context("trim")

################################################################################
# CHANGE LOG
# 22.03.2019: Changed deprecated 'matches' to 'expect_match'.
# 28.04.2014: First tests for 'trim'.
#
# test_dir("inst/tests/")
# test_file("tests/testthat/test-trim.r")
# test_dir("tests/testthat")

test_that("trim", {
  # Columns:
  allele <- c("10", "", "X", "Y", "12", "OL")
  nastrcol <- c("NA", "NA", "NA", "NA", "NA", "NA")
  nacol <- c(NA, NA, NA, NA, NA, NA)
  emptycol <- c("", "", "", "", "", "")

  # Samples:
  samples1 <- c(
    "01-Positive Control", "02-Negative Control", "03-Sample.1",
    "04-Sample.2", "05-Sample.3", "06-Allelic Ladder"
  )

  samples2 <- c(
    "C+", "C-", "C", "C1", "C2", "C3",
    "c+", "c-", "c", "c4", "c5", "c6"
  )

  # Create a dataframe for testing:

  # Test set 1.
  df1 <- data.frame(
    Sample.Name = samples1, Allele = allele,
    NAstr = nastrcol, NAcol = nacol, Empty = emptycol,
    stringsAsFactors = FALSE
  )

  # Test set 2.
  df2 <- data.frame(
    Sample.Name = samples2, Allele = allele,
    NAstr = nastrcol, NAcol = nacol, Empty = emptycol,
    stringsAsFactors = FALSE
  )



  # TEST 01 -------------------------------------------------------------------
  # Test nothing trimmed or changed using df1.

  # Analyse dataframe.
  res <- trim(
    data = df1, samples = NULL, columns = NULL, word = FALSE,
    ignore.case = TRUE, invert.s = FALSE, invert.c = FALSE,
    rm.na.col = FALSE, rm.empty.col = FALSE, missing = NULL, debug = FALSE
  )

  # Check return class.
  expect_match(class(res), class(data.frame()))

  # Check dimensions.
  expect_true(ncol(res) == 5)
  expect_true(nrow(res) == 6)

  # Check that expected columns exist.
  expect_true("Sample.Name" %in% names(res))
  expect_true("Allele" %in% names(res))
  expect_true("NAstr" %in% names(res))
  expect_true("NAcol" %in% names(res))
  expect_true("Empty" %in% names(res))

  # Check for NA's.
  expect_false(any(is.na(res$Sample.Name)))
  expect_false(any(is.na(res$Allele)))
  expect_false(any(is.na(res$NAstr)))
  expect_true(any(is.na(res$NAcol)))
  expect_false(any(is.na(res$Empty)))

  # TEST 02 -------------------------------------------------------------------
  # Test remove empty columns.

  # Analyse dataframe.
  res <- trim(
    data = df1, samples = NULL, columns = NULL, word = FALSE,
    ignore.case = TRUE, invert.s = FALSE, invert.c = FALSE,
    rm.na.col = FALSE, rm.empty.col = TRUE, missing = NULL, debug = FALSE
  )

  # Check return class.
  expect_match(class(res), class(data.frame()))

  # Check dimensions.
  expect_true(ncol(res) == 4)
  expect_true(nrow(res) == 6)

  # Check that expected columns exist.
  expect_true("Sample.Name" %in% names(res))
  expect_true("Allele" %in% names(res))
  expect_true("NAstr" %in% names(res))
  expect_true("NAcol" %in% names(res))
  expect_false("Empty" %in% names(res))

  # Check for NA's.
  expect_false(any(is.na(res$Sample.Name)))
  expect_false(any(is.na(res$Allele)))
  expect_false(any(is.na(res$NAstr)))
  expect_true(any(is.na(res$NAcol)))

  # TEST 03 -------------------------------------------------------------------
  # Test remove NA columns.

  # Analyse dataframe.
  res <- trim(
    data = df1, samples = NULL, columns = NULL, word = FALSE,
    ignore.case = TRUE, invert.s = FALSE, invert.c = FALSE,
    rm.na.col = TRUE, rm.empty.col = FALSE, missing = NULL, debug = FALSE
  )

  # Check return class.
  expect_match(class(res), class(data.frame()))

  # Check dimensions.
  expect_true(ncol(res) == 4)
  expect_true(nrow(res) == 6)

  # Check that expected columns exist.
  expect_true("Sample.Name" %in% names(res))
  expect_true("Allele" %in% names(res))
  expect_true("NAstr" %in% names(res))
  expect_false("NAcol" %in% names(res))
  expect_true("Empty" %in% names(res))

