ReviseAlgo Logo

Iterators & Generators

Infinite Iterators

Creating and managing infinite sequences — generators and itertools functions that produce values forever, with strategies for safe consumption.

Interview: Tests understanding of lazy evaluation and how to safely work with unbounded data streams.

Last Updated: June 12, 2026 • 8 min read

Infinite iterators produce values forever without exhausting. They're useful for ID generation, cycling through options, streaming data, and mathematical sequences. The key challenge is consuming them safely — always use islice(), takewhile(), or explicit break conditions.

itertools Infinite Iterators

  • count(start=0, step=1) — counts forever: start, start+step, start+2*step, ...
  • cycle(iterable) — repeats the iterable endlessly
  • repeat(object, times=None) — yields the same object forever (or N times)

Custom Infinite Generators

  • Any generator with while True is potentially infinite
  • Prime number generators, random data, timestamp streams
  • Always pair with a termination mechanism: islice, takewhile, or break

Safe Consumption Strategies

  • islice(gen, n) — take exactly n items
  • takewhile(pred, gen) — take items while condition is true
  • zip(gen, finite_iterable) — stops when the finite iterable ends
  • Explicit counter or break condition in a for loop

Warning

Never call list() on an infinite iterator — it will run forever and crash your program with a MemoryError. Always use a finite consumption pattern.

Use Cases

ID generation — unique identifiers for records, sessions, or requests

Streaming data — processing real-time feeds without buffering

Round-robin scheduling — cycling through workers, servers, or resources

Mathematical sequences — primes, Fibonacci, powers for algorithms

Testing — generating infinite test data for fuzzing or load testing

Common Mistakes

Calling list() on infinite iterator — runs forever, crashes with MemoryError

Forgetting to use islice/takewhile — the for loop never terminates

Using zip with an infinite iterator as the only argument — never stops

Not handling StopIteration in custom generators — return instead of raising

Building up state in infinite generators — memory grows without bound