Problem Statement:

Given a list of words representing search or action logs, your task is to return the top k most frequent words using a Heap structure.

If multiple words have the same frequency, return them in lexicographical order (dictionary order).

You must solve this using a Heap (e.g., PriorityQueue in Java). Sorting all words directly is not allowed due to performance constraints.

Input:

• words: A single string that represents a list of lowercase words (1 ≤ words.length ≤ 10⁵)
• k: An integer (1 ≤ k ≤ total unique words)
• Format: [word1, word2, ..., wordN], k = X

Output:

A single string representing a list of the top k frequent words in required order:

[wordX, wordY, ..., wordK]

Constraints:

• 1 ≤ total words ≤ 10⁵
• Words consist of lowercase English letters only
• Return exactly k words

Example 1:

input: [data, science, data, ai, ai, code], k = 3
output: [ai, data, code]

Explanation:
Both ai and data appear twice. ai is lexicographically smaller.
code and science appear once — code comes before science.
    

Example 2:

input: [hello, world, hello, java, java, python, hello], k = 2
output: [hello, java]

Explanation:
hello appears 3 times, java appears twice.
    

Example 3:

input: [zebra, apple, banana, apple, zebra, banana, zebra], k = 2
output: [zebra, apple]

Explanation:
zebra: 3 times, apple and banana: 2 each.
Between apple and banana, apple comes first.
    

Example 4:

input: [chat, chat, bot, bot, bot, ai, ai, data], k = 3
output: [bot, ai, chat]

Explanation:
bot appears 3 times, ai and chat appear 2 times.
Lexicographical order of tie: ai < chat.