefix search due to linear scan over keys -> Option A

4. Quick Check:

   Linear scan for prefix = inefficient [OK]

Hint: HashMap needs full key; prefix search requires scanning all keys [OK]

Common Mistakes:

• Thinking code has syntax or runtime errors
• Assuming HashMap keys are ordered
• Ignoring performance implications

5. You need to implement a system in Rust that supports fast insertion and prefix search for millions of words. Which approach is best and why?

// Option A: Use a Trie with HashMap children
// Option B: Use a HashMap storing all words as keys
// Option C: Use a sorted Vec and binary search for prefix
// Option D: Use a LinkedList of words and scan linearly

hard

A. Trie, because it shares prefixes and allows O(k) prefix search where k is prefix length

B. HashMap, because it provides O(1) exact lookup but slow prefix search

C. Sorted Vec, because binary search helps but prefix search needs extra logic

D. LinkedList, because insertion is fast but prefix search is very slow

Solution

1. Step 1: Evaluate insertion and prefix search efficiency

   Trie allows insertion and prefix search in O(k) time, where k is prefix length, by sharing common prefixes.

2. Step 2: Compare other data structures

   HashMap is fast for exact keys but slow for prefix search. Sorted Vec requires complex prefix range search. LinkedList is slow for prefix search.

3. Final Answer:

   Trie, because it shares prefixes and allows O(k) prefix search where k is prefix length -> Option A

4. Quick Check:

   Fast prefix search = Trie [OK]

Hint: Prefix search speed depends on prefix length, Trie is best [OK]

Common Mistakes:

• Choosing HashMap for prefix search
• Ignoring complexity of prefix range search in sorted Vec
• Underestimating LinkedList search cost