# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements.  See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership.  The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License.  You may obtain a copy of the License at
#
#   http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied.  See the License for the
# specific language governing permissions and limitations
# under the License.

# In 3.4 the lack of tzone attribute causes spurious failures
skip_on_r_older_than("3.5")

library(lubridate, warn.conflicts = FALSE)
library(dplyr, warn.conflicts = FALSE)

skip_if_not_available("acero")
# Skip these tests on CRAN due to build times > 10 mins
skip_on_cran()

# base::strptime() defaults to local timezone
# but arrow's strptime defaults to UTC.
# So that tests are consistent, set the local timezone to UTC
# TODO: consider reevaluating now that ARROW-12980 has merged
withr::local_timezone("UTC")

if (tolower(Sys.info()[["sysname"]]) == "windows") {
  withr::local_locale(LC_TIME = "C")
}

test_date <- as.POSIXct("2017-01-01 00:00:11.3456789", tz = "Pacific/Marquesas")

strptime_test_df <- tibble(
  string_a = c("2023-12-30-Sat", NA),
  string_A = c("2023-12-30-Saturday", NA),
  string_b = c("2023-12-30-Dec", NA),
  string_B = c("2023-12-30-December", NA),
  string_H = c("2023-12-30-01", NA),
  string_I = c("2023-12-30-01", NA),
  string_j = c("2023-12-30-364", NA),
  string_M = c("2023-12-30-45", NA),
  string_p = c("2023-12-30-AM", NA),
  string_q = c("2023.3", NA),
  string_S = c("2023-12-30-56", NA),
  string_OS = c("2023-12-30-12.345678", NA),
  string_U = c("2023-12-30-52", NA),
  string_w = c("2023-12-30-6", NA),
  string_W = c("2023-12-30-52", NA),
  string_y = c("23-12-30", NA),
  string_Y = c("2023-12-30", NA),
  string_m = c("2023-12-30", NA),
  string_r = c("2023-12-30-01", NA),
  string_R = c("2023-12-30-01:23", NA),
  string_T = c("2023-12-30-01:23:45", NA),
  string_z = c("2023-12-30-01:23:45z", NA)
)

test_df <- tibble::tibble(
  # test_date + 1 turns the tzone = "" to NULL, which is functionally equivalent
  # so we can run some tests on Windows, but this skirts around ARROW-13588.
  # That issue is tough because in C++, "" is the "no timezone" value
  # due to static typing, so we can't distinguish a literal "" from NULL
  datetime = c(test_date, NA) + 1,
  date = c(as.Date("2021-09-09"), NA),
  integer = 1:2
)


test_that("strptime", {
  t_string <- tibble(x = c("2018-10-07 19:04:05", NA))
  # lubridate defaults to "UTC" as timezone => t_stamp is in "UTC"
  t_stamp_with_utc_tz <- tibble(x = c(lubridate::ymd_hms("2018-10-07 19:04:05"), NA))
  t_stamp_with_pm_tz <- tibble(
    x = c(lubridate::ymd_hms("2018-10-07 19:04:05", tz = "Pacific/Marquesas"), NA)
  )

  # base::strptime returns a POSIXlt (a list) => we cannot use compare_dplyr_binding
  # => we use expect_equal for the tests below

  withr::with_timezone("Pacific/Marquesas", {
    # the default value for strptime's `tz` argument is "", which is interpreted
    # as the current timezone. we test here if the strptime binding picks up
    # correctly the current timezone (similarly to the base R version)
    expect_equal(
      t_string %>%
        record_batch() %>%
        mutate(
          x = strptime(x, format = "%Y-%m-%d %H:%M:%S")
        ) %>%
        collect(),
      t_stamp_with_pm_tz
    )

    expect_equal(
      t_string %>%
        record_batch() %>%
        mutate(
          x = base::strptime(x, format = "%Y-%m-%d %H:%M:%S")
        ) %>%
        collect(),
      t_stamp_with_pm_tz
    )
  })

  # adding a timezone to a timezone-naive timestamp works
  # and since our TZ when running the test is (typically) Pacific/Marquesas
  # this also tests that assigning a TZ different from the current session one
  # works as expected
  expect_equal(
    t_string %>%
      arrow_table() %>%
      mutate(
        x = strptime(x, format = "%Y-%m-%d %H:%M:%S", tz = "Pacific/Marquesas")
      ) %>%
      collect(),
    t_stamp_with_pm_tz
  )

  expect_equal(
    t_string %>%
      Table$create() %>%
      mutate(
        x = strptime(x, tz = "UTC")
      ) %>%
      collect(),
    t_stamp_with_utc_tz
  )

  expect_equal(
    t_string %>%
      Table$create() %>%
      mutate(
        x = strptime(x, format = "%Y-%m-%d %H:%M:%S", tz = "UTC")
      ) %>%
      collect(),
    t_stamp_with_utc_tz
  )

  expect_equal(
    t_string %>%
      Table$create() %>%
      mutate(
        x = strptime(x, format = "%Y-%m-%d %H:%M:%S", unit = "ns", tz = "UTC")
      ) %>%
      collect(),
    t_stamp_with_utc_tz
  )

  expect_equal(
    t_string %>%
      Table$create() %>%
      mutate(
        x = strptime(x, format = "%Y-%m-%d %H:%M:%S", unit = "s", tz = "UTC")
      ) %>%
      collect(),
    t_stamp_with_utc_tz
  )

  tstring <- tibble(x = c("08-05-2008", NA))
  tstamp <- strptime(c("08-05-2008", NA), format = "%m-%d-%Y")

  expect_equal(
    tstring %>%
      Table$create() %>%
      mutate(
        x = strptime(x, format = "%m-%d-%Y")
      ) %>%
      pull() %>%
      as.vector(),
    # R's strptime returns POSIXlt (list type)
    as.POSIXct(tstamp),
    ignore_attr = "tzone"
  )

  # these functions' internals use some string processing which requires the
  # RE2 library
  skip_if_not_available("re2")

  compare_dplyr_binding(
    .input %>%
      mutate(
        parsed_date_ymd = parse_date_time(string_1, orders = "Y-%m-d-%T")
      ) %>%
      collect(),
    tibble::tibble(string_1 = c("2022-02-11-12:23:45", NA))
  )
})

test_that("strptime works for individual formats", {
  # strptime format support is not consistent across platforms
  skip_on_cran()

  # these functions' internals use some string processing which requires the
  # RE2 library
  skip_if_not_available("re2")

  expect_equal(
    strptime_test_df %>%
      arrow_table() %>%
      mutate(
        parsed_H = strptime(string_H, format = "%Y-%m-%d-%H"),
        parsed_I = strptime(string_I, format = "%Y-%m-%d-%I"),
        parsed_j = strptime(string_j, format = "%Y-%m-%d-%j"),
        parsed_M = strptime(string_M, format = "%Y-%m-%d-%M"),
        parsed_S = strptime(string_S, format = "%Y-%m-%d-%S"),
        parsed_U = strptime(string_U, format = "%Y-%m-%d-%U"),
        parsed_w = strptime(string_w, format = "%Y-%m-%d-%w"),
        parsed_W = strptime(string_W, format = "%Y-%m-%d-%W"),
        parsed_y = strptime(string_y, format = "%y-%m-%d"),
        parsed_Y = strptime(string_Y, format = "%Y-%m-%d"),
        parsed_R = strptime(string_R, format = "%Y-%m-%d-%R"),
        parsed_T = strptime(string_T, format = "%Y-%m-%d-%T")
      ) %>%
      collect(),
    strptime_test_df %>%
      mutate(
        parsed_H = as.POSIXct(strptime(string_H, format = "%Y-%m-%d-%H")),
        parsed_I = as.POSIXct(strptime(string_I, format = "%Y-%m-%d-%I")),
        parsed_j = as.POSIXct(strptime(string_j, format = "%Y-%m-%d-%j")),
        parsed_M = as.POSIXct(strptime(string_M, format = "%Y-%m-%d-%M")),
        parsed_S = as.POSIXct(strptime(string_S, format = "%Y-%m-%d-%S")),
        parsed_U = as.POSIXct(strptime(string_U, format = "%Y-%m-%d-%U")),
        parsed_w = as.POSIXct(strptime(string_w, format = "%Y-%m-%d-%w")),
        parsed_W = as.POSIXct(strptime(string_W, format = "%Y-%m-%d-%W")),
        parsed_y = as.POSIXct(strptime(string_y, format = "%y-%m-%d")),
        parsed_Y = as.POSIXct(strptime(string_Y, format = "%Y-%m-%d")),
        parsed_R = as.POSIXct(strptime(string_R, format = "%Y-%m-%d-%R")),
        parsed_T = as.POSIXct(strptime(string_T, format = "%Y-%m-%d-%T"))
      ) %>%
      collect()
  )

  # Some formats are not supported on Windows
  skip_on_os("windows")
  expect_equal(
    strptime_test_df %>%
      arrow_table() %>%
      mutate(
        parsed_a = strptime(string_a, format = "%Y-%m-%d-%a"),
        parsed_A = strptime(string_A, format = "%Y-%m-%d-%A"),
        parsed_b = strptime(string_b, format = "%Y-%m-%d-%b"),
        parsed_B = strptime(string_B, format = "%Y-%m-%d-%B"),
        parsed_p = strptime(string_p, format = "%Y-%m-%d-%p"),
        parsed_r = strptime(string_r, format = "%Y-%m-%d-%r")
      ) %>%
      collect(),
    strptime_test_df %>%
      mutate(
        parsed_a = as.POSIXct(strptime(string_a, format = "%Y-%m-%d-%a")),
        parsed_A = as.POSIXct(strptime(string_A, format = "%Y-%m-%d-%A")),
        parsed_b = as.POSIXct(strptime(string_b, format = "%Y-%m-%d-%b")),
        parsed_B = as.POSIXct(strptime(string_B, format = "%Y-%m-%d-%B")),
        parsed_p = as.POSIXct(strptime(string_p, format = "%Y-%m-%d-%p")),
        parsed_r = as.POSIXct(strptime(string_r, format = "%Y-%m-%d-%r"))
      ) %>%
      collect()
  )
})

test_that("timestamp round trip correctly via strftime and strptime", {
  # strptime format support is not consistent across platforms
  skip_on_cran()

  # these functions' internals use some string processing which requires the
  # RE2 library
  skip_if_not_available("re2")

  tz <- "Pacific/Marquesas"
  set.seed(42)
  times <- seq(as.POSIXct("1999-02-07", tz = tz), as.POSIXct("2000-01-01", tz = tz), by = "sec")
  times <- sample(times, 100)

  # Op format is currently not supported by strptime
  formats <- c(
    "%d", "%H", "%j", "%m", "%T",
    "%S", "%q", "%M", "%U", "%w", "%W", "%y", "%Y", "%R", "%T"
  )
  formats2 <- c(
    "a", "A", "b", "B", "d", "H", "j", "m", "T", "OS", "Ip",
    "S", "q", "M", "U", "w", "W", "y", "Y", "r", "R", "Tz"
  )
  base_format <- "%Y-%m-%d"
  base_format2 <- "ymd"

  # Some formats are not supported on Windows
  if (!tolower(Sys.info()[["sysname"]]) == "windows") {
    # "%r" could also be here, though it is only valid in some locales (those
    # that use 12 hour formats, so skip for now)
    formats <- c(formats, "%a", "%A", "%b", "%B", "%OS", "%I%p", "%T%z")
  }

  for (fmt in formats) {
    fmt <- paste(base_format, fmt)
    test_df <- tibble::tibble(x = strftime(times, format = fmt))
    expect_equal(
      test_df %>%
        arrow_table() %>%
        mutate(!!fmt := strptime(x, format = fmt)) %>%
        collect(),
      test_df %>%
        mutate(!!fmt := as.POSIXct(strptime(x, format = fmt))) %>%
        collect()
    )
  }

  for (fmt in formats2) {
    fmt2 <- paste(base_format2, fmt)
    fmt <- paste(base_format, paste0("%", fmt))
    test_df <- tibble::tibble(x = strftime(times, format = fmt))
    expect_equal_data_frame(
      test_df %>%
        arrow_table() %>%
        mutate(!!fmt := strptime(x, format = fmt2)) %>%
        collect(),
      test_df %>%
        mutate(!!fmt := as.POSIXct(strptime(x, format = fmt2))) %>%
        collect()
    )
  }
})

test_that("strptime returns NA when format doesn't match the data", {
  df <- tibble(
    str_date = c("2022-02-07", "2012/02-07", "1975/01-02", "1981/01-07", NA)
  )

  # base::strptime() returns a POSIXlt object (a list), while the Arrow binding
  # returns a POSIXct (double) vector => we cannot use compare_dplyr_binding()
  expect_equal(
    df %>%
      arrow_table() %>%
      mutate(
        r_obj_parsed_date = strptime("03-27/2022", format = "%m-%d/%Y"),
        r_obj_parsed_na = strptime("03-27/2022", format = "Y%-%m-%d")
      ) %>%
      collect(),
    df %>%
      mutate(
        r_obj_parsed_date = as.POSIXct(strptime("03-27/2022", format = "%m-%d/%Y")),
        r_obj_parsed_na = as.POSIXct(strptime("03-27/2022", format = "Y%-%m-%d"))
      ),
    ignore_attr = "tzone"
  )

  expect_equal(
    df %>%
      record_batch() %>%
      mutate(parsed_date = strptime(str_date, format = "%Y-%m-%d")) %>%
      collect(),
    df %>%
      mutate(parsed_date = as.POSIXct(strptime(str_date, format = "%Y-%m-%d"))),
    ignore_attr = "tzone"
  )

  expect_equal(
    df %>%
      arrow_table() %>%
      mutate(parsed_date = strptime(str_date, format = "%Y/%m-%d")) %>%
      collect(),
    df %>%
      mutate(parsed_date = as.POSIXct(strptime(str_date, format = "%Y/%m-%d"))),
    ignore_attr = "tzone"
  )
})

test_that("strftime", {
  times <- tibble(
    datetime = c(lubridate::ymd_hms("2018-10-07 19:04:05", tz = "Etc/GMT+6"), NA),
    date = c(as.Date("2021-01-01"), NA)
  )
  formats <- "%a %A %w %d %b %B %m %y %Y %H %I %p %M %z %Z %j %U %W %x %X %% %G %V %u"
  formats_date <- "%a %A %w %d %b %B %m %y %Y %H %I %p %M %j %U %W %x %X %% %G %V %u"

  compare_dplyr_binding(
    .input %>%
      mutate(
        x = strftime(datetime, format = formats),
        x2 = base::strftime(datetime, format = formats)
      ) %>%
      collect(),
    times
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = strftime(date, format = formats_date)) %>%
      collect(),
    times
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = strftime(datetime, format = formats, tz = "Pacific/Marquesas")) %>%
      collect(),
    times
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = strftime(datetime, format = formats, tz = "EST", usetz = TRUE)) %>%
      collect(),
    times
  )

  withr::with_timezone(
    "Pacific/Marquesas",
    {
      compare_dplyr_binding(
        .input %>%
          mutate(
            x = strftime(datetime, format = formats, tz = "EST"),
            x_date = strftime(date, format = formats_date, tz = "EST")
          ) %>%
          collect(),
        times
      )

      compare_dplyr_binding(
        .input %>%
          mutate(
            x = strftime(datetime, format = formats),
            x_date = strftime(date, format = formats_date)
          ) %>%
          collect(),
        times
      )
    }
  )