  # Check for NA's.
  expect_false(any(is.na(res$Sample.Name)))
  expect_false(any(is.na(res$Allele)))
  expect_false(any(is.na(res$NAstr)))
  expect_false(any(is.na(res$Empty)))

  # TEST 04 -------------------------------------------------------------------
  # Test replace missing values with NA.

  # Analyse dataframe.
  res <- trim(
    data = df1, samples = NULL, columns = NULL, word = FALSE,
    ignore.case = TRUE, invert.s = FALSE, invert.c = FALSE,
    rm.na.col = FALSE, rm.empty.col = FALSE, missing = NA, debug = FALSE
  )

  # Check return class.
  expect_match(class(res), class(data.frame()))

  # Check dimensions.
  expect_true(ncol(res) == 5)
  expect_true(nrow(res) == 6)

  # Check that expected columns exist.
  expect_true("Sample.Name" %in% names(res))
  expect_true("Allele" %in% names(res))
  expect_true("NAstr" %in% names(res))
  expect_true("NAcol" %in% names(res))
  expect_true("Empty" %in% names(res))

  # Check for NA's.
  expect_false(any(is.na(res$Sample.Name)))
  expect_true(any(is.na(res$Allele)))
  expect_false(any(is.na(res$NAstr)))
  expect_true(any(is.na(res$NAcol)))
  expect_true(any(is.na(res$Empty)))

  # TEST 05 -------------------------------------------------------------------
  # Test replace missing values with string.

  # Analyse dataframe.
  res <- trim(
    data = df1, samples = NULL, columns = NULL, word = FALSE,
    ignore.case = TRUE, invert.s = FALSE, invert.c = FALSE,
    rm.na.col = FALSE, rm.empty.col = FALSE, missing = "N/A", debug = FALSE
  )

  # Check return class.
  expect_match(class(res), class(data.frame()))

  # Check dimensions.
  expect_true(ncol(res) == 5)
  expect_true(nrow(res) == 6)

  # Check that expected columns exist.
  expect_true("Sample.Name" %in% names(res))
  expect_true("Allele" %in% names(res))
  expect_true("NAstr" %in% names(res))
  expect_true("NAcol" %in% names(res))
  expect_true("Empty" %in% names(res))

  # Check for NA's.
  expect_false(any(is.na(res$Sample.Name)))
  expect_false(any(is.na(res$Allele)))
  expect_false(any(is.na(res$NAstr)))
  expect_true(any(is.na(res$NAcol)))
  expect_false(any(is.na(res$Empty)))

  # TEST 06 -------------------------------------------------------------------
  # Test to trim one samples and columns using df1.

  # Analyse dataframe.
  res <- trim(
    data = df1, samples = c("Sample.1"), columns = c("Sample.Name"), word = FALSE,
    ignore.case = TRUE, invert.s = FALSE, invert.c = FALSE,
    rm.na.col = FALSE, rm.empty.col = FALSE, missing = NA, debug = FALSE
  )

  # Check return class.
  expect_match(class(res), class(data.frame()))

  # Check dimensions.
  expect_true(ncol(res) == 1)
  expect_true(nrow(res) == 1)

  # Check that expected columns exist.
  expect_true("Sample.Name" %in% names(res))
  expect_false("Allele" %in% names(res))
  expect_false("NAstr" %in% names(res))
  expect_false("NAcol" %in% names(res))
  expect_false("Empty" %in% names(res))

  # Check for NA's.
  expect_false(any(is.na(res$Sample.Name)))

  # TEST 07 -------------------------------------------------------------------
  # Test to trim using invert for sample and column with df1.

  # Analyse dataframe.
  res <- trim(
    data = df1, samples = c("Sample.1"), columns = c("Sample.Name"), word = FALSE,
    ignore.case = TRUE, invert.s = TRUE, invert.c = TRUE,
    rm.na.col = FALSE, rm.empty.col = FALSE, missing = NA, debug = FALSE
  )

  # Check return class.
  expect_match(class(res), class(data.frame()))

  # Check dimensions.
  expect_true(ncol(res) == 4)
  expect_true(nrow(res) == 5)

  # Check that expected columns exist.
  expect_false("Sample.Name" %in% names(res))
  expect_true("Allele" %in% names(res))
  expect_true("NAstr" %in% names(res))
  expect_true("NAcol" %in% names(res))
  expect_true("Empty" %in% names(res))

  # Check for NA's.
  expect_true(any(is.na(res$Allele)))
  expect_false(any(is.na(res$NAstr)))
  expect_true(any(is.na(res$NAcol)))
  expect_true(any(is.na(res$Empty)))

  # TEST 08 -------------------------------------------------------------------
  # Test to trim controls and columns using df1.

