test_that("conceptIdColumn", {
  expect_no_error(
    standardConceptIdColumn("condition_occurrence")
  )
  expect_identical(
    standardConceptIdColumn("condition_occurrence"),
    "condition_concept_id"
  )
  expect_identical(
    standardConceptIdColumn("drug_exposure"),
    "drug_concept_id"
  )

  expect_identical(
    standardConceptIdColumn("cohort_k"),
    "cohort_definition_id"
  )

  expect_no_error(
    sourceConceptIdColumn("condition_occurrence")
  )
  expect_identical(
    sourceConceptIdColumn("condition_occurrence"),
    "condition_source_concept_id"
  )
  expect_identical(
    sourceConceptIdColumn("drug_exposure"),
    "drug_source_concept_id"
  )
  expect_true(
    is.na(sourceConceptIdColumn("cohort_k"))
  )
})

test_that("working examples", {
  # functionality
  cohort1 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2),
    cohort_start_date = as.Date(c(
      "2020-01-01", "2020-01-15", "2020-01-20", "2020-01-01", "2020-02-01"
    )),
    cohort_end_date = as.Date(c(
      "2020-01-01", "2020-01-15", "2020-01-20", "2020-01-01", "2020-02-01"
    ))
  )

  cohort2 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2, 2, 1),
    cohort_start_date = as.Date(c(
      "2020-01-15", "2020-01-25", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    )),
    cohort_end_date = as.Date(c(
      "2020-01-15", "2020-01-25", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    ))
  )

  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    cohort1 = cohort1,
    cohort2 = cohort2,
    numberIndividuals = 2
  )

  result <- cdm$cohort1 %>%
    .addIntersect(tableName = "cohort2", value = "date", nameStyle = "xx")

  expect_true(length(attributes(cdm$cohort1)) == length(attributes(result)))
  for (i in names(attributes(cdm$cohort1))) {
    if (i != "names" && i != "tbl_name") {
      x <- attr(cdm$cohort1, i)
      y <- attr(result, i)
      if (i == "class") {
        x <- x[x != "GeneratedCohortSet"]
        y <- y[y != "GeneratedCohortSet"]
      }
      expect_true(identical(x, y))
    }
  }

  result <- result %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(colnames(result)[1] == "cohort_definition_id")
  expect_true(colnames(result)[2] == "subject_id")
  expect_true(colnames(result)[3] == "cohort_start_date")
  expect_true(colnames(result)[4] == "cohort_end_date")

  expect_true(all(result %>% dplyr::pull("xx") == as.Date(c("2020-01-15", "2020-01-15", "2020-01-25", "2020-01-24", "2020-03-15"))))

  result1 <- cdm$cohort1 %>%
    .addIntersect(tableName = "cohort2", value = "count") %>%
    .addIntersect(tableName = "cohort2", value = "days") %>%
    .addIntersect(tableName = "cohort2", value = "flag") %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(all(result1$count_all_0_to_inf == c(4, 4, 3, 3, 1)))
  expect_true(all(result1$days_all_0_to_inf == c(14, 0, 5, 23, 43)))
  expect_true(all(result1$flag_all_0_to_inf == c(1, 1, 1, 1, 1)))

  result2 <-
    cdm$cohort1 %>%
    .addIntersect(tableName = "cohort2", value = "count", order = "last") %>%
    .addIntersect(tableName = "cohort2", value = "flag", order = "last") %>%
    .addIntersect(tableName = "cohort2", value = "date", order = "last") %>%
    .addIntersect(tableName = "cohort2", value = "days", order = "last") %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  result6 <-
    cdm$cohort1 %>%
    .addIntersect(tableName = "cohort2", value = c("date", "count", "days", "flag"), order = "last") %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  for (col in colnames(result2)) {
    expect_true(all(result2[[col]][!is.na(result2[[col]])] == result6[[col]][!is.na(result6[[col]])]))
  }

  expect_true(all(result2$date_all_0_to_inf == as.Date(c(
    "2020-02-16", "2020-02-16", "2020-02-16", "2020-03-15", "2020-03-15"
  ))))
  expect_true(all(result2$days_all_0_to_inf == c(46, 32, 27, 74, 43)))
  expect_true(all(result2$count_all_0_to_inf == c(4, 4, 3, 3, 1)))
  expect_true(all(result2$flag_all_0_to_inf == c(1, 1, 1, 1, 1)))