  # This check is due to differences in the way %c currently works in Arrow and R's strftime.
  # We can revisit after https://github.com/HowardHinnant/date/issues/704 is resolved.
  if (Sys.getlocale("LC_TIME") != "C") {
    expect_error(
      times %>%
        Table$create() %>%
        mutate(x = strftime(datetime, format = "%c")) %>%
        collect(),
      "%c flag is not supported in non-C locales."
    )
  }

  # Output precision of %S depends on the input timestamp precision.
  # Timestamps with second precision are represented as integers while
  # milliseconds, microsecond and nanoseconds are represented as fixed floating
  # point numbers with 3, 6 and 9 decimal places respectively.
  compare_dplyr_binding(
    .input %>%
      mutate(x = strftime(datetime, format = "%S")) %>%
      transmute(as.double(substr(x, 1, 2))) %>%
      collect(),
    times,
    tolerance = 1e-6
  )
})

test_that("format_ISO8601", {
  # https://issues.apache.org/jira/projects/ARROW/issues/ARROW-15266
  skip_if_not_available("re2")
  # A change in R altered the behavior of lubridate::format_ISO8601:
  # https://github.com/wch/r-source/commit/f6fd993f8a2f799a56dbecbd8238f155191fc31b
  # Fixed in lubridate here:
  # https://github.com/tidyverse/lubridate/pull/1068
  skip_if_not(packageVersion("lubridate") > "1.8")

  times <- tibble(x = c(lubridate::ymd_hms("2018-10-07 19:04:05", tz = "Etc/GMT+6"), NA))

  compare_dplyr_binding(
    .input %>%
      mutate(
        a = format_ISO8601(x, precision = "ymd", usetz = FALSE),
        a2 = lubridate::format_ISO8601(x, precision = "ymd", usetz = FALSE)
      ) %>%
      collect(),
    times
  )

  if (getRversion() < "3.5") {
    # before 3.5, times$x will have no timezone attribute, so Arrow faithfully
    # errors that there is no timezone to format:
    expect_error(
      times %>%
        Table$create() %>%
        mutate(x = format_ISO8601(x, precision = "ymd", usetz = TRUE)) %>%
        collect(),
      "Timezone not present, cannot convert to string with timezone: %Y-%m-%d%z"
    )

    # See comment regarding %S flag in strftime tests
    expect_error(
      times %>%
        Table$create() %>%
        mutate(x = format_ISO8601(x, precision = "ymdhms", usetz = TRUE)) %>%
        mutate(x = gsub("\\.0*", "", x)) %>%
        collect(),
      "Timezone not present, cannot convert to string with timezone: %Y-%m-%dT%H:%M:%S%z"
    )
  } else {
    compare_dplyr_binding(
      .input %>%
        mutate(x = format_ISO8601(x, precision = "ymd", usetz = TRUE)) %>%
        collect(),
      times
    )

    # See comment regarding %S flag in strftime tests
    compare_dplyr_binding(
      .input %>%
        mutate(x = format_ISO8601(x, precision = "ymdhms", usetz = TRUE)) %>%
        mutate(x = gsub("\\.0*", "", x)) %>%
        collect(),
      times
    )
  }


  # See comment regarding %S flag in strftime tests
  compare_dplyr_binding(
    .input %>%
      mutate(x = format_ISO8601(x, precision = "ymdhms", usetz = FALSE)) %>%
      mutate(x = gsub("\\.0*", "", x)) %>%
      collect(),
    times
  )
})

# These tests test detection of dates and times

test_that("is.* functions from lubridate", {
  # make sure all true and at least one false value is considered
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = is.POSIXct(datetime),
        y = is.POSIXct(integer),
        x2 = lubridate::is.POSIXct(datetime)
      ) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        x = is.Date(date),
        y = is.Date(integer),
        x2 = lubridate::is.Date(date)
      ) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        x = is.instant(datetime),
        y = is.instant(date),
        z = is.instant(integer)
      ) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        x = is.timepoint(datetime),
        y = is.instant(date),
        z = is.timepoint(integer),
        x2 = lubridate::is.timepoint(datetime),
        y2 = lubridate::is.instant(date),
        z2 = lubridate::is.timepoint(integer)
      ) %>%
      collect(),
    test_df
  )
})

# These tests test component extraction from timestamp objects

test_that("extract year from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = year(datetime)) %>%
      collect(),
    test_df
  )
})

test_that("extract isoyear from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = isoyear(datetime)) %>%
      collect(),
    test_df
  )
})

test_that("extract epiyear from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = epiyear(datetime),
        x2 = lubridate::epiyear(datetime)
      ) %>%
      collect(),
    test_df
  )
})

test_that("extract quarter from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = quarter(datetime)) %>%
      collect(),
    test_df
  )
})

test_that("extract month from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = month(datetime),
        x2 = lubridate::month(datetime)
      ) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      # R returns ordered factor whereas Arrow returns character
      mutate(x = as.character(month(datetime, label = TRUE))) %>%
      collect(),
    test_df,
    ignore_attr = TRUE
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = as.character(month(datetime, label = TRUE, abbr = TRUE))) %>%
      collect(),
    test_df,
    ignore_attr = TRUE
  )
})

test_that("extract isoweek from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = isoweek(datetime),
        x2 = lubridate::isoweek(datetime)
      ) %>%
      collect(),
    test_df
  )
})

test_that("extract epiweek from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = epiweek(datetime)) %>%
      collect(),
    test_df
  )
})

test_that("extract week from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = week(datetime),
        x2 = lubridate::week(datetime)
      ) %>%
      collect(),
    test_df
  )
})

test_that("extract day from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = day(datetime)) %>%
      collect(),
    test_df
  )
})

test_that("extract wday from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = wday(datetime)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = wday(date, week_start = 3)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = wday(date, week_start = 1)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = wday(date, label = TRUE)) %>%
      mutate(x = as.character(x)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = wday(datetime, label = TRUE, abbr = TRUE)) %>%
      mutate(x = as.character(x)) %>%
      collect(),
    test_df
  )
})

test_that("extract mday from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = mday(datetime)) %>%
      collect(),
    test_df
  )
})

test_that("extract yday from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = yday(datetime),
        x2 = lubridate::yday(datetime)
      ) %>%
      collect(),
    test_df
  )
})

test_that("extract qday from timestamp", {
  test_df <- tibble::tibble(
    time = as.POSIXct(seq(as.Date("1999-12-31", tz = "UTC"), as.Date("2001-01-01", tz = "UTC"), by = "day"))
  )

  compare_dplyr_binding(
    .input %>%
      transmute(x = qday(time)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      transmute(x = qday(as.POSIXct("2022-06-29 12:35"))) %>%
      collect(),
    test_df
  )
})

test_that("extract hour from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = hour(datetime),
        x2 = lubridate::hour(datetime)
      ) %>%
      collect(),
    test_df
  )
})

test_that("extract minute from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = minute(datetime),
        x2 = lubridate::minute(datetime)
      ) %>%
      collect(),
    test_df
  )
})

test_that("extract second from timestamp", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = second(datetime),
        x2 = lubridate::second(datetime)
      ) %>%
      collect(),
    test_df,
    # arrow supports nanosecond resolution but lubridate does not
    tolerance = 1e-6
  )
})

# These tests test extraction of components from date32 objects

test_that("extract year from date", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = year(date),
        x2 = lubridate::year(date)
      ) %>%
      collect(),
    test_df
  )
})

test_that("extract isoyear from date", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = isoyear(date),
        x2 = lubridate::isoyear(date)
      ) %>%
      collect(),
    test_df
  )
})

test_that("extract epiyear from date", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = epiyear(date)) %>%
      collect(),
    test_df
  )
})

test_that("extract quarter from date", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = quarter(date),
        x2 = lubridate::quarter(date)
      ) %>%
      collect(),
    test_df
  )
})

test_that("extract isoweek from date", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = isoweek(date)) %>%
      collect(),
    test_df
  )
})

test_that("extract epiweek from date", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = epiweek(date),
        x2 = lubridate::epiweek(date)
      ) %>%
      collect(),
    test_df
  )
})

test_that("extract week from date", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = week(date)) %>%
      collect(),
    test_df
  )
})

test_that("extract month from date", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = month(date)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      # R returns ordered factor whereas Arrow returns character
      mutate(x = as.character(month(date, label = TRUE))) %>%
      collect(),
    test_df,
    ignore_attr = TRUE
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = as.character(month(date, label = TRUE, abbr = TRUE))) %>%
      collect(),
    test_df,
    ignore_attr = TRUE
  )
})

test_that("extract day from date", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = day(date),
        x2 = lubridate::day(date)
      ) %>%
      collect(),
    test_df
  )
})

test_that("extract wday from date", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = wday(date)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        x = wday(date, week_start = 3),
        x2 = lubridate::wday(date, week_start = 3)
      ) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = wday(date, week_start = 1)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = wday(date, label = TRUE, abbr = TRUE)) %>%
      mutate(x = as.character(x)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = wday(date, label = TRUE)) %>%
      mutate(x = as.character(x)) %>%
      collect(),
    test_df
  )
})

test_that("extract mday from date", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = mday(date),
        x2 = lubridate::mday(date)
      ) %>%
      collect(),
    test_df
  )
})

test_that("extract yday from date", {
  compare_dplyr_binding(
    .input %>%
      mutate(x = yday(date)) %>%
      collect(),
    test_df
  )
})

test_that("extract qday from date", {
  test_df <- tibble::tibble(
    date = seq(as.Date("1999-12-31"), as.Date("2001-01-01"), by = "day")
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = qday(date)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(y = qday(as.Date("2022-06-29"))) %>%
      collect(),
    test_df
  )
})

test_that("leap_year mirror lubridate", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        x = leap_year(date),
        x2 = lubridate::leap_year(date)
      ) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = leap_year(datetime)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(x = leap_year(test_year)) %>%
      collect(),
    tibble::tibble(
      test_year = as.Date(c(
        "1998-01-01", # not leap year
        "1996-01-01", # leap year (divide by 4 rule)
        "1900-01-01", # not leap year (divide by 100 rule)
        "2000-01-01" # leap year (divide by 400 rule)
      ))
    )
  )
})

test_that("am/pm mirror lubridate", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        am = am(test_time),
        pm = pm(test_time),
        am2 = lubridate::am(test_time),
        pm2 = lubridate::pm(test_time)
      ) %>%
      # can't use collect() here due to how tibbles store datetimes
      # TODO: add better explanation above
      as.data.frame(),
    data.frame(
      test_time = strptime(
        x = c(
          "2022-01-25 11:50:59",
          "2022-01-25 12:00:00",
          "2022-01-25 00:00:00"
        ),
        format = "%Y-%m-%d %H:%M:%S"
      )
    )
  )
})

test_that("extract tz", {
  df <- tibble(
    posixct_date = as.POSIXct(c("2022-02-07", "2022-02-10"), tz = "Pacific/Marquesas"),
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        timezone_posixct_date = tz(posixct_date),
        timezone_posixct_date2 = lubridate::tz(posixct_date)
      ) %>%
      collect(),
    df
  )

  # test a few types directly from R objects
  expect_error(
    call_binding("tz", "2020-10-01"),
    "timezone extraction for objects of class `string` not supported in Arrow"
  )
  expect_error(
    call_binding("tz", as.Date("2020-10-01")),
    "timezone extraction for objects of class `date32[day]` not supported in Arrow",
    fixed = TRUE
  )
  expect_error(
    call_binding("tz", 1L),
    "timezone extraction for objects of class `int32` not supported in Arrow"
  )
  expect_error(
    call_binding("tz", 1.1),
    "timezone extraction for objects of class `double` not supported in Arrow"
  )

  # Test one expression
  expect_error(
    call_binding("tz", Expression$scalar("2020-10-01")),
    "timezone extraction for objects of class `string` not supported in Arrow"
  )
})

test_that("semester works with temporal types and integers", {
  test_df <- tibble(
    month_as_int = c(1:12, NA),
    month_as_char_pad = sprintf("%02i", month_as_int),
    dates = as.Date(paste0("2021-", month_as_char_pad, "-15"))
  )

