test_that("Arvind distribution functions evaluate correctly and handle edge cases", {
  theta <- 0.5
  
  # PDF
  expect_equal(darvind(0, theta), 0.5)
  expect_true(darvind(1, theta) > 0)
  expect_equal(darvind(-1, theta), 0)
  expect_true(is.na(darvind(NA, theta)))
  
  # CDF
  expect_equal(parvind(0, theta), 0)
  expect_true(parvind(1, theta) > 0 && parvind(1, theta) < 1)
  expect_equal(parvind(1, theta, lower.tail = FALSE), 1 - parvind(1, theta))
  
  # Quantile & CDF consistency
  q95 <- qarvind(0.95, theta)
  expect_equal(parvind(q95, theta), 0.95, tolerance = 1e-4)
  expect_equal(qarvind(0, theta), 0)
  expect_equal(qarvind(1, theta), Inf)
  
  # Hazard and Survival
  h1 <- harvind(1, theta)
  s1 <- sarvind(1, theta)
  expect_true(h1 > 0)
  expect_true(s1 > 0 && s1 < 1)
  expect_equal(darvind(1, theta), h1 * s1, tolerance = 1e-5)
  
  # Random Generation
  set.seed(42)
  r_samp <- rarvind(100, theta)
  expect_length(r_samp, 100)
  expect_true(all(r_samp > 0))
  
  # Character boolean checks
  expect_equal(darvind(1, theta, log = "TRUE"), log(darvind(1, theta)))
  expect_equal(parvind(1, theta, lower.tail = "FALSE"), 1 - parvind(1, theta))
})