  result3 <-
    cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2",
      window = list(c(-Inf, 0)), value = "date"
    ) %>%
    .addIntersect(
      tableName = "cohort2",
      window = list(c(-Inf, 0)), value = "days"
    ) %>%
    .addIntersect(
      tableName = "cohort2",
      window = list(c(-Inf, 0)), value = "count"
    ) %>%
    .addIntersect(
      tableName = "cohort2",
      window = list(c(-Inf, 0)), value = "flag"
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(all(result3$date_all_minf_to_0 %in% as.Date(
    c(NA, "2020-01-15", "2020-01-15", NA, "2020-01-24")
  )))
  expect_true(all(result3$days_all_minf_to_0 %in% c(NA, 0, -5, NA, -8)))
  expect_true(all(result3$count_all_minf_to_0 == c(0, 1, 1, 0, 2)))
  expect_true(all(result3$flag_all_minf_to_0 == c(0, 1, 1, 0, 1)))

  result4 <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2", window = list(c(-30, 30)), value = "date"
    ) %>%
    .addIntersect(
      tableName = "cohort2", window = list(c(-30, 30)), value = "days"
    ) %>%
    .addIntersect(
      tableName = "cohort2", window = list(c(-30, 30)), value = "count"
    ) %>%
    .addIntersect(
      tableName = "cohort2", window = list(c(-30, 30)), value = "flag"
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(all(result4$date_all_m30_to_30 == as.Date(c(
    "2020-01-15", "2020-01-15", "2020-01-15", "2020-01-24", "2020-01-24"
  ))))
  expect_true(all(result4$days_all_m30_to_30 == c(14, 0, -5, 23, -8)))
  expect_true(all(result4$count_all_m30_to_30 == c(3, 3, 4, 2, 2)))
  expect_true(all(result4$flag_all_m30_to_30 == c(1, 1, 1, 1, 1)))

  result5 <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2",
      window = list(c(-30, 30)),
      value = "date",
      order = "last"
    ) %>%
    .addIntersect(
      tableName = "cohort2",
      window = list(c(-30, 30)),
      value = "days",
      order = "last"
    ) %>%
    .addIntersect(
      tableName = "cohort2",
      window = list(c(-30, 30)),
      value = "count",
      order = "last"
    ) %>%
    .addIntersect(
      tableName = "cohort2",
      window = list(c(-30, 30)),
      value = "flag",
      order = "last"
    ) %>%
    dplyr::arrange(subject_id, cohort_start_date) %>%
    dplyr::collect()


  expect_true(all(result5$count_all_m30_to_30 == c(3, 3, 4, 2, 2)))
  expect_true(all(result5$flag_all_m30_to_30 == c(1, 1, 1, 1, 1)))
  expect_true(all(result5$days_all_m30_to_30 == c(25, 11, 27, 28, -3)))
  expect_true(all(result5$date_all_m30_to_30 == as.Date(c(
    "2020-01-26", "2020-01-26", "2020-02-16", "2020-01-29", "2020-01-29"
  ))))

  mockDisconnect(cdm = cdm)
})

test_that("working examples with cohort_end_date", {
  # functionality
  cohort1 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2),
    cohort_start_date = as.Date(c(
      "2020-01-01", "2020-01-15", "2020-01-20", "2020-01-01", "2020-02-01"
    )),
    cohort_end_date = as.Date(c(
      "2020-01-01", "2020-01-15", "2020-01-20", "2020-01-01", "2020-02-01"
    ))
  )

  cohort2 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2, 2, 1),
    cohort_start_date = as.Date(c(
      "2020-01-15", "2020-01-25", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    )),
    cohort_end_date = as.Date(c(
      "2020-01-15", "2020-01-25", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    ))
  )

  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    cohort1 = cohort1,
    cohort2 = cohort2,
    numberIndividuals = 2
  )

  result <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2", value = "date", indexDate = "cohort_end_date"
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(all(
    result$date_all_0_to_inf == as.Date(c(
      "2020-01-15", "2020-01-15", "2020-01-25", "2020-01-24", "2020-03-15"
    ))
  ))

  mockDisconnect(cdm = cdm)
})

test_that("working examples with extra column", {
  # functionality
  cohort1 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2),
    cohort_start_date = as.Date(c(
      "2020-01-01", "2020-01-15", "2020-01-20", "2020-01-01", "2020-02-01"
    )),
    cohort_end_date = as.Date(c(
      "2020-01-14", "2020-01-15", "2020-01-20", "2020-01-01", "2020-02-01"
    ))
  )

  cohort2 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2, 2, 1),
    cohort_start_date = as.Date(c(
      "2020-01-15", "2020-01-25", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    )),
    cohort_end_date = as.Date(c(
      "2020-01-15", "2020-01-25", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    ))
  )