  # semester extraction from dates
  compare_dplyr_binding(
    .input %>%
      mutate(
        sem_wo_year = semester(dates),
        sem_wo_year2 = lubridate::semester(dates),
        sem_w_year = semester(dates, with_year = TRUE)
      ) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(sem_month_as_int = semester(month_as_int)) %>%
      collect(),
    test_df
  )

  expect_error(
    test_df %>%
      arrow_table() %>%
      mutate(sem_month_as_char_pad = semester(month_as_char_pad)) %>%
      collect(),
    regexp = "NotImplemented: Function 'month' has no kernel matching input types (string)",
    fixed = TRUE
  )
})

test_that("dst extracts daylight savings time correctly", {
  test_df <- tibble(
    dates = as.POSIXct(c("2021-02-20", "2021-07-31", "2021-10-31", "2021-01-31"), tz = "Europe/London")
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        dst = dst(dates),
        dst2 = lubridate::dst(dates)
      ) %>%
      collect(),
    test_df
  )
})

test_that("month() supports integer input", {
  test_df_month <- tibble(
    month_as_int = c(1:12, NA)
  )

  compare_dplyr_binding(
    .input %>%
      mutate(month_int_input = month(month_as_int)) %>%
      collect(),
    test_df_month
  )

  compare_dplyr_binding(
    .input %>%
      # R returns ordered factor whereas Arrow returns character
      mutate(
        month_int_input = as.character(month(month_as_int, label = TRUE))
      ) %>%
      collect(),
    test_df_month
  )

  compare_dplyr_binding(
    .input %>%
      # R returns ordered factor whereas Arrow returns character
      mutate(
        month_int_input = as.character(
          month(month_as_int, label = TRUE, abbr = FALSE)
        )
      ) %>%
      collect(),
    test_df_month
  )
})

test_that("month() errors with double input and returns NA with int outside 1:12", {
  test_df_month <- tibble(
    month_as_int = c(-1L, 1L, 13L, NA),
    month_as_double = month_as_int + 0.1
  )

  expect_equal(
    test_df_month %>%
      arrow_table() %>%
      select(month_as_int) %>%
      mutate(month_int_input = month(month_as_int)) %>%
      collect(),
    tibble(
      month_as_int = c(-1L, 1L, 13L, NA),
      month_int_input = c(NA, 1L, NA, NA)
    )
  )

  expect_error(
    test_df_month %>%
      arrow_table() %>%
      mutate(month_dbl_input = month(month_as_double)) %>%
      collect(),
    regexp = "Function 'month' has no kernel matching input types (double)",
    fixed = TRUE
  )

  expect_error(
    test_df_month %>%
      record_batch() %>%
      mutate(month_dbl_input = month(month_as_double)) %>%
      collect(),
    regexp = "Function 'month' has no kernel matching input types (double)",
    fixed = TRUE
  )
})

test_that("date works in arrow", {
  # this date is specific since lubridate::date() is different from base::as.Date()
  # since as.Date returns the UTC date and date() doesn't
  test_df <- tibble(
    posixct_date = as.POSIXct(c("2012-03-26 23:12:13", NA), tz = "America/New_York"),
    posixct_fractional_second = as_datetime(c("2012-03-26 23:12:13.676632", NA)),
    integer_var = c(32L, NA)
  )

  r_date_object <- lubridate::ymd_hms("2012-03-26 23:12:13")

  compare_dplyr_binding(
    .input %>%
      mutate(a_date = lubridate::date(posixct_date)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(a_date_base = as.Date(posixct_date)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(a_date_base = as.Date(posixct_fractional_second)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(date_from_r_object = lubridate::date(r_date_object)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(as_date_from_r_object = as.Date(r_date_object)) %>%
      collect(),
    test_df
  )

  # date from integer supported in arrow (similar to base::as.Date()), but in
  # Arrow it assumes a fixed origin "1970-01-01". However this is not supported
  # by lubridate. lubridate::date(integer_var) errors without an `origin`
  expect_equal(
    test_df %>%
      arrow_table() %>%
      select(integer_var) %>%
      mutate(date_int = date(integer_var)) %>%
      collect(),
    tibble(
      integer_var = c(32L, NA),
      date_int = as.Date(c("1970-02-02", NA))
    )
  )
})

test_that("date() errors with unsupported inputs", {
  # Use InMemoryDataset here so that abandon_ship() errors instead of warns.
  # The lubridate version errors too.
  skip_if_not_available("dataset")
  expect_error(
    example_data %>%
      InMemoryDataset$create() %>%
      mutate(date_bool = lubridate::date(TRUE)) %>%
      collect(),
    regexp = "Unsupported cast from bool to date32 using function cast_date32"
  )
})

test_that("make_date & make_datetime", {
  test_df <- expand.grid(
    year = c(1999, 1969, 2069, NA),
    month = c(1, 2, 7, 12, NA),
    day = c(1, 9, 13, 28, NA),
    hour = c(0, 7, 23, NA),
    min = c(0, 59, NA),
    sec = c(0, 59, NA)
  ) %>%
    tibble()

  compare_dplyr_binding(
    .input %>%
      mutate(
        composed_date = make_date(year, month, day),
        composed_date2 = lubridate::make_date(year, month, day)
      ) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(composed_date_r_obj = make_date(1999, 12, 31)) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        composed_datetime = make_datetime(year, month, day, hour, min, sec),
        composed_datetime2 = lubridate::make_datetime(year, month, day, hour, min, sec)
      ) %>%
      collect(),
    test_df,
    # the make_datetime binding uses strptime which does not support tz, hence
    # a mismatch in tzone attribute (ARROW-12820)
    ignore_attr = TRUE
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        composed_datetime_r_obj = make_datetime(1999, 12, 31, 14, 15, 16)
      ) %>%
      collect(),
    test_df,
    # the make_datetime binding uses strptime which does not support tz, hence
    # a mismatch in tzone attribute (ARROW-12820)
    ignore_attr = TRUE
  )
})

test_that("ISO_datetime & ISOdate", {
  test_df <- expand.grid(
    year = c(1999, 1969, 2069, NA),
    month = c(1, 2, 7, 12, NA),
    day = c(1, 9, 13, 28, NA),
    hour = c(0, 7, 23, NA),
    min = c(0, 59, NA),
    sec = c(0, 59, NA)
  ) %>%
    tibble()

  compare_dplyr_binding(
    .input %>%
      mutate(
        composed_date = ISOdate(year, month, day),
        composed_date2 = base::ISOdate(year, month, day)
      ) %>%
      collect(),
    test_df,
    # the make_datetime binding uses strptime which does not support tz, hence
    # a mismatch in tzone attribute (ARROW-12820)
    ignore_attr = TRUE
  )

  compare_dplyr_binding(
    .input %>%
      mutate(composed_date_r_obj = ISOdate(1999, 12, 31)) %>%
      collect(),
    test_df,
    # the make_datetime binding uses strptime which does not support tz, hence
    # a mismatch in tzone attribute (ARROW-12820)
    ignore_attr = TRUE
  )

  # the default `tz` for base::ISOdatetime is "", but in Arrow it's "UTC"
  compare_dplyr_binding(
    .input %>%
      mutate(
        composed_datetime = ISOdatetime(year, month, day, hour, min, sec, tz = "UTC"),
        composed_datetime2 = base::ISOdatetime(year, month, day, hour, min, sec, tz = "UTC")
      ) %>%
      collect(),
    test_df,
    # the make_datetime binding uses strptime which does not support tz, hence
    # a mismatch in tzone attribute (ARROW-12820)
    ignore_attr = TRUE
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        composed_datetime_r_obj = ISOdatetime(1999, 12, 31, 14, 15, 16)
      ) %>%
      collect(),
    test_df,
    # the make_datetime binding uses strptime which does not support tz, hence
    # a mismatch in tzone attribute (ARROW-12820)
    ignore_attr = TRUE
  )
})

test_that("difftime()", {
  test_df <- tibble(
    time1 = as.POSIXct(
      c("2021-02-20", "2021-07-31 0:0:0", "2021-10-30", "2021-01-31 0:0:0")
    ),
    time2 = as.POSIXct(
      c("2021-02-20 00:02:01", "2021-07-31 00:03:54", "2021-10-30 00:05:45", "2021-01-31 00:07:36")
    ),
    secs = c(121L, 234L, 345L, 456L)
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        secs = difftime(time1, time2, units = "secs"),
        secs2 = base::difftime(time1, time2, units = "secs")
      ) %>%
      collect(),
    test_df,
    ignore_attr = TRUE
  )

  # units other than "secs" not supported in arrow
  compare_dplyr_binding(
    .input %>%
      mutate(
        mins = difftime(time1, time2, units = "mins")
      ) %>%
      collect(),
    test_df,
    warning = TRUE,
    ignore_attr = TRUE
  )

  test_df_with_tz <- tibble(
    time1 = as.POSIXct(
      c("2021-02-20", "2021-07-31", "2021-10-30", "2021-01-31"),
      tz = "Pacific/Marquesas"
    ),
    time2 = as.POSIXct(
      c("2021-02-20 00:02:01", "2021-07-31 00:03:54", "2021-10-30 00:05:45", "2021-01-31 00:07:36"),
      tz = "Asia/Kathmandu"
    ),
    secs = c(121L, 234L, 345L, 456L)
  )

  compare_dplyr_binding(
    .input %>%
      mutate(secs2 = difftime(time2, time1, units = "secs")) %>%
      collect(),
    test_df_with_tz
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        secs2 = difftime(
          as.POSIXct("2022-03-07", tz = "Pacific/Marquesas"),
          time1,
          units = "secs"
        )
      ) %>%
      collect(),
    test_df_with_tz
  )

  # `tz` is effectively ignored both in R (used only if inputs are POSIXlt) and Arrow
  compare_dplyr_binding(
    .input %>%
      mutate(secs2 = difftime(time2, time1, units = "secs", tz = "Pacific/Marquesas")) %>%
      collect(),
    test_df_with_tz,
    warning = "`tz` argument is not supported in Arrow, so it will be ignored"
  )
})

test_that("as.difftime()", {
  test_df <- tibble(
    hms_string = c("0:7:45", "12:34:56"),
    hm_string = c("7:45", "12:34"),
    int = c(30L, 75L),
    integerish_dbl = c(31, 76),
    dbl = c(31.2, 76.4)
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        hms_difftime = as.difftime(hms_string, units = "secs"),
        hms_difftime2 = base::as.difftime(hms_string, units = "secs")
      ) %>%
      collect(),
    test_df
  )

  # TODO add test with `format` mismatch returning NA once
  # https://issues.apache.org/jira/browse/ARROW-15659 is solved
  # for example: as.difftime("07:", format = "%H:%M") should return NA
  compare_dplyr_binding(
    .input %>%
      mutate(hm_difftime = as.difftime(hm_string, units = "secs", format = "%H:%M")) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(int_difftime = as.difftime(int, units = "secs")) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(integerish_dbl_difftime = as.difftime(integerish_dbl, units = "secs")) %>%
      collect(),
    test_df
  )

  # "mins" or other values for units cannot be handled in Arrow
  compare_dplyr_binding(
    .input %>%
      mutate(int_difftime = as.difftime(int, units = "mins")) %>%
      collect(),
    test_df,
    warning = TRUE
  )

  # only integer (or integer-like) -> duration conversion supported in Arrow.
  # double -> duration not supported. We aren't testing the content of the
  # error message as it is being generated in the C++ code and it might change,
  # but we want to make sure that this error is raised in our binding implementation
  expect_error(
    test_df %>%
      arrow_table() %>%
      mutate(dbl_difftime = as.difftime(dbl, units = "secs")) %>%
      collect()
  )
})

test_that("`decimal_date()` and `date_decimal()`", {
  test_df <- tibble(
    a = c(
      2007.38998954347, 1970.77732069883, 2020.96061799722,
      2009.43465948477, 1975.71251467871, NA
    ),
    b = as.POSIXct(
      c(
        "2007-05-23 08:18:30", "1970-10-11 17:19:45", "2020-12-17 14:04:06",
        "2009-06-08 15:37:01", "1975-09-18 01:37:42", NA
      )
    ),
    c = as.Date(
      c("2007-05-23", "1970-10-11", "2020-12-17", "2009-06-08", "1975-09-18", NA)
    )
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        decimal_date_from_POSIXct = decimal_date(b),
        decimal_date_from_POSIXct2 = lubridate::decimal_date(b),
        decimal_date_from_r_POSIXct_obj = decimal_date(as.POSIXct("2022-03-25 15:37:01")),
        decimal_date_from_r_date_obj = decimal_date(as.Date("2022-03-25")),
        decimal_date_from_date = decimal_date(c),
        date_from_decimal = date_decimal(a),
        date_from_decimal2 = lubridate::date_decimal(a),
        date_from_decimal_r_obj = date_decimal(2022.178)
      ) %>%
      collect(),
    test_df,
    ignore_attr = "tzone"
  )
})

test_that("dminutes, dhours, ddays, dweeks, dmonths, dyears", {
  example_d <- tibble(x = c(1:10, NA))
  date_to_add <- ymd("2009-08-03", tz = "Pacific/Marquesas")

  # When comparing results we use ignore_attr = TRUE because of the diff in:
  # attribute 'package' (absent vs. 'lubridate')
  # class (difftime vs Duration)
  # attribute 'units' (character vector ('secs') vs. absent)

  compare_dplyr_binding(
    .input %>%
      mutate(
        dminutes = dminutes(x),
        dhours = dhours(x),
        ddays = ddays(x),
        dweeks = dweeks(x),
        dmonths = dmonths(x),
        dyears = dyears(x)
      ) %>%
      collect(),
    example_d,
    ignore_attr = TRUE
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        dhours = dhours(x),
        ddays = ddays(x),
        new_date_1 = date_to_add + ddays,
        new_date_2 = date_to_add + ddays - dhours(3),
        new_duration = dhours - ddays
      ) %>%
      collect(),
    example_d,
    ignore_attr = TRUE
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        r_obj_dminutes = dminutes(1),
        r_obj_dhours = dhours(2),
        r_obj_ddays = ddays(3),
        r_obj_dweeks = dweeks(4),
        r_obj_dmonths = dmonths(5),
        r_obj_dyears = dyears(6),
        r_obj_dminutes2 = lubridate::dminutes(1),
        r_obj_dhours2 = lubridate::dhours(2),
        r_obj_ddays2 = lubridate::ddays(3),
        r_obj_dweeks2 = lubridate::dweeks(4),
        r_obj_dmonths2 = lubridate::dmonths(5),
        r_obj_dyears2 = lubridate::dyears(6)
      ) %>%
      collect(),
    tibble(),
    ignore_attr = TRUE
  )

  # double -> duration not supported in Arrow.
  # With a scalar, cast to int64 error in mutate() -> abandon_ship warning
  expect_warning(
    test_df %>%
      arrow_table() %>%
      mutate(r_obj_dminutes = dminutes(1.12345)),
    "not supported in Arrow"
  )

  # When operating on a column, it doesn't happen until collect()
  expect_error(
    arrow_table(dbl = 1.948230) %>%
      mutate(r_obj_dminutes = dminutes(dbl)) %>%
      collect(),
    "truncated converting to int64"
  )
})

test_that("dseconds, dmilliseconds, dmicroseconds, dnanoseconds, dpicoseconds", {
  example_d <- tibble(x = c(1:10, NA))
  date_to_add <- ymd("2009-08-03", tz = "Pacific/Marquesas")