  # Analyse dataframe.
  res <- trim(
    data = df1, samples = c("pos|neg|ladder"), columns = c("Sample.Name|Allele"), word = FALSE,
    ignore.case = TRUE, invert.s = TRUE, invert.c = FALSE,
    rm.na.col = FALSE, rm.empty.col = FALSE, missing = NA, debug = FALSE
  )

  # Check return class.
  expect_match(class(res), class(data.frame()))

  # Check dimensions.
  expect_true(ncol(res) == 2)
  expect_true(nrow(res) == 3)

  # Check that expected columns exist.
  expect_true("Sample.Name" %in% names(res))
  expect_true("Allele" %in% names(res))
  expect_false("NAstr" %in% names(res))
  expect_false("NAcol" %in% names(res))
  expect_false("Empty" %in% names(res))

  # Check for NA's.
  expect_false(any(is.na(res$Sample.Name)))
  expect_false(any(is.na(res$Allele)))

  # TEST 09 -------------------------------------------------------------------
  # Test to trim controls and columns with +/-, and respect case using df2.
  # ...also invert columns, remove NA/empty columns.

  # Analyse dataframe.
  res <- trim(
    data = df2, samples = c("c+|c-"), columns = c("NAstr"), word = FALSE,
    ignore.case = FALSE, invert.s = TRUE, invert.c = TRUE,
    rm.na.col = TRUE, rm.empty.col = TRUE, missing = NA, debug = FALSE
  )

  # Check return class.
  expect_match(class(res), class(data.frame()))

  # Check dimensions.
  expect_true(ncol(res) == 2)
  expect_true(nrow(res) == 10)

  # Check that expected columns exist.
  expect_true("Sample.Name" %in% names(res))
  expect_true("Allele" %in% names(res))
  expect_false("NAstr" %in% names(res))
  expect_false("NAcol" %in% names(res))
  expect_false("Empty" %in% names(res))

  # Check for NA's.
  expect_false(any(is.na(res$Sample.Name)))
  expect_true(any(is.na(res$Allele)))

  # TEST 10 -------------------------------------------------------------------
  # Test to trim controls and columns with +/-, and ignore case using df2.
  # ...also invert columns, remove NA/empty columns.

  # Analyse dataframe.
  res <- trim(
    data = df2, samples = c("c+|c-"), columns = c("NAstr"), word = FALSE,
    ignore.case = TRUE, invert.s = TRUE, invert.c = TRUE,
    rm.na.col = TRUE, rm.empty.col = TRUE, missing = NA, debug = FALSE
  )

  # Check return class.
  expect_match(class(res), class(data.frame()))

  # Check dimensions.
  expect_true(ncol(res) == 2)
  expect_true(nrow(res) == 8)

  # Check that expected columns exist.
  expect_true("Sample.Name" %in% names(res))
  expect_true("Allele" %in% names(res))
  expect_false("NAstr" %in% names(res))
  expect_false("NAcol" %in% names(res))
  expect_false("Empty" %in% names(res))

  # Check for NA's.
  expect_false(any(is.na(res$Sample.Name)))
  expect_false(any(is.na(res$Allele)))

  # TEST 11 -------------------------------------------------------------------
  # Test word boundary using df2.

  # Analyse dataframe.
  res <- trim(
    data = df2, samples = c("c"), columns = c("NAstr"), word = TRUE,
    ignore.case = TRUE, invert.s = FALSE, invert.c = TRUE,
    rm.na.col = FALSE, rm.empty.col = FALSE, missing = "*", debug = FALSE
  )

  # Check return class.
  expect_match(class(res), class(data.frame()))

  # Check dimensions.
  expect_true(ncol(res) == 4)
  expect_true(nrow(res) == 6)

  # Check that expected columns exist.
  expect_true("Sample.Name" %in% names(res))
  expect_true("Allele" %in% names(res))
  expect_false("NAstr" %in% names(res))
  expect_true("NAcol" %in% names(res))
  expect_true("Empty" %in% names(res))

  # Check for NA's.
  expect_false(any(is.na(res$Sample.Name)))
  expect_false(any(is.na(res$Allele)))
  expect_true(any(is.na(res$NAcol)))
  expect_false(any(is.na(res$Empty)))
})