  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    cohort1 = cohort1,
    cohort2 = cohort2,
    numberIndividuals = 2
  )

  cdm$cohort2 <- cdm$cohort2 %>%
    dbplyr::window_order(
      .data$cohort_definition_id, .data$subject_id, .data$cohort_start_date
    ) %>%
    dplyr::mutate(measurment_result = dplyr::row_number()) %>%
    dbplyr::window_order() %>%
    dplyr::compute()

  result <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2",
      value = c("flag", "measurment_result"),
      filterVariable = "cohort_definition_id",
      filterId = 1,
      idName = "covid",
      window = list(c(0, Inf))
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  result1 <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2",
      value = "measurment_result",
      filterVariable = "cohort_definition_id",
      filterId = 2,
      idName = "covid",
      window = list(c(0, Inf))
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  cohort2 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2, 2, 1),
    cohort_start_date = as.Date(c(
      "2020-01-15", "2020-01-16", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    )),
    cohort_end_date = as.Date(c(
      "2020-01-15", "2020-01-16", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    ))
  )

  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    cohort1 = cohort1,
    cohort2 = cohort2,
    numberIndividuals = 2
  )

  cdm$cohort2 <- cdm$cohort2 %>%
    dbplyr::window_order(
      .data$cohort_definition_id, .data$subject_id, .data$cohort_start_date
    ) %>%
    dplyr::mutate(measurment_result = dplyr::row_number()) %>%
    dbplyr::window_order() %>%
    dplyr::compute()

  result2 <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2", value = "measurment_result",
      nameStyle = "{value}_{window_name}"
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  result3 <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2", value = c("flag", "measurment_result"),
      nameStyle = "{value}_{window_name}",
      window = list(c(-400, -200))
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(all(result$measurment_result_covid_0_to_inf == c(1, 1, 2, 5, 7)))
  expect_true(all(is.na(result1$measurment_result_covid_0_to_inf)))
  expect_true(all(result2$measurment_result_0_to_inf == c("1", "1", 3, 5, 7)))
  expect_true(all(is.na(result3$measurment_result_m400_to_m200)))

  mockDisconnect(cdm = cdm)
})

test_that("working examples with multiple cohort Ids", {
  # functionality
  cohort1 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2),
    cohort_start_date = as.Date(c(
      "2020-01-01", "2020-01-15", "2020-01-20", "2020-01-01", "2020-02-01"
    )),
    cohort_end_date = as.Date(c(
      "2020-01-01", "2020-01-15", "2020-01-20", "2020-01-01", "2020-02-01"
    ))
  )

  cohort2 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 2, 2, 3, 3, 3),
    subject_id = c(1, 1, 1, 2, 2, 2, 1),
    cohort_start_date = as.Date(c(
      "2020-01-15", "2020-01-25", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    )),
    cohort_end_date = as.Date(c(
      "2020-01-15", "2020-01-25", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    ))
  )

  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    cohort1 = cohort1,
    cohort2 = cohort2,
    numberIndividuals = 2
  )

  compareNA <- function(v1, v2) {
    same <- (v1 == v2) | (is.na(v1) & is.na(v2))
    same[is.na(same)] <- FALSE
    return(same)
  }

  result <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2", filterVariable = "cohort_definition_id",
      filterId = 1, value = "date"
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(all(compareNA(result$date_id1_0_to_inf, as.Date(c(
    "2020-01-15", "2020-01-15", "2020-01-25", NA, NA
  )))))

  result1 <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2", filterVariable = "cohort_definition_id",
      filterId = 2, value = "count"
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(all(result1$count_id2_0_to_inf == c(1, 1, 1, 1, 0)))

  result2 <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2", filterVariable = "cohort_definition_id",
      filterId = c(1, 3), value = "flag"
    ) %>%
    .addIntersect(
      tableName = "cohort2", filterVariable = "cohort_definition_id",
      filterId = c(1, 3), value = "count"
    ) %>%
    .addIntersect(
      tableName = "cohort2", filterVariable = "cohort_definition_id",
      filterId = c(1, 3), value = "days"
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  result3 <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2", filterVariable = "cohort_definition_id",
      filterId = c(1, 3), value = c("count", "days", "flag")
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  for (col in colnames(result2)) {
    expect_true(all(
      result2[[col]][!is.na(result2[[col]])] ==
        result3[[col]][!is.na(result3[[col]])]
    ))
  }

  expect_true(all(result2$count_id1_0_to_inf == c(2, 2, 1, 0, 0)))
  expect_true(all(compareNA(result2$days_id1_0_to_inf, c(14, 0, 5, NA, NA))))
  expect_true(all(result2$flag_id1_0_to_inf == c(1, 1, 1, 0, 0)))
  expect_true(all(result2$count_id3_0_to_inf == c(1, 1, 1, 2, 1)))
  expect_true(all(result2$days_id3_0_to_inf == c(46, 32, 27, 23, 43)))
  expect_true(all(result2$flag_id3_0_to_inf == c(1, 1, 1, 1, 1)))

  mockDisconnect(cdm = cdm)
})

test_that("working examples calculating as incidence target cohort", {
  # functionality
  cohort1 <- dplyr::tibble(
    cohort_definition_id = c(1),
    subject_id = c(1),
    cohort_start_date = as.Date("2020-01-15"),
    cohort_end_date = as.Date("2021-01-01")
  )

  cohort2 <- dplyr::tibble(
    cohort_definition_id = c(1),
    subject_id = c(1),
    cohort_start_date = as.Date("2020-01-01"),
    cohort_end_date = as.Date("2020-02-15")
  )