  # When comparing results we use ignore_attr = TRUE because of the diff in:
  # attribute 'package' (absent vs. 'lubridate')
  # class (difftime vs Duration)
  # attribute 'units' (character vector ('secs') vs. absent)

  compare_dplyr_binding(
    .input %>%
      mutate(
        dseconds = dseconds(x),
        dmilliseconds = dmilliseconds(x),
        dmicroseconds = dmicroseconds(x),
        dnanoseconds = dnanoseconds(x),
        dseconds2 = lubridate::dseconds(x),
        dmilliseconds2 = lubridate::dmilliseconds(x),
        dmicroseconds2 = lubridate::dmicroseconds(x),
        dnanoseconds2 = lubridate::dnanoseconds(x),
      ) %>%
      collect(),
    example_d,
    ignore_attr = TRUE
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        dseconds = dseconds(x),
        dmicroseconds = dmicroseconds(x),
        new_date_1 = date_to_add + dseconds,
        new_date_2 = date_to_add + dseconds - dmicroseconds,
        new_duration = dseconds - dmicroseconds
      ) %>%
      collect(),
    example_d,
    ignore_attr = TRUE
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        r_obj_dseconds = dseconds(1),
        r_obj_dmilliseconds = dmilliseconds(2),
        r_obj_dmicroseconds = dmicroseconds(3),
        r_obj_dnanoseconds = dnanoseconds(4)
      ) %>%
      collect(),
    tibble(),
    ignore_attr = TRUE
  )

  expect_error(
    call_binding("dpicoseconds"),
    "Duration in picoseconds not supported in Arrow"
  )

  expect_error(
    call_binding("lubridate::dpicoseconds"),
    "Duration in picoseconds not supported in Arrow"
  )
})

test_that("make_difftime()", {
  test_df <- tibble(
    seconds = c(3, 4, 5, 6),
    minutes = c(1.5, 2.3, 4.5, 6.7),
    hours = c(2, 3, 4, 5),
    days = c(6, 7, 8, 9),
    weeks = c(1, 3, 5, NA),
    number = 10:13
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        duration_from_parts = make_difftime(
          second = seconds,
          minute = minutes,
          hour = hours,
          day = days,
          week = weeks,
          units = "secs"
        ),
        duration_from_num = make_difftime(
          num = number,
          units = "secs"
        ),
        duration_from_r_num = make_difftime(
          num = 154,
          units = "secs"
        ),
        duration_from_r_parts = make_difftime(
          minute = 45,
          day = 2,
          week = 4,
          units = "secs"
        ),
        duration_from_parts2 = lubridate::make_difftime(
          second = seconds,
          minute = minutes,
          hour = hours,
          day = days,
          week = weeks,
          units = "secs"
        )
      ) %>%
      collect(),
    test_df
  )

  # named difftime parts other than `second`, `minute`, `hour`, `day` and `week`
  # are not supported
  expect_error(
    expect_warning(
      test_df %>%
        arrow_table() %>%
        mutate(
          err_difftime = make_difftime(month = 2)
        ) %>%
        collect(),
      paste0(
        "named `difftime` units other than: `second`, `minute`, `hour`,",
        " `day`, and `week` not supported in Arrow."
      )
    )
  )

  # units other than "secs" not supported since they are the only ones in common
  # between R and Arrow
  compare_dplyr_binding(
    .input %>%
      mutate(error_difftime = make_difftime(num = number, units = "mins")) %>%
      collect(),
    test_df,
    warning = TRUE
  )

  # constructing a difftime from both `num` and parts passed through `...` while
  # possible with the lubridate function (resulting in a concatenation of the 2
  # resulting objects), it errors in a dplyr context
  expect_error(
    expect_warning(
      test_df %>%
        arrow_table() %>%
        mutate(
          duration_from_num_and_parts = make_difftime(
            num = number,
            second = seconds,
            minute = minutes,
            hour = hours,
            day = days,
            week = weeks,
            units = "secs"
          )
        ) %>%
        collect(),
      "with both `num` and `...` not supported in Arrow"
    )
  )
})

test_that("`as.Date()` and `as_date()`", {
  test_df <- tibble::tibble(
    posixct_var = as.POSIXct(c("2022-02-25 00:00:01", "1987-11-24 12:34:56", NA), tz = "Pacific/Marquesas"),
    dt_europe = ymd_hms("2010-08-03 00:50:50", "1987-11-24 12:34:56", NA, tz = "Europe/London"),
    dt_utc = ymd_hms("2010-08-03 00:50:50", "1987-11-24 12:34:56", NA),
    date_var = as.Date(c("2022-02-25", "1987-11-24", NA)),
    difference_date = ymd_hms("2010-08-03 00:50:50", "1987-11-24 12:34:56", NA, tz = "Pacific/Marquesas"),
    try_formats_string = c(NA, "2022-01-01", "2022/01/01"),
    character_ymd_hms_var = c("2022-02-25 00:00:01", "1987-11-24 12:34:56", NA),
    character_ydm_hms_var = c("2022/25/02 00:00:01", "1987/24/11 12:34:56", NA),
    character_ymd_var = c("2022-02-25", "1987-11-24", NA),
    character_ydm_var = c("2022/25/02", "1987/24/11", NA),
    integer_var = c(21L, 32L, NA),
    integerish_var = c(21, 32, NA),
    double_var = c(12.34, 56.78, NA)
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        date_dv1 = as.Date(date_var),
        date_dv1_nmspc = base::as.Date(date_var),
        date_pv1 = as.Date(posixct_var),
        date_pv_tz1 = as.Date(posixct_var, tz = "Pacific/Marquesas"),
        date_utc1 = as.Date(dt_utc),
        date_europe1 = as.Date(dt_europe),
        date_char_ymd_hms1 = as.Date(character_ymd_hms_var, format = "%Y-%m-%d %H:%M:%S"),
        date_char_ydm_hms1 = as.Date(character_ydm_hms_var, format = "%Y/%d/%m %H:%M:%S"),
        date_int1 = as.Date(integer_var, origin = "1970-01-01"),
        date_int_origin1 = as.Date(integer_var, origin = "1970-01-03"),
        date_integerish1 = as.Date(integerish_var, origin = "1970-01-01"),
        date_dv2 = as_date(date_var),
        date_dv2_nmspc = lubridate::as_date(date_var),
        date_pv2 = as_date(posixct_var),
        date_pv_tz2 = as_date(posixct_var, tz = "Pacific/Marquesas"),
        date_utc2 = as_date(dt_utc),
        date_europe2 = as_date(dt_europe),
        date_char_ymd2 = as_date(character_ymd_hms_var, format = "%Y-%m-%d %H:%M:%S"),
        date_char_ydm2 = as_date(character_ydm_hms_var, format = "%Y/%d/%m %H:%M:%S"),
        date_int2 = as_date(integer_var, origin = "1970-01-01"),
        date_int_origin2 = as_date(integer_var, origin = "1970-01-03"),
        date_integerish2 = as_date(integerish_var, origin = "1970-01-01")
      ) %>%
      collect(),
    test_df
  )

  # strptime does not support a partial format - Arrow returns NA, while
  # lubridate parses correctly
  # TODO: revisit after ARROW-15813
  expect_error(
    expect_equal(
      test_df %>%
        arrow_table() %>%
        mutate(date_char_ymd_hms = as_date(character_ymd_hms_var)) %>%
        collect(),
      test_df %>%
        mutate(date_char_ymd_hms = as_date(character_ymd_hms_var)) %>%
        collect()
    )
  )

  # same as above
  expect_error(
    expect_equal(
      test_df %>%
        arrow_table() %>%
        mutate(date_char_ymd_hms = as.Date(character_ymd_hms_var)) %>%
        collect(),
      test_df %>%
        mutate(date_char_ymd_hms = as.Date(character_ymd_hms_var)) %>%
        collect()
    )
  )

  # we do not support as.Date() with double/ float (error surfaced from C++)
  # TODO: revisit after ARROW-15798
  expect_error(
    test_df %>%
      arrow_table() %>%
      mutate(date_double = as.Date(double_var, origin = "1970-01-01")) %>%
      collect()
  )
  expect_error(
    test_df %>%
      arrow_table() %>%
      mutate(date_double = as_date(double_var, origin = "1970-01-01")) %>%
      collect()
  )

  # difference between as.Date() and as_date():
  # `as.Date()` ignores the `tzone` attribute and uses the value of the `tz` arg
  # to `as.Date()`
  # `as_date()` does the opposite: uses the tzone attribute of the POSIXct object
  # passed if`tz` is NULL
  compare_dplyr_binding(
    .input %>%
      transmute(
        date_diff_lubridate = as_date(difference_date),
        date_diff_base = as.Date(difference_date)
      ) %>%
      collect(),
    test_df
  )

  skip_if_not_available("dataset")
  # we do not support multiple tryFormats
  # Use a dataset to test the alternative suggestion message
  expect_snapshot(
    test_df %>%
      InMemoryDataset$create() %>%
      transmute(
        date_char_ymd = as.Date(
          character_ymd_var,
          tryFormats = c("%Y-%m-%d", "%Y/%m/%d")
        )
      ) %>%
      collect(),
    error = TRUE
  )
})

test_that("`as_date()` and `as.Date()` work with R objects", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        date1 = as.Date("2022-05-10"),
        date2 = as.Date(12, origin = "2022-05-01"),
        date3 = as.Date("2022-10-03", tryFormats = "%Y-%m-%d"),
        date4 = as_date("2022-05-10"),
        date5 = as_date(12, origin = "2022-05-01"),
        date6 = as_date("2022-10-03")
      ) %>%
      collect(),
    tibble(
      a = 1
    )
  )
})

test_that("`as_datetime()`", {
  test_df <- tibble(
    date = as.Date(c("2022-03-22", "2021-07-30", NA)),
    char_date = c("2022-03-22", "2021-07-30 14:32:47", NA),
    char_date_subsec = c("1970-01-01T00:00:59.123456789", "2000-02-29T23:23:23.999999999", NA),
    char_date_non_iso = c("2022-22-03 12:34:56", "2021-30-07 14:32:47", NA),
    int_date = c(10L, 25L, NA),
    integerish_date = c(10, 25, NA),
    double_date = c(10.1, 25.2, NA)
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        ddate = as_datetime(date),
        ddate2 = lubridate::as_datetime(date),
        dchar_date_no_tz = as_datetime(char_date),
        dchar_date_with_tz = as_datetime(char_date, tz = "Pacific/Marquesas"),
        dchar_date_subsec_no_tz = as_datetime(char_date_subsec),
        dchar_date_subsec_with_tz = as_datetime(char_date_subsec, tz = "Pacific/Marquesas"),
        dint_date = as_datetime(int_date, origin = "1970-01-02"),
        dintegerish_date = as_datetime(integerish_date, origin = "1970-01-02"),
        dintegerish_date2 = as_datetime(integerish_date, origin = "1970-01-01"),
        ddouble_date = as_datetime(double_date)
      ) %>%
      collect(),
    test_df
  )

  expect_identical(
    test_df %>%
      arrow_table() %>%
      mutate(
        x = cast(as_datetime(double_date, unit = "ns"), int64()),
        y = cast(as_datetime(double_date, unit = "us"), int64()),
        z = cast(as_datetime(double_date, unit = "ms"), int64()),
        .keep = "none"
      ) %>%
      collect(),
    tibble(
      x = bit64::as.integer64(c(10100000000, 25200000000, NA)),
      y = as.integer(c(10100000, 25200000, NA)),
      z = as.integer(c(10100, 25200, NA))
    )
  )
})

test_that("as_datetime() works with other functions", {
  test_df <- tibble(
    char_date = c("2022-03-22", "2021-07-30 14:32:47", "1970-01-01 00:00:59.123456789", NA)
  )

  compare_dplyr_binding(
    .input %>%
      transmute(
        ddchar_date = as_datetime(char_date),
        ddchar_date_date32_1 = as.Date(ddchar_date),
        ddchar_date_date32_2 = as_date(ddchar_date),
        ddchar_date_floored = floor_date(ddchar_date, unit = "days")
      ) %>%
      collect(),
    test_df
  )

  # ARROW-17428 - Arrow does not support conversion of timestamp to int32
  expect_error(
    test_df %>%
      arrow_table() %>%
      mutate(
        dchar_date = as_datetime(char_date),
        dchar_date_int = as.integer(dchar_date)
      ) %>%
      collect()
  )

  # ARROW-17428 - Arrow does not support conversion of timestamp to double
  expect_error(
    test_df %>%
      arrow_table() %>%
      mutate(
        dchar_date = as_datetime(char_date),
        dchar_date_num = as.numeric(dchar_date)
      ) %>%
      collect()
  )
})

test_that("parse_date_time() works with year, month, and date components", {
  # these functions' internals use some string processing which requires the
  # RE2 library
  skip_if_not_available("re2")
  compare_dplyr_binding(
    .input %>%
      mutate(
        parsed_date_ymd = parse_date_time(string_ymd, orders = "ymd"),
        parsed_date_ymd2 = lubridate::parse_date_time(string_ymd, orders = "ymd"),
        parsed_date_dmy = parse_date_time(string_dmy, orders = "dmy"),
        parsed_date_mdy = parse_date_time(string_mdy, orders = "mdy")
      ) %>%
      collect(),
    tibble::tibble(
      string_ymd = c(
        "2021-09-1", "2021/09///2", "2021.09.03", "2021,09,4", "2021:09::5",
        "2021 09   6", "21-09-07", "21/09/08", "21.09.9", "21,09,10", "21:09:11",
        "20210912", "210913", NA
      ),
      string_dmy = c(
        "1-09-2021", "2/09//2021", "03.09.2021", "04,09,2021", "5:::09:2021",
        "6  09  2021", "07-09-21", "08/09/21", "9.09.21", "10,09,21", "11:09:21",
        "12092021", "130921", NA
      ),
      string_mdy = c(
        "09-01-2021", "09/2/2021", "09.3.2021", "09,04,2021", "09:05:2021",
        "09 6 2021", "09-7-21", "09/08/21", "09.9.21", "09,10,21", "09:11:21",
        "09122021", "091321", NA
      )
    )
  )

