tests/testthat/test-ngrams.R

In jiebaRS: Chinese Text Segmentation, POS Tagging, and Keyword Extraction

test_that("count_ngrams counts a single n", {
  expect_identical(
    count_ngrams(c("我", "爱", "R", "语言"), n = 2),
    data.frame(
      term = c("我 爱", "爱 R", "R 语言"),
      n = c(2L, 2L, 2L),
      count = c(1L, 1L, 1L)
    )
  )
})

test_that("count_ngrams requires optional arguments to be named", {
  expect_snapshot(
    count_ngrams(c("我", "爱", "R", "语言"), 2),
    error = TRUE
  )
})

test_that("count_ngrams counts multiple n values", {
  expect_identical(
    count_ngrams(c("sd", "sd", "sd", "rd"), n = 2:3, sep = ""),
    data.frame(
      term = c("sdsd", "sdrd", "sdsdsd", "sdsdrd"),
      n = c(2L, 2L, 3L, 3L),
      count = c(2L, 1L, 1L, 1L)
    )
  )
})

test_that("count_ngrams supports list input", {
  expect_identical(
    count_ngrams(list(c("a", "b", "c"), c("a", "b")), n = 2),
    data.frame(
      term = c("a b", "b c"),
      n = c(2L, 2L),
      count = c(2L, 1L)
    )
  )
})

test_that("count_ngrams vector format uses raw term names", {
  expect_identical(
    count_ngrams(c("sd", "sd", "sd", "rd"), n = 2:3, sep = "", format = "vector"),
    stats::setNames(
      c(2L, 1L, 1L, 1L),
      c("sdsd", "sdrd", "sdsdsd", "sdsdrd")
    )
  )
})

test_that("count_ngrams sorts by count by default", {
  expect_identical(
    count_ngrams(c("a", "b", "b", "b", "a"), n = 1),
    data.frame(
      term = c("b", "a"),
      n = c(1L, 1L),
      count = c(3L, 2L)
    )
  )

  expect_identical(
    count_ngrams(c("a", "b", "b", "b", "a"), n = 1, format = "vector"),
    stats::setNames(c(3L, 2L), c("b", "a"))
  )
})

test_that("count_ngrams can keep first-appearance order", {
  expect_identical(
    count_ngrams(c("a", "b", "b", "b", "a"), n = 1, sort = FALSE),
    data.frame(
      term = c("a", "b"),
      n = c(1L, 1L),
      count = c(2L, 3L)
    )
  )

  expect_identical(
    count_ngrams(c("a", "b", "b", "b", "a"), n = 1, sort = FALSE, format = "vector"),
    stats::setNames(c(2L, 3L), c("a", "b"))
  )
})

test_that("count_ngrams snapshots invalid inputs", {
  expect_snapshot(count_ngrams(1:3), error = TRUE)
  expect_snapshot(count_ngrams(c("a", "b"), n = c(2, 0)), error = TRUE)
  expect_snapshot(count_ngrams(c("a", "b"), sort = NA), error = TRUE)
})