  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    cohort1 = cohort1,
    cohort2 = cohort2
  )

  compareNA <- function(v1, v2) {
    same <- (v1 == v2) | (is.na(v1) & is.na(v2))
    same[is.na(same)] <- FALSE
    return(same)
  }

  result <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2",
      window = c(-Inf, Inf),
      value = "date",
      nameStyle = "test_{id_name}_{window_name}"
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  result1 <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2",
      window = c(-Inf, Inf),
      value = "date",
      targetEndDate = NULL
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(all(result$test_all_minf_to_inf == as.Date("2020-01-01")))
  expect_true(("date_all_minf_to_inf" %in% colnames(result1)))

  mockDisconnect(cdm = cdm)
})

test_that("working examples with more than one window", {
  # functionality
  cohort1 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2),
    cohort_start_date = as.Date(c(
      "2020-01-01", "2020-01-15", "2020-01-20", "2020-01-01", "2020-02-01"
    )),
    cohort_end_date = as.Date(c(
      "2020-01-01", "2020-01-15", "2020-01-20", "2020-01-01", "2020-02-01"
    ))
  )

  cohort2 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 2, 2, 3, 3, 3),
    subject_id = c(1, 1, 1, 2, 2, 2, 1),
    cohort_start_date = as.Date(c(
      "2020-01-15", "2020-01-25", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    )),
    cohort_end_date = as.Date(c(
      "2020-01-15", "2020-01-25", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    ))
  )

  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    cohort1 = cohort1,
    cohort2 = cohort2,
    numberIndividuals = 2
  )

  compareNA <- function(v1, v2) {
    same <- (v1 == v2) | (is.na(v1) & is.na(v2))
    same[is.na(same)] <- FALSE
    return(same)
  }

  result <- cdm$cohort1 %>%
    .addIntersect(tableName = "cohort2", value = "date") %>%
    .addIntersect(
      tableName = "cohort2", value = "date", window = list(c(-Inf, 0))
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  result1 <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2",
      value = "date",
      window = list(c(0, Inf), c(-Inf, 0))
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(all(
    result$date_all_0_to_inf == result1$date_all_0_to_inf
  ))
  expect_true(all(compareNA(
    result$date_all_minf_to_0, result1$date_all_minf_to_0
  )))

  mockDisconnect(cdm = cdm)
})

test_that("working examples with tables, not cohorts", {
  # functionality
  cohort1 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2),
    cohort_start_date = as.Date(
      c(
        "2020-01-01",
        "2020-01-15",
        "2020-01-20",
        "2020-01-01",
        "2020-02-01"
      )
    ),
    cohort_end_date = as.Date(
      c(
        "2020-01-01",
        "2020-01-15",
        "2020-01-20",
        "2020-01-01",
        "2020-02-01"
      )
    )
  )

  drugExposure <- dplyr::tibble(
    drug_exposure_id = 0,
    drug_type_concept_id = 0,
    person_id = c(1, 1),
    drug_concept_id = c(1, 2),
    drug_exposure_start_date = as.Date(c("2020-02-10", "2019-09-01")),
    drug_exposure_end_date = as.Date(NA)
  )

  conditionOccurrence <- dplyr::tibble(
    condition_concept_id = c(1, 1, 2, 2, 3, 3, 3),
    condition_occurrence_id = 0,
    condition_type_concept_id = 0,
    person_id = c(1, 1, 1, 2, 2, 2, 1),
    condition_start_date = as.Date(
      c(
        "2020-01-15",
        "2020-01-25",
        "2020-01-26",
        "2020-01-29",
        "2020-03-15",
        "2020-01-24",
        "2020-02-16"
      )
    ),
    condition_end_date = as.Date(
      c(
        "2020-01-15",
        "2020-01-25",
        "2020-01-26",
        "2020-01-29",
        "2020-03-15",
        "2020-01-24",
        "2020-02-16"
      )
    ),
  )