  # TODO(ARROW-16443): locale (affecting "%b% and "%B") does not work on Windows
  skip_on_os("windows")
  compare_dplyr_binding(
    .input %>%
      mutate(
        parsed_date_ymd = parse_date_time(string_ymd, orders = "ymd"),
        parsed_date_dmy = parse_date_time(string_dmy, orders = "dmy"),
        parsed_date_mdy = parse_date_time(string_mdy, orders = "mdy")
      ) %>%
      collect(),
    tibble::tibble(
      string_ymd = c(
        "2021 Sep 12", "2021 September 13", "21 Sep 14", "21 September 15",
        "2021Sep16", NA
      ),
      string_dmy = c(
        "12 Sep 2021", "13 September 2021", "14 Sep 21", "15 September 21",
        "16Sep2021", NA
      ),
      string_mdy = c(
        "Sep 12 2021", "September 13 2021", "Sep 14 21", "September 15 21",
        "Sep1621", NA
      )
    )
  )
})

test_that("parse_date_time() works with a mix of formats and orders", {
  # these functions' internals use some string processing which requires the
  # RE2 library
  skip_if_not_available("re2")
  test_df <- tibble(
    string_combi = c("2021-09-1", "2/09//2021", "09.3.2021")
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        date_from_string = parse_date_time(
          string_combi,
          orders = c("ymd", "%d/%m//%Y", "%m.%d.%Y")
        )
      ) %>%
      collect(),
    test_df
  )
})

test_that("year, month, day date/time parsers", {
  test_df <- tibble::tibble(
    ymd_string = c("2022-05-11", "2022/05/12", "22.05-13"),
    ydm_string = c("2022-11-05", "2022/12/05", "22.13-05"),
    mdy_string = c("05-11-2022", "05/12/2022", "05.13-22"),
    myd_string = c("05-2022-11", "05/2022/12", "05.22-14"),
    dmy_string = c("11-05-2022", "12/05/2022", "13.05-22"),
    dym_string = c("11-2022-05", "12/2022/05", "13.22-05")
  )

  # these functions' internals use some string processing which requires the
  # RE2 library
  skip_if_not_available("re2")
  compare_dplyr_binding(
    .input %>%
      mutate(
        ymd_date = ymd(ymd_string),
        ydm_date = ydm(ydm_string),
        mdy_date = mdy(mdy_string),
        myd_date = myd(myd_string),
        dmy_date = dmy(dmy_string),
        dym_date = dym(dym_string),
        ymd_date2 = lubridate::ymd(ymd_string),
        ydm_date2 = lubridate::ydm(ydm_string),
        mdy_date2 = lubridate::mdy(mdy_string),
        myd_date2 = lubridate::myd(myd_string),
        dmy_date2 = lubridate::dmy(dmy_string),
        dym_date2 = lubridate::dym(dym_string)
      ) %>%
      collect(),
    test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        ymd_date = ymd(ymd_string, tz = "Pacific/Marquesas"),
        ydm_date = ydm(ydm_string, tz = "Pacific/Marquesas"),
        mdy_date = mdy(mdy_string, tz = "Pacific/Marquesas"),
        myd_date = myd(myd_string, tz = "Pacific/Marquesas"),
        dmy_date = dmy(dmy_string, tz = "Pacific/Marquesas"),
        dym_date = dym(dym_string, tz = "Pacific/Marquesas")
      ) %>%
      collect(),
    test_df
  )
})

test_that("ym, my & yq parsers", {
  test_df <- tibble::tibble(
    ym_string = c("2022-05", "2022/02", "22.3", "1979//12", "88.09", NA),
    my_string = c("05-2022", "02/2022", "03.22", "12//1979", "09.88", NA),
    Ym_string = c("2022-05", "2022/02", "2022.03", "1979//12", "1988.09", NA),
    mY_string = c("05-2022", "02/2022", "03.2022", "12//1979", "09.1988", NA),
    yq_string = c("2007.3", "1971.2", "2021.1", "2009.4", "1975.1", NA),
    yq_numeric = c(2007.3, 1971.2, 2021.1, 2009.4, 1975.1, NA),
    yq_space = c("2007 3", "1970 2", "2020 1", "2009 4", "1975 1", NA),
    qy_string = c("3.2007", "2.1971", "1.2020", "4.2009", "1.1975", NA),
    qy_numeric = c(3.2007, 2.1971, 1.2021, 4.2009, 1.1975, NA),
    qy_space = c("3 2007", "2 1971", "1 2021", "4 2009", "1 1975", NA)
  )

  # these functions' internals use some string processing which requires the
  # RE2 library
  skip_if_not_available("re2")
  compare_dplyr_binding(
    .input %>%
      mutate(
        ym_date = ym(ym_string),
        ym_date2 = lubridate::ym(ym_string),
        ym_datetime = ym(ym_string, tz = "Pacific/Marquesas"),
        Ym_date = ym(Ym_string),
        Ym_datetime = ym(Ym_string, tz = "Pacific/Marquesas"),
        my_date = my(my_string),
        my_date2 = lubridate::my(my_string),
        my_datetime = my(my_string, tz = "Pacific/Marquesas"),
        mY_date = my(mY_string),
        mY_datetime = my(mY_string, tz = "Pacific/Marquesas"),
        yq_date_from_string = yq(yq_string),
        yq_date_from_string2 = lubridate::yq(yq_string),
        yq_datetime_from_string = yq(yq_string, tz = "Pacific/Marquesas"),
        yq_date_from_numeric = yq(yq_numeric),
        yq_datetime_from_numeric = yq(yq_numeric, tz = "Pacific/Marquesas"),
        yq_date_from_string_with_space = yq(yq_space),
        yq_datetime_from_string_with_space = yq(yq_space, tz = "Pacific/Marquesas"),
        ym_date2 = parse_date_time(ym_string, orders = c("ym", "ymd")),
        my_date2 = parse_date_time(my_string, orders = c("my", "myd")),
        Ym_date2 = parse_date_time(Ym_string, orders = c("Ym", "ymd")),
        mY_date2 = parse_date_time(mY_string, orders = c("mY", "myd")),
        yq_date_from_string2 = parse_date_time(yq_string, orders = "yq"),
        yq_date_from_numeric2 = parse_date_time(yq_numeric, orders = "yq"),
        yq_date_from_string_with_space2 = parse_date_time(yq_space, orders = "yq"),
        # testing with Yq
        yq_date_from_string3 = parse_date_time(yq_string, orders = "Yq"),
        yq_date_from_numeric3 = parse_date_time(yq_numeric, orders = "Yq"),
        yq_date_from_string_with_space3 = parse_date_time(yq_space, orders = "Yq"),
        # testing with qy
        qy_date_from_string = parse_date_time(qy_string, orders = "qy"),
        qy_date_from_numeric = parse_date_time(qy_numeric, orders = "qy"),
        qy_date_from_string_with_space = parse_date_time(qy_space, orders = "qy"),
        # testing with qY
        qy_date_from_string2 = parse_date_time(qy_string, orders = "qY"),
        qy_date_from_numeric2 = parse_date_time(qy_numeric, orders = "qY"),
        qy_date_from_string_with_space2 = parse_date_time(qy_space, orders = "qY")
      ) %>%
      collect(),
    test_df
  )
})

test_that("parse_date_time's other formats", {
  # these functions' internals use some string processing which requires the
  # RE2 library
  skip_if_not_available("re2")

  compare_dplyr_binding(
    .input %>%
      mutate(
        parsed_H = parse_date_time(string_H, orders = "%Y-%m-%d-%H"),
        parsed_I = parse_date_time(string_I, orders = "%Y-%m-%d-%I"),
        parsed_j = parse_date_time(string_j, orders = "%Y-%m-%d-%j"),
        parsed_M = parse_date_time(string_M, orders = "%Y-%m-%d-%M"),
        parsed_S = parse_date_time(string_S, orders = "%Y-%m-%d-%S"),
        parsed_U = parse_date_time(string_U, orders = "%Y-%m-%d-%U"),
        parsed_w = parse_date_time(string_w, orders = "%Y-%m-%d-%w"),
        parsed_W = parse_date_time(string_W, orders = "%Y-%m-%d-%W"),
        parsed_y = parse_date_time(string_y, orders = "%y-%m-%d"),
        parsed_Y = parse_date_time(string_Y, orders = "%Y-%m-%d"),
        parsed_R = parse_date_time(string_R, orders = "%Y-%m-%d-%R"),
        parsed_T = parse_date_time(string_T, orders = "%Y-%m-%d-%T")
      ) %>%
      collect(),
    strptime_test_df
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        parsed_H = parse_date_time(string_H, orders = "ymdH"),
        parsed_I = parse_date_time(string_I, orders = "ymdI"),
        parsed_j = parse_date_time(string_j, orders = "ymdj"),
        parsed_M = parse_date_time(string_M, orders = "ymdM"),
        parsed_S = parse_date_time(string_S, orders = "ymdS"),
        parsed_U = parse_date_time(string_U, orders = "ymdU"),
        parsed_w = parse_date_time(string_w, orders = "ymdw"),
        parsed_W = parse_date_time(string_W, orders = "ymdW"),
        parsed_y = parse_date_time(string_y, orders = "ymd"),
        parsed_Y = parse_date_time(string_Y, orders = "Ymd"),
        parsed_R = parse_date_time(string_R, orders = "ymdR"),
        parsed_T = parse_date_time(string_T, orders = "ymdT")
      ) %>%
      collect(),
    strptime_test_df
  )

  # Some formats are not supported on Windows
  if (!tolower(Sys.info()[["sysname"]]) == "windows") {
    compare_dplyr_binding(
      .input %>%
        mutate(
          parsed_a = parse_date_time(string_a, orders = "%Y-%m-%d-%a"),
          parsed_A = parse_date_time(string_A, orders = "%Y-%m-%d-%A"),
          parsed_b = parse_date_time(string_b, orders = "%Y-%m-%d-%b"),
          parsed_B = parse_date_time(string_B, orders = "%Y-%m-%d-%B"),
          parsed_p = parse_date_time(string_p, orders = "%Y-%m-%d-%p"),
          parsed_r = parse_date_time(string_r, orders = "%Y-%m-%d-%r")
        ) %>%
        collect(),
      strptime_test_df
    )

    compare_dplyr_binding(
      .input %>%
        mutate(
          parsed_a = parse_date_time(string_a, orders = "ymda"),
          parsed_A = parse_date_time(string_A, orders = "ymdA"),
          parsed_b = parse_date_time(string_b, orders = "ymdb"),
          parsed_B = parse_date_time(string_B, orders = "ymdB"),
          parsed_p = parse_date_time(string_p, orders = "ymdp"),
          parsed_r = parse_date_time(string_r, orders = "ymdr")
        ) %>%
        collect(),
      strptime_test_df
    )

    compare_dplyr_binding(
      .input %>%
        mutate(
          parsed_date_ymd = parse_date_time(string_1, orders = "Y-%b-d-%T")
        ) %>%
        collect(),
      tibble::tibble(string_1 = c("2022-Feb-11-12:23:45", NA))
    )
  }
})

test_that("lubridate's fast_strptime", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        y = fast_strptime(x, format = "%Y-%m-%d %H:%M:%S", lt = FALSE),
        y2 = lubridate::fast_strptime(x, format = "%Y-%m-%d %H:%M:%S", lt = FALSE)
      ) %>%
      collect(),
    tibble(
      x = c("2018-10-07 19:04:05", "2022-05-17 21:23:45", NA)
    )
  )

  # R object
  compare_dplyr_binding(
    .input %>%
      mutate(
        y =
          fast_strptime(
            "68-10-07 19:04:05",
            format = "%y-%m-%d %H:%M:%S",
            lt = FALSE
          )
      ) %>%
      collect(),
    tibble(
      x = c("2018-10-07 19:04:05", NA)
    )
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        date_multi_formats =
          fast_strptime(
            x,
            format = c("%Y-%m-%d %H:%M:%S", "%m-%d-%Y %H:%M:%S"),
            lt = FALSE
          )
      ) %>%
      collect(),
    tibble(
      x = c("2018-10-07 19:04:05", "10-07-1968 19:04:05")
    )
  )

  # these functions' internals use some string processing which requires the
  # RE2 library
  skip_if_not_available("re2")

  compare_dplyr_binding(
    .input %>%
      mutate(
        dttm_with_tz = fast_strptime(
          dttm_as_string,
          format = "%Y-%m-%d %H:%M:%S",
          tz = "Pacific/Marquesas",
          lt = FALSE
        )
      ) %>%
      collect(),
    tibble(
      dttm_as_string =
        c("2018-10-07 19:04:05", "1969-10-07 19:04:05", NA)
    )
  )

  # fast_strptime()'s `cutoff_2000` argument is not supported, but its value is
  # implicitly set to 68L both in lubridate and in Arrow
  compare_dplyr_binding(
    .input %>%
      mutate(
        date_short_year =
          fast_strptime(
            x,
            format = "%y-%m-%d %H:%M:%S",
            lt = FALSE
          )
      ) %>%
      collect(),
    tibble(
      x =
        c("68-10-07 19:04:05", "69-10-07 19:04:05", NA)
    )
  )

  # the arrow binding errors for a value different from 68L for `cutoff_2000`
  compare_dplyr_binding(
    .input %>%
      mutate(
        date_short_year =
          fast_strptime(
            x,
            format = "%y-%m-%d %H:%M:%S",
            lt = FALSE,
            cutoff_2000 = 69L
          )
      ) %>%
      collect(),
    tibble(
      x = c("68-10-07 19:04:05", "69-10-07 19:04:05", NA)
    ),
    warning = TRUE
  )

  # compare_dplyr_binding would not work here since lt = TRUE returns a list
  # and it also errors in regular dplyr pipelines
  expect_warning(
    tibble(
      x = c("68-10-07 19:04:05", "69-10-07 19:04:05", NA)
    ) %>%
      arrow_table() %>%
      mutate(
        date_short_year =
          fast_strptime(
            x,
            format = "%y-%m-%d %H:%M:%S",
            lt = TRUE
          )
      ) %>%
      collect()
  )
})

test_that("parse_date_time with hours, minutes and seconds components", {
  test_dates_times <- tibble(
    ymd_hms_string =
      c("67-01-09 12:34:56", "1970-05-22 20:13:59", "870822201359", NA),
    ymd_hm_string =
      c("67-01-09 12:34", "1970-05-22 20:13", "8708222013", NA),
    ymd_h_string =
      c("67-01-09 12", "1970-05-22 20", "87082220", NA),
    dmy_hms_string =
      c("09-01-67 12:34:56", "22-05-1970 20:13:59", "220887201359", NA),
    dmy_hm_string =
      c("09-01-67 12:34", "22-05-1970 20:13", "2208872013", NA),
    dmy_h_string =
      c("09-01-67 12", "22-05-1970 20", "22088720", NA),
    mdy_hms_string =
      c("01-09-67 12:34:56", "05-22-1970 20:13:59", "082287201359", NA),
    mdy_hm_string =
      c("01-09-67 12:34", "05-22-1970 20:13", "0822872013", NA),
    mdy_h_string =
      c("01-09-67 12", "05-22-1970 20", "08228720", NA),
    ydm_hms_string =
      c("67-09-01 12:34:56", "1970-22-05 20:13:59", "872208201359", NA),
    ydm_hm_string =
      c("67-09-01 12:34", "1970-22-05 20:13", "8722082013", NA),
    ydm_h_string =
      c("67-09-01 12", "1970-22-05 20", "87220820", NA)
  )
  # the unseparated strings are versions of "1987-08-22 20:13:59" (with %y)

