Skip to content

RAGFlow is an agent-native database: a Retrieval-Augmented Generation (RAG) engine that grew into context infrastructure.

1.3 KB
Raw Blame
import { formatSecondsToHumanReadable } from '../date';

describe('formatSecondsToHumanReadable', () => {
  test('formats sub-minute durations in seconds', () => {
    expect(formatSecondsToHumanReadable(0)).toBe('0s');
    expect(formatSecondsToHumanReadable(30)).toBe('30s');
  });

  test('strips trailing zeros from fractional seconds', () => {
    expect(formatSecondsToHumanReadable(1.5)).toBe('1.5s');
  });

  test('does not emit a trailing space for whole minutes or hours', () => {
    expect(formatSecondsToHumanReadable(60)).toBe('1m');
    expect(formatSecondsToHumanReadable(3600)).toBe('1h');
    expect(formatSecondsToHumanReadable(7200)).toBe('2h');
  });

  test('does not emit a trailing space when seconds are zero', () => {
    expect(formatSecondsToHumanReadable(3660)).toBe('1h 1m');
    expect(formatSecondsToHumanReadable(5400)).toBe('1h 30m');
  });

  test('joins non-zero hour/minute/second parts with a single space', () => {
    expect(formatSecondsToHumanReadable(90)).toBe('1m 30s');
    expect(formatSecondsToHumanReadable(125)).toBe('2m 5s');
    expect(formatSecondsToHumanReadable(3661)).toBe('1h 1m 1s');
  });

  test('returns "0s" for negative or non-numeric input', () => {
    expect(formatSecondsToHumanReadable(-5)).toBe('0s');
    expect(formatSecondsToHumanReadable(NaN)).toBe('0s');
  });
});