  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    cohort1 = cohort1,
    condition_occurrence = conditionOccurrence,
    drug_exposure = drugExposure,
    numberIndividuals = 2
  )

  compareNA <- function(v1, v2) {
    same <- (v1 == v2) | (is.na(v1) & is.na(v2))
    same[is.na(same)] <- FALSE
    return(same)
  }

  result <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "condition_occurrence", value = "date",
      targetStartDate = "condition_start_date",
      targetEndDate = "condition_end_date",
      window = list(c(0, Inf), c(-Inf, 0))
    ) %>%
    dplyr::collect()

  expect_true(all(result$date_all_0_to_inf %in% as.Date(c("2020-01-15", "2020-01-15", "2020-01-25", "2020-01-24", "2020-03-15"))))

  result1 <- cdm$condition_occurrence %>%
    .addIntersect(
      tableName = "drug_exposure", value = "count",
      indexDate = "condition_start_date",
      targetStartDate = "drug_exposure_start_date", targetEndDate = NULL,
      window = list(c(0, Inf), c(-Inf, 0)), filterVariable = "drug_concept_id",
      filterId = c(1, 2)
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(person_id, condition_start_date)

  expect_true(all(result1$count_id1_0_to_inf == c(1, 1, 1, 0, 0, 0, 0)))
  # test output all zero column when no result found
  expect_true(all(result1$count_id2_0_to_inf == c(0, 0, 0, 0, 0, 0, 0)))
  expect_true(all(result1$count_id1_minf_to_0 == c(0, 0, 0, 1, 0, 0, 0)))
  expect_true(all(result1$count_id2_minf_to_0 == c(1, 1, 1, 1, 0, 0, 0)))

  result2 <- cdm$condition_occurrence %>%
    .addIntersect(
      tableName = "drug_exposure", value = "count",
      indexDate = "condition_start_date",
      targetStartDate = "drug_exposure_start_date", targetEndDate = NULL,
      window = list(c(0, Inf), c(-Inf, 0))
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(person_id, condition_start_date)

  expect_true(all(result1$count_id1_0_to_inf + result1$count_id2_0_to_inf == result2$count_all_0_to_inf))
  expect_true(all(result1$count_id1_minf_to_0 + result1$count_id2_minf_to_0 == result2$count_all_minf_to_0))

  result3 <- cdm$condition_occurrence %>%
    .addIntersect(
      tableName = "drug_exposure", value = "date",
      indexDate = "condition_start_date",
      targetStartDate = "drug_exposure_start_date", targetEndDate = NULL,
      window = list(c(0, Inf)), filterVariable = "drug_concept_id",
      filterId = c(1, 2)
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(person_id, condition_start_date)
  # test output all zero column when no result found
  expect_true(all(is.na(result3$date_id2_0_to_inf)))

  result4 <- cdm$condition_occurrence %>%
    .addIntersect(
      tableName = "drug_exposure", value = "days",
      indexDate = "condition_start_date",
      targetStartDate = "drug_exposure_start_date", targetEndDate = NULL,
      window = list(c(0, Inf)), filterVariable = "drug_concept_id",
      filterId = c(1, 2)
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(person_id, condition_start_date)
  # test output all zero column when no result found
  expect_true(all(is.na(result4$days_id2_0_to_inf)))

  mockDisconnect(cdm = cdm)
})

test_that("check input length and type for each of the arguments", {
  cohort1 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2),
    cohort_start_date = as.Date(
      c(
        "2020-01-01",
        "2020-01-15",
        "2020-01-20",
        "2020-01-01",
        "2020-02-01"
      )
    ),
    cohort_end_date = as.Date(
      c(
        "2020-01-01",
        "2020-01-15",
        "2020-01-20",
        "2020-01-01",
        "2020-02-01"
      )
    )
  )

  cohort2 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2, 2, 1),
    cohort_start_date = as.Date(
      c(
        "2020-01-15",
        "2020-01-25",
        "2020-01-26",
        "2020-01-29",
        "2020-03-15",
        "2020-01-24",
        "2020-02-16"
      )
    ),
    cohort_end_date = as.Date(
      c(
        "2020-01-15",
        "2020-01-25",
        "2020-01-26",
        "2020-01-29",
        "2020-03-15",
        "2020-01-24",
        "2020-02-16"
      )
    ),
  )