  # these functions' internals use some string processing which requires the
  # RE2 library
  skip_if_not_available("re2")

  compare_dplyr_binding(
    .input %>%
      mutate(
        ymd_hms_dttm = parse_date_time(ymd_hms_string, orders = "ymd_HMS"),
        ymd_hm_dttm  = parse_date_time(ymd_hm_string, orders = "ymd_HM"),
        ymd_h_dttm   = parse_date_time(ymd_h_string, orders = "ymd_H"),
        dmy_hms_dttm = parse_date_time(dmy_hms_string, orders = "dmy_HMS"),
        dmy_hm_dttm  = parse_date_time(dmy_hm_string, orders = "dmy_HM"),
        dmy_h_dttm   = parse_date_time(dmy_h_string, orders = "dmy_H"),
        mdy_hms_dttm = parse_date_time(mdy_hms_string, orders = "mdy_HMS"),
        mdy_hm_dttm  = parse_date_time(mdy_hm_string, orders = "mdy_HM"),
        mdy_h_dttm   = parse_date_time(mdy_h_string, orders = "mdy_H"),
        ydm_hms_dttm = parse_date_time(ydm_hms_string, orders = "ydm_HMS"),
        ydm_hm_dttm  = parse_date_time(ydm_hm_string, orders = "ydmHM"),
        ydm_h_dttm   = parse_date_time(ydm_h_string, orders = "ydmH")
      ) %>%
      collect(),
    test_dates_times
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        ymd_hms_dttm = ymd_hms(ymd_hms_string),
        ymd_hm_dttm = ymd_hm(ymd_hm_string),
        ymd_h_dttm = ymd_h(ymd_h_string),
        dmy_hms_dttm = dmy_hms(dmy_hms_string),
        dmy_hm_dttm = dmy_hm(dmy_hm_string),
        dmy_h_dttm = dmy_h(dmy_h_string),
        mdy_hms_dttm = mdy_hms(mdy_hms_string),
        mdy_hm_dttm = mdy_hm(mdy_hm_string),
        mdy_h_dttm = mdy_h(mdy_h_string),
        ydm_hms_dttm = ydm_hms(ydm_hms_string),
        ydm_hm_dttm = ydm_hm(ydm_hm_string),
        ydm_h_dttm = ydm_h(ydm_h_string)
      ) %>%
      collect(),
    test_dates_times
  )

  # parse_date_time with timezone
  pm_tz <- "Pacific/Marquesas"
  compare_dplyr_binding(
    .input %>%
      mutate(
        ymd_hms_dttm = parse_date_time(ymd_hms_string, orders = "ymd_HMS", tz = pm_tz),
        ymd_hm_dttm = parse_date_time(ymd_hm_string, orders = "ymd_HM", tz = pm_tz),
        ymd_h_dttm = parse_date_time(ymd_h_string, orders = "ymd_H", tz = pm_tz),
        dmy_hms_dttm = parse_date_time(dmy_hms_string, orders = "dmy_HMS", tz = pm_tz),
        dmy_hm_dttm = parse_date_time(dmy_hm_string, orders = "dmy_HM", tz = pm_tz),
        dmy_h_dttm = parse_date_time(dmy_h_string, orders = "dmy_H", tz = pm_tz),
        mdy_hms_dttm = parse_date_time(mdy_hms_string, orders = "mdy_HMS", tz = pm_tz),
        mdy_hm_dttm = parse_date_time(mdy_hm_string, orders = "mdy_HM", tz = pm_tz),
        mdy_h_dttm = parse_date_time(mdy_h_string, orders = "mdy_H", tz = pm_tz),
        ydm_hms_dttm = parse_date_time(ydm_hms_string, orders = "ydm_HMS", tz = pm_tz),
        ydm_hm_dttm = parse_date_time(ydm_hm_string, orders = "ydm_HM", tz = pm_tz),
        ydm_h_dttm = parse_date_time(ydm_h_string, orders = "ydm_H", tz = pm_tz)
      ) %>%
      collect(),
    test_dates_times
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        ymd_hms_dttm = ymd_hms(ymd_hms_string, tz = pm_tz),
        ymd_hm_dttm = ymd_hm(ymd_hm_string, tz = pm_tz),
        ymd_h_dttm = ymd_h(ymd_h_string, tz = pm_tz),
        dmy_hms_dttm = dmy_hms(dmy_hms_string, tz = pm_tz),
        dmy_hm_dttm = dmy_hm(dmy_hm_string, tz = pm_tz),
        dmy_h_dttm = dmy_h(dmy_h_string, tz = pm_tz),
        mdy_hms_dttm = mdy_hms(mdy_hms_string, tz = pm_tz),
        mdy_hm_dttm = mdy_hm(mdy_hm_string, tz = pm_tz),
        mdy_h_dttm = mdy_h(mdy_h_string, tz = pm_tz),
        ydm_hms_dttm = ydm_hms(ydm_hms_string, tz = pm_tz),
        ydm_hm_dttm = ydm_hm(ydm_hm_string, tz = pm_tz),
        ydm_h_dttm = ydm_h(ydm_h_string, tz = pm_tz),
      ) %>%
      collect(),
    test_dates_times
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        ymd_hms_dttm = ymd_hms("2022-07-19 20:24:43"),
        ymd_hm_dttm = ymd_hm("2022-07-19 20:24"),
        ymd_h_dttm = ymd_h("2022-07-19 20"),
        dmy_hms_dttm = dmy_hms("19-07-2022 20:24:43"),
        dmy_hm_dttm = dmy_hm("19-07-2022 20:24"),
        dmy_h_dttm = dmy_h("19-07-2022 20"),
        mdy_hms_dttm = mdy_hms("07-19-2022 20:24:43"),
        mdy_hm_dttm = mdy_hm("07-19-2022 20:24"),
        mdy_h_dttm = mdy_h("07-19-2022 20"),
        ydm_hms_dttm = ydm_hms("2022-19-07 20:24:43"),
        ydm_hm_dttm = ydm_hm("2022-19-07 20:24"),
        ydm_h_dttm = ydm_h("2022-19-07 20")
      ) %>%
      collect(),
    test_dates_times
  )

  # test ymd_ims
  compare_dplyr_binding(
    .input %>%
      mutate(
        ymd_ims_dttm =
          parse_date_time(
            ymd_ims_string,
            orders = "ymd_IMS",
            # lubridate is chatty and will warn 1 format failed to parse
            quiet = TRUE
          )
      ) %>%
      collect(),
    tibble(
      ymd_ims_string =
        c("67-01-09 9:34:56", "1970-05-22 10:13:59", "19870822171359", NA)
    )
  )
})

test_that("parse_date_time with month names and HMS", {
  # TODO(ARROW-16443): locale (affecting "%b% and "%B") does not work on Windows
  skip_on_os("windows")

  # these functions' internals use some string processing which requires the
  # RE2 library (not available in the minimal nightly builds)
  skip_if_not_available("re2")

  test_dates_times2 <- tibble(
    ymd_hms_string =
      c("67-Jan-09 12:34:56", "1970-June-22 20:13:59", "87Aug22201359", NA),
    ymd_hm_string =
      c("67-Jan-09 12:34", "1970-June-22 20:13", "87Aug222013", NA),
    ymd_h_string =
      c("67-Jan-09 12", "1970-June-22 20", "87Aug2220", NA),
    dmy_hms_string =
      c("09-Jan-67 12:34:56", "22-June-1970 20:13:59", "22Aug87201359", NA),
    dmy_hm_string =
      c("09-Jan-67 12:34", "22-June-1970 20:13", "22Aug872013", NA),
    dmy_h_string =
      c("09-Jan-67 12", "22-June-1970 20", "22Aug8720", NA),
    mdy_hms_string =
      c("Jan-09-67 12:34:56", "June-22-1970 20:13:59", "Aug2287201359", NA),
    mdy_hm_string =
      c("Jan-09-67 12:34", "June-22-1970 20:13", "Aug22872013", NA),
    mdy_h_string =
      c("Jan-09-67 12", "June-22-1970 20", "Aug228720", NA),
    ydm_hms_string =
      c("67-09-Jan 12:34:56", "1970-22-June 20:13:59", "8722Aug201359", NA),
    ydm_hm_string =
      c("67-09-Jan 12:34", "1970-22-June 20:13", "8722Aug2013", NA),
    ydm_h_string =
      c("67-09-Jan 12", "1970-22-June 20", "8722Aug20", NA)
  )
  # the un-separated strings are versions of "1987-08-22 20:13:59" (with %y)

  compare_dplyr_binding(
    .input %>%
      mutate(
        ymd_hms_dttm = parse_date_time(ymd_hms_string, orders = "ymd_HMS"),
        ymd_hm_dttm  = parse_date_time(ymd_hm_string, orders = "ymdHM"),
        ymd_h_dttm   = parse_date_time(ymd_h_string, orders = "ymd_H"),
        dmy_hms_dttm = parse_date_time(dmy_hms_string, orders = "dmy_HMS"),
        dmy_hm_dttm  = parse_date_time(dmy_hm_string, orders = "dmyHM"),
        dmy_h_dttm   = parse_date_time(dmy_h_string, orders = "dmy_H"),
        mdy_hms_dttm = parse_date_time(mdy_hms_string, orders = "mdy_HMS"),
        mdy_hm_dttm  = parse_date_time(mdy_hm_string, orders = "mdyHM"),
        mdy_h_dttm   = parse_date_time(mdy_h_string, orders = "mdy_H"),
        ydm_hms_dttm = parse_date_time(ydm_hms_string, orders = "ydm_HMS"),
        ydm_hm_dttm  = parse_date_time(ydm_hm_string, orders = "ydmHM"),
        ydm_h_dttm   = parse_date_time(ydm_h_string, orders = "ydm_H")
      ) %>%
      collect(),
    test_dates_times2
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        ymd_hms_dttm = ymd_hms(ymd_hms_string),
        ymd_hm_dttm  = ymd_hm(ymd_hm_string),
        ymd_h_dttm   = ymd_h(ymd_h_string),
        dmy_hms_dttm = dmy_hms(dmy_hms_string),
        dmy_hm_dttm  = dmy_hm(dmy_hm_string),
        dmy_h_dttm   = dmy_h(dmy_h_string),
        mdy_hms_dttm = mdy_hms(mdy_hms_string),
        mdy_hm_dttm  = mdy_hm(mdy_hm_string),
        mdy_h_dttm   = mdy_h(mdy_h_string),
        ydm_hms_dttm = ydm_hms(ydm_hms_string),
        ydm_hm_dttm  = ydm_hm(ydm_hm_string),
        ydm_h_dttm   = ydm_h(ydm_h_string)
      ) %>%
      collect(),
    test_dates_times2
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        ymd_hms_dttm = ymd_hms("2022-June-19 20:24:43"),
        ymd_hm_dttm = ymd_hm("2022-June-19 20:24"),
        ymd_h_dttm = ymd_h("2022-June-19 20"),
        dmy_hms_dttm = dmy_hms("19-June-2022 20:24:43"),
        dmy_hm_dttm = dmy_hm("19-June-2022 20:24"),
        dmy_h_dttm = dmy_h("19-June-2022 20"),
        mdy_hms_dttm = mdy_hms("June-19-2022 20:24:43"),
        mdy_hm_dttm = mdy_hm("June-19-2022 20:24"),
        mdy_h_dttm = mdy_h("June-19-2022 20"),
        ydm_hms_dttm = ydm_hms("2022-19-June 20:24:43"),
        ydm_hm_dttm = ydm_hm("2022-19-June 20:24"),
        ydm_h_dttm = ydm_h("2022-19-June 20")
      ) %>%
      collect(),
    test_dates_times2
  )
})

test_that("parse_date_time with `quiet = FALSE` not supported", {
  # we need expect_warning twice as both the arrow pipeline (because quiet =
  # FALSE is not supported) and the fallback dplyr/lubridate one throw
  # warnings (the lubridate one because quiet is FALSE)
  # https://issues.apache.org/jira/browse/ARROW-17146

  # these functions' internals use some string processing which requires the
  # RE2 library (not available in the minimal nightly builds)
  skip_if_not_available("re2")

  expect_warning(
    expect_warning(
      tibble(x = c("2022-05-19 13:46:51")) %>%
        arrow_table() %>%
        mutate(
          x_dttm = parse_date_time(x, orders = "dmy_HMS", quiet = FALSE)
        ) %>%
        collect(),
      "`quiet = FALSE` not supported in Arrow"
    ),
    "All formats failed to parse"
  )

  expect_warning(
    tibble(x = c("2022-05-19 13:46:51")) %>%
      arrow_table() %>%
      mutate(
        x_dttm = ymd_hms(x, quiet = FALSE)
      ) %>%
      collect(),
    "`quiet = FALSE` not supported in Arrow"
  )
})

test_that("parse_date_time with truncated formats", {
  # these functions' internals use some string processing which requires the
  # RE2 library
  skip_if_not_available("re2")

  test_truncation_df <- tibble(
    truncated_ymd_string =
      c(
        "2022-05-19 13:46:51",
        "2022-05-18 13:46",
        "2022-05-17 13",
        "2022-05-16"
      )
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        dttm =
          parse_date_time(
            truncated_ymd_string,
            orders = "ymd_HMS",
            truncated = 3
          ),
        dttm2 =
          ymd_hms(
            truncated_ymd_string,
            truncated = 3
          )
      ) %>%
      collect(),
    test_truncation_df
  )

