Wrapper Over a Paginated Page API That Fetches the Next N Words

You are given an external API that returns content one page at a time, and each page holds a variable number of words. Write a wrapper class around this API that lets callers fetch by word count: each call asks for the next n words and receives them in order, continuing exactly where the previous call stopped.

Assume the external API looks like this (confirm the real shape with the interviewer):

def fetch_page(page: int) -> tuple[list[str], bool]:
 """Return (words, has_more) for page number `page`, counting from 0.

 `words` holds the words on that page, in order, and may be of any length.
 `has_more` is False on the last page.
 """

Implement:

class WordFetcher:
 def __init__(self, fetch_page): ...

 def fetch(self, n: int) -> list[str]:
 """Return the next n words. Return fewer only when the content runs out."""

Constraints and Clarifications

• Calls to the external API are slow compared with in-memory work, so the wrapper should call it only when it needs more words.
• The wrapper must never request the same page twice or skip a page.

Clarifying Questions

• Does the API return a list of words, or raw text that the wrapper must split? If raw text, can a word be cut in half at a page boundary?
• Can a page in the middle of the content be empty?
• What should fetch do for n = 0, or for a negative n?
• What should happen when an API call fails or times out?