  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    cohort1 = cohort1,
    cohort2 = cohort2,
    numberIndividuals = 2
  )

  expect_error(.addIntersect("cdm$cohort1"))

  expect_error(.addIntersect(cdm$cohort1))

  expect_error(.addIntersect(cdm$cohort1, tableName = "drug"))

  expect_error(.addIntersect(cdm$cohort1, tableName = "cohort2", value = "end_date"))

  expect_error(.addIntersect(cdm$cohort1, tableName = "cohort2", value = "flag", cohortId = "3"))

  expect_error(.addIntersect(cdm$cohort1, tableName = "cohort2", value = "flag", indexDate = 3))

  expect_error(.addIntersect(cdm$cohort1, tableName = "cohort2", value = "flag", targetStartDate = "no"))

  expect_error(.addIntersect(cdm$cohort1, tableName = "cohort2", value = "flag", targetEndDate = NA))

  expect_error(.addIntersect(cdm$cohort1, tableName = "cohort2", value = "flag", nameStyle = "test_{nowindow}_{cohortName}"))

  mockDisconnect(cdm = cdm)
})

test_that("test checkWindow function", {
  cdm <- mockPatientProfiles(con = connection(), writeSchema = writeSchema())

  expect_error(cdm$cohort1 %>%
    .addIntersect(
      value = "days",
      filterId = 1,
      filterVariable = "cohort_definition_id",
      tableName = "cohort2",
      window = c(150, -90)
    ))
})

test_that("test if column exist, overwrite", {
  cohort1 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2),
    cohort_start_date = as.Date(c(
      "2020-01-01", "2020-01-15", "2020-01-20", "2020-01-01", "2020-02-01"
    )),
    cohort_end_date = cohort_start_date,
    flag_all_0_to_30 = c(2, 2, 2, 2, 2),
    count_all_0_to_30 = c(1, 1, 1, 1, 1),
    days_all_0_to_30 = c(1, 1, 1, 1, 1),
    date_all_0_to_30 = c(1, 1, 1, 1, 1)
  )

  cohort2 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2, 2, 1),
    cohort_start_date = as.Date(c(
      "2020-01-15", "2020-01-25", "2020-01-26", "2020-01-29", "2020-03-15",
      "2020-01-24", "2020-02-16"
    )),
    cohort_end_date = cohort_start_date,
  )

  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    cohort1 = cohort1,
    cohort2 = cohort2,
    numberIndividuals = 2
  )

  expect_message(
    result <- cdm$cohort1 %>%
      .addIntersect(
        tableName = "cohort2",
        value = c("flag", "date", "days", "count"), window = list(c(0, 30))
      ) %>%
      dplyr::collect()
  )

  expect_true(sum(colnames(result) == "flag_all_0_to_30") == 1)
  expect_true(all(result %>% dplyr::arrange(cohort_start_date, subject_id) %>%
    dplyr::select(flag_all_0_to_30) !=
    cohort1 %>%
      dplyr::arrange(cohort_start_date, subject_id) %>%
      dplyr::select(flag_all_0_to_30), na.rm = TRUE))

  expect_true(all(result %>% dplyr::arrange(cohort_start_date, subject_id) %>%
    dplyr::select(count_all_0_to_30) !=
    cohort1 %>%
      dplyr::arrange(cohort_start_date, subject_id) %>%
      dplyr::select(count_all_0_to_30), na.rm = TRUE))

  expect_true(all(result %>% dplyr::arrange(cohort_start_date, subject_id) %>%
    dplyr::select(days_all_0_to_30) !=
    cohort1 %>%
      dplyr::arrange(cohort_start_date, subject_id) %>%
      dplyr::select(days_all_0_to_30), na.rm = TRUE))

  expect_true(all(result %>% dplyr::arrange(cohort_start_date, subject_id) %>%
    dplyr::select(date_all_0_to_30) !=
    cohort1 %>%
      dplyr::arrange(cohort_start_date, subject_id) %>%
      dplyr::select(date_all_0_to_30), na.rm = TRUE))

  mockDisconnect(cdm = cdm)
})

test_that("overlapTable is empty, check return columns", {
  # functionality
  cohort2 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1),
    subject_id = c(1, 1, 1, 2, 2),
    cohort_start_date = as.Date(
      c(
        "2020-01-01",
        "2020-01-15",
        "2020-01-20",
        "2020-01-01",
        "2020-02-01"
      )
    ),
    cohort_end_date = as.Date(
      c(
        "2020-01-01",
        "2020-01-15",
        "2020-01-20",
        "2020-01-01",
        "2020-02-01"
      )
    )
  )

  cohort1 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 2, 2, 3, 3, 3),
    subject_id = c(1, 1, 1, 2, 2, 2, 1),
    cohort_start_date = as.Date(
      c(
        "2020-01-15",
        "2020-01-25",
        "2020-01-26",
        "2020-01-29",
        "2020-03-15",
        "2020-01-24",
        "2020-02-16"
      )
    ),
    cohort_end_date = as.Date(
      c(
        "2020-01-15",
        "2020-01-25",
        "2020-01-26",
        "2020-01-29",
        "2020-03-15",
        "2020-01-24",
        "2020-02-16"
      )
    ),
  )