  # values for truncated greater than nchar(orders) - 3 not supported in Arrow
  compare_dplyr_binding(
    .input %>%
      mutate(
        dttm =
          parse_date_time(
            truncated_ymd_string,
            orders = "ymd_HMS",
            truncated = 5
          )
      ) %>%
      collect(),
    test_truncation_df,
    warning = "a value for `truncated` > 4 not supported in Arrow"
  )

  # values for truncated greater than nchar(orders) - 3 not supported in Arrow
  compare_dplyr_binding(
    .input %>%
      mutate(
        dttm =
          ymd_hms(
            truncated_ymd_string,
            truncated = 5
          )
      ) %>%
      collect(),
    test_truncation_df,
    warning = "a value for `truncated` > 4 not supported in Arrow"
  )
})

test_that("parse_date_time with `locale != NULL` not supported", {
  # parse_date_time currently doesn't take locale parameter which will be
  # addressed in https://issues.apache.org/jira/browse/ARROW-17147
  skip_if_not_available("re2")

  expect_warning(
    tibble(x = c("2022-05-19 13:46:51")) %>%
      arrow_table() %>%
      mutate(
        x_dttm = ymd_hms(x, locale = "C")
      ) %>%
      collect(),
    "`locale` not supported in Arrow"
  )
})

test_that("parse_date_time with `exact = TRUE`, and with regular R objects", {
  test_df <- tibble(
    x = c("2022-12-31 12:59:59", "2022-01-01 12:11", "2022-01-01 12", "2022-01-01", NA),
    y = c("11/23/1998 07:00:00", "6/18/1952 0135", "2/25/1974 0523", "9/07/1985 01", NA)
  )

  # these functions' internals use some string processing which requires the
  # RE2 library
  skip_if_not_available("re2")
  compare_dplyr_binding(
    .input %>%
      mutate(
        parsed_x =
          parse_date_time(
            x,
            c("%Y-%m-%d %H:%M:%S", "%Y-%m-%d %H:%M", "%Y-%m-%d %H", "%Y-%m-%d"),
            exact = TRUE
          ),
        parsed_y =
          parse_date_time(
            y,
            c("%m/%d/%Y %I:%M:%S", "%m/%d/%Y %H%M", "%m/%d/%Y %H"),
            exact = TRUE
          )
      ) %>%
      collect(),
    test_df
  )
  compare_dplyr_binding(
    .input %>%
      mutate(
        b = parse_date_time("2022-12-31 12:59:59", orders = "ymd_HMS")
      ) %>%
      collect(),
    tibble(
      a = 1
    )
  )
})

test_that("build_formats() and build_format_from_order()", {
  ymd_formats <- c(
    "%y-%m-%d", "%Y-%m-%d", "%y-%B-%d", "%Y-%B-%d", "%y-%b-%d", "%Y-%b-%d",
    "%y%m%d", "%Y%m%d", "%y%B%d", "%Y%B%d", "%y%b%d", "%Y%b%d"
  )

  ymd_hms_formats <- c(
    "%y-%m-%d-%H-%M-%S", "%Y-%m-%d-%H-%M-%S", "%y-%B-%d-%H-%M-%S",
    "%Y-%B-%d-%H-%M-%S", "%y-%b-%d-%H-%M-%S", "%Y-%b-%d-%H-%M-%S",
    "%y%m%d%H%M%S", "%Y%m%d%H%M%S", "%y%B%d%H%M%S", "%Y%B%d%H%M%S",
    "%y%b%d%H%M%S", "%Y%b%d%H%M%S"
  )

  expect_equal(
    build_formats(c("ym", "myd", "%Y-%d-%m")),
    c(
      # formats from "ym" order
      "%y-%m-%d", "%Y-%m-%d", "%y-%B-%d", "%Y-%B-%d", "%y-%b-%d", "%Y-%b-%d",
      "%y%m%d", "%Y%m%d", "%y%B%d", "%Y%B%d", "%y%b%d", "%Y%b%d",
      # formats from "myd" order
      "%m-%y-%d", "%B-%y-%d", "%b-%y-%d", "%m-%Y-%d", "%B-%Y-%d", "%b-%Y-%d",
      "%m%y%d", "%B%y%d", "%b%y%d", "%m%Y%d", "%B%Y%d", "%b%Y%d",
      # formats from "%Y-%d-%m" format
      "%y-%d-%m", "%Y-%d-%m", "%y-%d-%B", "%Y-%d-%B", "%y-%d-%b", "%Y-%d-%b",
      "%y%d%m", "%Y%d%m", "%y%d%B", "%Y%d%B", "%y%d%b", "%Y%d%b"
    )
  )

  expect_equal(
    build_formats("ymd_HMS"),
    ymd_hms_formats
  )

  # when order is one of "yq", "qy", "ym" or "my" the data is augmented to "ymd"
  # or "ydm" and the formats are built accordingly
  expect_equal(
    build_formats("yq"),
    ymd_formats
  )

  expect_equal(
    build_formats("ym"),
    ymd_formats
  )

  expect_equal(
    build_formats("qy"),
    ymd_formats
  )

  # build formats will output unique formats
  expect_equal(
    build_formats(c("yq", "ym", "qy")),
    ymd_formats
  )

  expect_equal(
    build_formats("my"),
    c(
      "%m-%y-%d", "%B-%y-%d", "%b-%y-%d", "%m-%Y-%d", "%B-%Y-%d", "%b-%Y-%d",
      "%m%y%d", "%B%y%d", "%b%y%d", "%m%Y%d", "%B%Y%d", "%b%Y%d"
    )
  )

  expect_equal(
    build_format_from_order("abp"),
    c(
      "%a-%m-%p", "%A-%m-%p", "%a-%B-%p", "%A-%B-%p", "%a-%b-%p", "%A-%b-%p",
      "%a%m%p", "%A%m%p", "%a%B%p", "%A%B%p", "%a%b%p", "%A%b%p"
    )
  )

  expect_error(
    build_formats(c("vu", "ymd")),
    '"vu" `orders` not supported in Arrow'
  )

  expect_error(
    build_formats(c("abc")),
    '"abc" `orders` not supported in Arrow'
  )

  expect_equal(
    build_formats("wIpz"),
    c("%w-%I-%p-%z", "%w%I%p%z")
  )

  expect_equal(
    build_formats("yOmd"),
    ymd_formats
  )

  expect_equal(
    build_format_from_order("ymd"),
    ymd_formats
  )

  expect_equal(
    build_format_from_order("ymdHMS"),
    ymd_hms_formats
  )

  expect_equal(
    build_format_from_order("ymdHM"),
    c(
      "%y-%m-%d-%H-%M", "%Y-%m-%d-%H-%M", "%y-%B-%d-%H-%M",
      "%Y-%B-%d-%H-%M", "%y-%b-%d-%H-%M", "%Y-%b-%d-%H-%M",
      "%y%m%d%H%M", "%Y%m%d%H%M", "%y%B%d%H%M", "%Y%B%d%H%M",
      "%y%b%d%H%M", "%Y%b%d%H%M"
    )
  )

  expect_equal(
    build_format_from_order("ymdH"),
    c(
      "%y-%m-%d-%H", "%Y-%m-%d-%H", "%y-%B-%d-%H",
      "%Y-%B-%d-%H", "%y-%b-%d-%H", "%Y-%b-%d-%H",
      "%y%m%d%H", "%Y%m%d%H", "%y%B%d%H", "%Y%B%d%H",
      "%y%b%d%H", "%Y%b%d%H"
    )
  )

  expect_equal(
    build_formats("y-%b-d-%T"),
    c(
      "%y-%m-%d-%I-%M-%S-%p", "%Y-%m-%d-%I-%M-%S-%p", "%y-%B-%d-%I-%M-%S-%p", "%Y-%B-%d-%I-%M-%S-%p",
      "%y-%b-%d-%I-%M-%S-%p", "%Y-%b-%d-%I-%M-%S-%p", "%y-%m-%d-%H-%M-%S", "%Y-%m-%d-%H-%M-%S",
      "%y-%B-%d-%H-%M-%S", "%Y-%B-%d-%H-%M-%S", "%y-%b-%d-%H-%M-%S", "%Y-%b-%d-%H-%M-%S",
      "%y-%m-%d-%H-%M-%OS", "%Y-%m-%d-%H-%M-%OS", "%y-%B-%d-%H-%M-%OS", "%Y-%B-%d-%H-%M-%OS",
      "%y-%b-%d-%H-%M-%OS", "%Y-%b-%d-%H-%M-%OS", "%y%m%d%I%M%S%p", "%Y%m%d%I%M%S%p",
      "%y%B%d%I%M%S%p", "%Y%B%d%I%M%S%p", "%y%b%d%I%M%S%p", "%Y%b%d%I%M%S%p", "%y%m%d%H%M%S",
      "%Y%m%d%H%M%S", "%y%B%d%H%M%S", "%Y%B%d%H%M%S", "%y%b%d%H%M%S", "%Y%b%d%H%M%S", "%y%m%d%H%M%OS",
      "%Y%m%d%H%M%OS", "%y%B%d%H%M%OS", "%Y%B%d%H%M%OS", "%y%b%d%H%M%OS", "%Y%b%d%H%M%OS"
    )
  )

  expect_equal(
    build_formats("%YdmH%p"),
    c(
      "%y-%d-%m-%H-%p", "%Y-%d-%m-%H-%p", "%y-%d-%B-%H-%p", "%Y-%d-%B-%H-%p",
      "%y-%d-%b-%H-%p", "%Y-%d-%b-%H-%p", "%y%d%m%H%p", "%Y%d%m%H%p",
      "%y%d%B%H%p", "%Y%d%B%H%p", "%y%d%b%H%p", "%Y%d%b%H%p"
    )
  )
})



# tests for datetime rounding ---------------------------------------------

# an "easy" date to avoid conflating tests of different things (i.e., it's
# UTC time, and not one of the edge cases on or extremely close to the
# rounding boundary)
easy_date <- as.POSIXct("2022-10-11 12:00:00", tz = "UTC")
easy_df <- tibble::tibble(datetime = easy_date)

# dates near month boundaries over the course of 1 year
month_boundaries <- c(
  "2021-01-01 00:01:00", "2021-02-01 00:01:00", "2021-03-01 00:01:00",
  "2021-04-01 00:01:00", "2021-05-01 00:01:00", "2021-06-01 00:01:00",
  "2021-07-01 00:01:00", "2021-08-01 00:01:00", "2021-09-01 00:01:00",
  "2021-10-01 00:01:00", "2021-11-01 00:01:00", "2021-12-01 00:01:00",
  "2021-01-31 23:59:00", "2021-02-28 23:59:00", "2021-03-31 23:59:00",
  "2021-04-30 23:59:00", "2021-05-31 23:59:00", "2021-06-30 23:59:00",
  "2021-07-31 23:59:00", "2021-08-31 23:59:00", "2021-09-30 23:59:00",
  "2021-10-31 23:59:00", "2021-11-30 23:59:00", "2021-12-31 23:59:00"
)
year_of_dates <- tibble::tibble(
  datetime = as.POSIXct(month_boundaries, tz = "UTC"),
  date = as.Date(datetime)
)

# test case used to check we catch week boundaries for all week_start values
fortnight <- tibble::tibble(
  date = seq(
    from = as.Date("2022-04-04"),
    to = as.Date("2022-04-17"),
    by = "day"
  ),
  datetime = as.POSIXct(date)
)

# test case to check we catch interval lower boundaries for ceiling_date
boundary_times <- tibble::tibble(
  datetime = as.POSIXct(strptime(c(
    "2022-05-10 00:00:00", # boundary for week when week_start = 7 (Sunday)
    "2022-05-11 00:00:00", # boundary for week when week_start = 1 (Monday)
    "2022-05-12 00:00:00", # boundary for week when week_start = 2 (Tuesday)
    "2022-03-10 00:00:00", # boundary for day, hour, minute, second, millisecond
    "2022-03-10 00:00:01", # boundary for second, millisecond
    "2022-03-10 00:01:00", # boundary for second, millisecond, minute
    "2022-03-10 01:00:00", # boundary for second, millisecond, minute, hour
    "2022-01-01 00:00:00" # boundary for year
  ), tz = "UTC", format = "%F %T")),
  date = as.Date(datetime)
)

# test case to check rounding takes place in local time
datestrings <- c(
  "1970-01-01 00:00:59.123456789",
  "2000-02-29 23:23:23.999999999",
  "1899-01-01 00:59:20.001001001",
  "2033-05-18 03:33:20.000000000",
  "2020-01-01 01:05:05.001",
  "2019-12-31 02:10:10.002",
  "2019-12-30 03:15:15.003",
  "2009-12-31 04:20:20.004132",
  "2010-01-01 05:25:25.005321",
  "2010-01-03 06:30:30.006163",
  "2010-01-04 07:35:35",
  "2006-01-01 08:40:40",
  "2005-12-31 09:45:45",
  "2008-12-28 00:00:00",
  "2008-12-29 00:00:00",
  "2012-01-01 01:02:03"
)
tz_times <- tibble::tibble(
  utc_time = as.POSIXct(datestrings, tz = "UTC"),
  syd_time = as.POSIXct(datestrings, tz = "Australia/Sydney"), # UTC +10   (UTC +11 with DST)
  adl_time = as.POSIXct(datestrings, tz = "Australia/Adelaide"), # UTC +9:30 (UTC +10:30 with DST)
  mar_time = as.POSIXct(datestrings, tz = "Pacific/Marquesas"), # UTC -9:30 (no DST)
  kat_time = as.POSIXct(datestrings, tz = "Asia/Kathmandu") # UTC +5:45 (no DST)
)

test_that("timestamp round/floor/ceiling works for a minimal test", {
  compare_dplyr_binding(
    .input %>%
      mutate(
        round_datetime = round_date(datetime),
        floor_datetime = floor_date(datetime),
        ceiling_datetime = ceiling_date(datetime, change_on_boundary = FALSE)
      ) %>%
      collect(),
    test_df
  )
})

test_that("timestamp round/floor/ceiling accepts period unit abbreviation", {
  # test helper to ensure standard abbreviations of period names
  # are understood by arrow and mirror the lubridate behaviour
  check_period_abbreviation <- function(unit, synonyms) {
    # check arrow against lubridate
    compare_dplyr_binding(
      .input %>%
        mutate(out_1 = round_date(datetime, unit)) %>%
        collect(),
      easy_df
    )