  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    cohort1 = cohort1,
    cohort2 = cohort2,
    numberIndividuals = 2
  )


  result <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2", value = c("date", "days", "count", "flag"),
      filterVariable = "cohort_definition_id",
      filterId = 2
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(all(c(
    "count_id2_0_to_inf", "flag_id2_0_to_inf", "days_id2_0_to_inf",
    "date_id2_0_to_inf"
  ) %in% colnames(result)))

  expect_true(all(result$count_id2_0_to_inf == 0))

  expect_true(all(result$flag_id2_0_to_inf == 0))

  expect_true(all(is.na(result$days_id2_0_to_inf)))

  expect_true(all(is.na(result$date_id2_0_to_inf)))

  mockDisconnect(cdm = cdm)
})

test_that("overlap is empty or not, multiple ids, check return columns", {
  # functionality
  cohort2 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 1, 1, 1, 3),
    subject_id = c(1, 1, 1, 2, 2, 3),
    cohort_start_date = as.Date(
      c(
        "2020-01-01",
        "2020-01-15",
        "2020-01-20",
        "2020-01-01",
        "2020-02-01",
        "2020-03-03"
      )
    ),
    cohort_end_date = as.Date(
      c(
        "2020-01-01",
        "2020-01-15",
        "2020-01-20",
        "2020-01-01",
        "2020-02-01",
        "2020-03-03"
      )
    )
  )

  cohort1 <- dplyr::tibble(
    cohort_definition_id = c(1, 1, 2, 2, 3, 3, 3),
    subject_id = c(3, 3, 3, 3, 3, 3, 3),
    cohort_start_date = as.Date(
      c(
        "2020-01-15",
        "2020-01-25",
        "2020-01-26",
        "2020-01-29",
        "2020-03-15",
        "2020-01-24",
        "2020-02-16"
      )
    ),
    cohort_end_date = as.Date(
      c(
        "2020-01-15",
        "2020-01-25",
        "2020-01-26",
        "2020-01-29",
        "2020-03-15",
        "2020-01-24",
        "2020-02-16"
      )
    ),
  )

  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    cohort1 = cohort1,
    cohort2 = cohort2,
    numberIndividuals = 3
  )

  compareNA <- function(v1, v2) {
    same <- (v1 == v2) | (is.na(v1) & is.na(v2))
    same[is.na(same)] <- FALSE
    return(same)
  }

  result <- cdm$cohort1 %>%
    addCohortIntersectCount(
      targetCohortTable = "cohort2"
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true("cohort_1_0_to_inf" %in% colnames(result))

  expect_true(all(result$cohort_1_0_to_inf == 0))

  result <- cdm$cohort1 %>%
    addCohortIntersectFlag(
      targetCohortTable = "cohort2"
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true("cohort_1_0_to_inf" %in% colnames(result))

  expect_true(all(result$cohort_1_0_to_inf == 0))

  result <- cdm$cohort1 %>%
    addCohortIntersectDate(
      targetCohortTable = "cohort2"
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true("cohort_1_0_to_inf" %in% colnames(result))

  expect_true(all(is.na(result$cohort_1_0_to_inf)))

  result <- cdm$cohort1 %>%
    addCohortIntersectDays(
      targetCohortTable = "cohort2"
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true("cohort_1_0_to_inf" %in% colnames(result))

  expect_true(all(is.na(result$cohort_1_0_to_inf)))

  result <- cdm$cohort1 %>%
    .addIntersect(
      tableName = "cohort2",
      value = c("flag", "date"),
      filterVariable = "cohort_definition_id",
      filterId = c(1, 2, 3),
      window = list(c(0, Inf), c(-30, -1)),
      idName = c("num1", "num2", "num3")
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(all(c(
    "flag_num1_0_to_inf", "flag_num2_0_to_inf", "flag_num3_0_to_inf",
    "date_num1_0_to_inf", "date_num2_0_to_inf", "date_num3_0_to_inf",
    "flag_num1_m30_to_m1", "flag_num2_m30_to_m1", "flag_num3_m30_to_m1",
    "date_num1_m30_to_m1", "date_num2_m30_to_m1", "date_num3_m30_to_m1"
  )
  %in% colnames(result)))

  expect_true(all(compareNA(result$date_num3_0_to_inf, c("2020-03-03", "2020-03-03", "2020-03-03", "2020-03-03", "2020-03-03", "2020-03-03", NA))))
  expect_true(all(compareNA(result$date_num3_m30_to_m1, c(NA, NA, NA, NA, NA, NA, "2020-03-03"))))
  expect_true(all(result$flag_num3_0_to_inf == c(1, 1, 1, 1, 1, 1, 0)))
  expect_true(all(result$flag_num3_m30_to_m1 == c(0, 0, 0, 0, 0, 0, 1)))

  expect_true(all(result$flag_num2_0_to_inf == c(0, 0, 0, 0, 0, 0, 0)))
  expect_true(all(result$flag_num1_0_to_inf == c(0, 0, 0, 0, 0, 0, 0)))
  expect_true(all(result$flag_num2_m30_to_m1 == c(0, 0, 0, 0, 0, 0, 0)))
  expect_true(all(result$flag_num1_m30_to_m1 == c(0, 0, 0, 0, 0, 0, 0)))

  expect_true(all(is.na(result$date_num2_0_to_inf)))
  expect_true(all(is.na(result$date_num1_0_to_inf)))
  expect_true(all(is.na(result$date_num2_m30_to_m1)))
  expect_true(all(is.na(result$date_num1_m30_to_m1)))

  expect_error(cdm$cohort1 %>%
    addCohortIntersectDate(
      targetCohortTable = "cohort2",
      targetCohortId = c(1, 2, 3),
      window = list(c(0, Inf), c(-30, -1))
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date))

  result <- cdm$cohort1 %>%
    addCohortIntersectDate(
      targetCohortTable = "cohort2",
      targetCohortId = c(1, 3),
      window = list(c(0, Inf), c(-30, -1))
    ) %>%
    dplyr::collect() |>
    dplyr::arrange(subject_id, cohort_start_date)

  expect_true(all(c(
    "cohort_3_m30_to_m1", "cohort_1_m30_to_m1",
    "cohort_3_0_to_inf", "cohort_1_0_to_inf"
  )
  %in% colnames(result)))

  expect_true(all(compareNA(result$cohort_3_0_to_inf, c("2020-03-03", "2020-03-03", "2020-03-03", "2020-03-03", "2020-03-03", "2020-03-03", NA))))
  expect_true(all(compareNA(result$cohort_3_m30_to_m1, c(NA, NA, NA, NA, NA, NA, "2020-03-03"))))

  expect_true(all(is.na(result$cohort_1_m30_to_m1)))
  expect_true(all(is.na(result$cohort_1_0_to_inf)))

  mockDisconnect(cdm = cdm)
})

test_that("non snake columns not repeated in output", {
  cdm <- mockPatientProfiles(con = connection(), writeSchema = writeSchema())
  attr(cdm$cohort1, "cohort_set") <- attr(cdm$cohort1, "cohort_set") %>% dplyr::mutate(cohort_name = toupper(cohort_name))
  cdm$cohort2 <- cdm$cohort2 %>%
    addCohortIntersectFlag(targetCohortTable = "cohort1")

  expect_true("cohort_1_0_to_inf" %in% colnames(cdm$cohort2))
  expect_false("COHORT_1_0_to_inf" %in% colnames(cdm$cohort2))
})

test_that("no NA when overwrite column", {
  cdm <- mockPatientProfiles(
    con = connection(),
    writeSchema = writeSchema(),
    numberIndividuals = 1000
  )

  # Presence in characteristis 'cohort 1' in 180 days before cohort start
  cdm$cohort1 <- cdm$cohort1 %>%
    addCohortIntersectFlag(
      targetCohortTable = "cohort2",
      window = list(c(-180, -1)),
      targetCohortId = 1,
      nameStyle = "{cohort_name}"
    )


  # Trying to overwrite the previous created variable, for example because the characteristics cohort has changed.
  expect_message(
    cdm$cohort1 <- cdm$cohort1 %>%
      addCohortIntersectFlag(
        targetCohortTable = "cohort2",
        window = list(c(-180, -1)),
        targetCohortId = 1,
        nameStyle = "{cohort_name}"
      )
  )

  expect_true(!any(is.na(cdm$cohort1 %>% dplyr::pull("cohort_1"))))

  # subject 2, who has no record for cohort_definition_id 1 in the characteristics cohort, now gets a "NA"

  # Moving the "cohort_definition_id == 1" records from subject 1 to subject 2:
  cdm$cohort2 <- cdm$cohort2 %>%
    dplyr::mutate(subject_id = dplyr::if_else(cohort_definition_id == 1 & subject_id == 1, 2, subject_id))

  expect_message(
    cdm$cohort1 <- cdm$cohort1 %>%
      addCohortIntersectFlag(
        targetCohortTable = "cohort2",
        window = list(c(-180, -1)),
        targetCohortId = 1,
        nameStyle = "{cohort_name}"
      )
  )

  expect_true(!any(is.na(cdm$cohort1 %>% dplyr::pull("cohort_1"))))

  mockDisconnect(cdm = cdm)
})