    # check synonyms
    base <- call_binding("round_date", Expression$scalar(easy_date), unit)
    for (syn in synonyms) {
      expect_equal(
        call_binding("round_date", Expression$scalar(easy_date), syn),
        base
      )
    }
  }

  check_period_abbreviation("minute", synonyms = c("minutes", "min", "mins"))
  check_period_abbreviation("second", synonyms = c("seconds", "sec", "secs"))
  check_period_abbreviation("month", synonyms = c("months", "mon", "mons"))
})

test_that("temporal round/floor/ceiling accepts periods with multiple units", {
  check_multiple_unit_period <- function(unit, multiplier) {
    unit_string <- paste(multiplier, unit)
    compare_dplyr_binding(
      .input %>%
        mutate(
          round_datetime = round_date(datetime, unit_string),
          floor_datetime = floor_date(datetime, unit_string),
          ceiling_datetime = ceiling_date(datetime, unit_string)
        ) %>%
        collect(),
      easy_df
    )
  }

  for (multiplier in c(1, 2, 10)) {
    for (unit in c("second", "minute", "day", "year")) {
      check_multiple_unit_period(unit, multiplier)
    }
  }
})

# Test helper functions for checking equivalence of outputs regardless of
# the unit specified. The lubridate_unit argument allows for cases where
# arrow supports a unit name (e.g., nanosecond) that lubridate doesn't. Also
# note that in the check_date_rounding helper the lubridate output is coerced
# to ensure type stable output (arrow output should be type stable without this)

check_date_rounding <- function(data, unit, lubridate_unit = unit, ...) {
  expect_equal(
    data %>%
      arrow_table() %>%
      mutate(
        date_rounded = round_date(date, unit),
        date_floored = floor_date(date, unit),
        date_ceiling = ceiling_date(date, unit)
      ) %>%
      collect(),
    data %>%
      mutate(
        date_rounded = as.Date(round_date(date, lubridate_unit)),
        date_floored = as.Date(floor_date(date, lubridate_unit)),
        date_ceiling = as.Date(ceiling_date(date, lubridate_unit))
      ),
    ...
  )
}

check_timestamp_rounding <- function(data, unit, lubridate_unit = unit, ...) {
  expect_equal(
    data %>%
      arrow_table() %>%
      mutate(
        datetime_rounded = round_date(datetime, unit),
        datetime_floored = floor_date(datetime, unit),
        datetime_ceiling = ceiling_date(datetime, unit)
      ) %>%
      collect(),
    data %>%
      mutate(
        datetime_rounded = round_date(datetime, lubridate_unit),
        datetime_floored = floor_date(datetime, lubridate_unit),
        datetime_ceiling = ceiling_date(datetime, lubridate_unit)
      ),
    ...
  )
}

test_that("date round/floor/ceil works for units of 1 day or less", {
  test_df %>% check_date_rounding("1 millisecond", lubridate_unit = ".001 second")
  test_df %>% check_date_rounding("1 second")
  test_df %>% check_date_rounding("1 hour")

  skip("floor_date(as.Date(NA), '1 day') is no longer NA on latest R-devel")
  # Possibly https://github.com/wch/r-source/commit/4f70ce0d79eeda7464cf97448e515275cbef754b
  test_df %>% check_date_rounding("1 day")
})

test_that("timestamp round/floor/ceil works for units of 1 day or less", {
  test_df %>% check_timestamp_rounding("second")
  test_df %>% check_timestamp_rounding("minute")
  test_df %>% check_timestamp_rounding("hour")
  test_df %>% check_timestamp_rounding("day")

  test_df %>% check_timestamp_rounding(".01 second")
  test_df %>% check_timestamp_rounding(".001 second")
  test_df %>% check_timestamp_rounding(".00001 second")

  test_df %>% check_timestamp_rounding("1 millisecond", lubridate_unit = ".001 second")
  test_df %>% check_timestamp_rounding("1 microsecond", lubridate_unit = ".000001 second")
  test_df %>% check_timestamp_rounding("1 nanosecond", lubridate_unit = ".000000001 second")
})

test_that("timestamp round/floor/ceil works for units: month/quarter/year", {
  year_of_dates %>% check_timestamp_rounding("month", ignore_attr = TRUE)
  year_of_dates %>% check_timestamp_rounding("quarter", ignore_attr = TRUE)
  year_of_dates %>% check_timestamp_rounding("year", ignore_attr = TRUE)
})

# check helper invoked when we need to avoid the lubridate rounding bug
check_date_rounding_1051_bypass <- function(data, unit, ignore_attr = TRUE, ...) {
  # directly compare arrow to lubridate for floor and ceiling
  compare_dplyr_binding(
    .input %>%
      mutate(
        date_floored = floor_date(date, unit),
        date_ceiling = ceiling_date(date, unit)
      ) %>%
      collect(),
    data,
    ignore_attr = ignore_attr,
    ...
  )

  # The rounding tests for dates is run against Arrow timestamp behaviour
  # because of a lubridate bug specific to Date objects with week and
  # higher-unit rounding (see lubridate issue 1051)
  # https://github.com/tidyverse/lubridate/issues/1051
  out <- data %>%
    arrow_table() %>%
    mutate(
      out_date = date %>% round_date(unit), # Date
      out_time = datetime %>% round_date(unit) # POSIXct
    ) %>%
    collect()

  expect_equal(
    out$out_date,
    as.Date(out$out_time)
  )
}

test_that("date round/floor/ceil works for units: month/quarter/year", {
  # these test cases are affected by lubridate issue 1051 so we bypass
  # lubridate::round_date() for Date objects with large rounding units
  # https://github.com/tidyverse/lubridate/issues/1051

  check_date_rounding_1051_bypass(year_of_dates, "month", ignore_attr = TRUE)
  check_date_rounding_1051_bypass(year_of_dates, "quarter", ignore_attr = TRUE)
  check_date_rounding_1051_bypass(year_of_dates, "year", ignore_attr = TRUE)
})

check_date_week_rounding <- function(data, week_start, ignore_attr = TRUE, ...) {
  expect_equal(
    data %>%
      arrow_table() %>%
      mutate(
        date_rounded = round_date(date, unit),
        date_floored = floor_date(date, unit),
        date_ceiling = ceiling_date(date, unit)
      ) %>%
      collect(),
    data %>%
      mutate(
        date_rounded = as.Date(round_date(date, lubridate_unit)),
        date_floored = as.Date(floor_date(date, lubridate_unit)),
        date_ceiling = as.Date(ceiling_date(date, lubridate_unit))
      ),
    ignore_attr = ignore_attr,
    ...
  )
}

check_timestamp_week_rounding <- function(data, week_start, ignore_attr = TRUE, ...) {
  compare_dplyr_binding(
    .input %>%
      mutate(
        datetime_rounded = round_date(datetime, "week", week_start = week_start),
        datetime_floored = floor_date(datetime, "week", week_start = week_start),
        datetime_ceiling = ceiling_date(datetime, "week", week_start = week_start)
      ) %>%
      collect(),
    data,
    ignore_attr = ignore_attr,
    ...
  )
}

test_that("timestamp round/floor/ceil works for week units (standard week_start)", {
  fortnight %>% check_timestamp_week_rounding(week_start = 1) # Monday
  fortnight %>% check_timestamp_week_rounding(week_start = 7) # Sunday
})

test_that("timestamp round/floor/ceil works for week units (non-standard week_start)", {
  fortnight %>% check_timestamp_week_rounding(week_start = 2) # Tuesday
  fortnight %>% check_timestamp_week_rounding(week_start = 3) # Wednesday
  fortnight %>% check_timestamp_week_rounding(week_start = 4) # Thursday
  fortnight %>% check_timestamp_week_rounding(week_start = 5) # Friday
  fortnight %>% check_timestamp_week_rounding(week_start = 6) # Saturday
})

check_date_week_rounding <- function(data, week_start, ignore_attr = TRUE, ...) {
  # directly compare arrow to lubridate for floor and ceiling
  compare_dplyr_binding(
    .input %>%
      mutate(
        date_floored = floor_date(date, "week", week_start = week_start),
        date_ceiling = ceiling_date(date, "week", week_start = week_start)
      ) %>%
      collect(),
    data,
    ignore_attr = ignore_attr,
    ...
  )

  # use the bypass method to avoid the lubridate-1051 bug for week units
  # https://github.com/tidyverse/lubridate/issues/1051
  out <- data %>%
    arrow_table() %>%
    mutate(
      out_date = date %>% round_date("week", week_start = week_start), # Date
      out_time = datetime %>% round_date("week", week_start = week_start) # POSIXct
    ) %>%
    collect()

  expect_equal(
    out$out_date,
    as.Date(out$out_time)
  )
}

test_that("date round/floor/ceil works for week units (standard week_start)", {
  check_date_week_rounding(fortnight, week_start = 1) # Monday
  check_date_week_rounding(fortnight, week_start = 7) # Sunday
})

test_that("date round/floor/ceil works for week units (non-standard week_start)", {
  check_date_week_rounding(fortnight, week_start = 2) # Tuesday
  check_date_week_rounding(fortnight, week_start = 3) # Wednesday
  check_date_week_rounding(fortnight, week_start = 4) # Thursday
  check_date_week_rounding(fortnight, week_start = 5) # Friday
  check_date_week_rounding(fortnight, week_start = 6) # Saturday
})

# Test helper used to check that the change_on_boundary argument to
# ceiling_date behaves identically to the lubridate version. It takes
# unit as an argument to run tests separately for different rounding units
check_boundary_with_unit <- function(unit, ...) {
  # timestamps
  compare_dplyr_binding(
    .input %>%
      mutate(
        cob_null = ceiling_date(datetime, unit, change_on_boundary = NULL),
        cob_true = ceiling_date(datetime, unit, change_on_boundary = TRUE),
        cob_false = ceiling_date(datetime, unit, change_on_boundary = FALSE)
      ) %>%
      collect(),
    boundary_times,
    ...
  )

  # dates
  expect_equal(
    boundary_times %>%
      arrow_table() %>%
      mutate(
        cob_null = ceiling_date(date, unit, change_on_boundary = NULL),
        cob_true = ceiling_date(date, unit, change_on_boundary = TRUE),
        cob_false = ceiling_date(date, unit, change_on_boundary = FALSE)
      ) %>%
      collect(),
    boundary_times %>%
      mutate(
        cob_null = as.Date(ceiling_date(date, unit, change_on_boundary = NULL)),
        cob_true = as.Date(ceiling_date(date, unit, change_on_boundary = TRUE)),
        cob_false = as.Date(ceiling_date(date, unit, change_on_boundary = FALSE))
      ),
    ...
  )
}

test_that("ceiling_date() applies change_on_boundary correctly", {
  check_boundary_with_unit(".001 second")
  check_boundary_with_unit("second")
  check_boundary_with_unit("minute", tolerance = 0.001) # floating point issue?
  check_boundary_with_unit("hour")
  check_boundary_with_unit("day")
})

# In lubridate, an error is thrown when 60 sec/60 min/24 hour thresholds are
# exceeded. Checks that arrow mimics this behaviour and throws an identically
# worded error message
test_that("temporal round/floor/ceil period unit maxima are enforced", {
  expect_error(
    call_binding("round_date", Expression$scalar(Sys.time()), "61 seconds"),
    "Rounding with second > 60 is not supported"
  )
  expect_error(
    call_binding("round_date", Expression$scalar(Sys.time()), "61 minutes"),
    "Rounding with minute > 60 is not supported"
  )
  expect_error(
    call_binding("round_date", Expression$scalar(Sys.time()), "25 hours"),
    "Rounding with hour > 24 is not supported"
  )
  expect_error(
    call_binding("round_date", Expression$scalar(Sys.Date()), "25 hours"),
    "Rounding with hour > 24 is not supported"
  )
})

# one method to test that temporal rounding takes place in local time is to
# use lubridate as a ground truth and compare arrow results to lubridate
# results. this test helper runs that test, skipping cases where lubridate
# produces incorrect answers
check_timezone_rounding_vs_lubridate <- function(data, unit) {
  # esoteric lubridate bug: on windows and macOS (not linux), lubridate returns
  # incorrect ceiling/floor for timezoned POSIXct times (syd, adl, kat zones,
  # but not mar) but not utc, and not for round, and only for these two
  # timestamps where high-precision timing is relevant to the outcome
  if (unit %in% c(".001 second", "second", "minute")) {
    if (tolower(Sys.info()[["sysname"]]) %in% c("windows", "darwin")) {
      data <- data[-c(1, 3), ]
    }
  }

  # external validity check: compare lubridate to arrow
  compare_dplyr_binding(
    .input %>%
      mutate(
        utc_floored = floor_date(utc_time, unit = unit),
        utc_rounded = round_date(utc_time, unit = unit),
        utc_ceiling = ceiling_date(utc_time, unit = unit),
        syd_floored = floor_date(syd_time, unit = unit),
        syd_rounded = round_date(syd_time, unit = unit),
        syd_ceiling = ceiling_date(syd_time, unit = unit),
        adl_floored = floor_date(adl_time, unit = unit),
        adl_rounded = round_date(adl_time, unit = unit),
        adl_ceiling = ceiling_date(adl_time, unit = unit),
        mar_floored = floor_date(mar_time, unit = unit),
        mar_rounded = round_date(mar_time, unit = unit),
        mar_ceiling = ceiling_date(mar_time, unit = unit),
        kat_floored = floor_date(kat_time, unit = unit),
        kat_rounded = round_date(kat_time, unit = unit),
        kat_ceiling = ceiling_date(kat_time, unit = unit)
      ) %>%
      collect(),
    data
  )
}

# another method to check that temporal rounding takes place in local
# time is to test the internal consistency of the YMD HMS values returned
# by temporal rounding functions: these should be the same regardless of
# timezone and should always be identical to the equivalent result calculated
# for UTC test. this test isn't useful for subsecond resolution but avoids
# dependency on lubridate
check_timezone_rounding_for_consistency <- function(data, unit) {
  shifted_times <- data %>%
    arrow_table() %>%
    mutate(
      utc_floored = floor_date(utc_time, unit = unit),
      utc_rounded = round_date(utc_time, unit = unit),
      utc_ceiling = ceiling_date(utc_time, unit = unit),
      syd_floored = floor_date(syd_time, unit = unit),
      syd_rounded = round_date(syd_time, unit = unit),
      syd_ceiling = ceiling_date(syd_time, unit = unit),
      adl_floored = floor_date(adl_time, unit = unit),
      adl_rounded = round_date(adl_time, unit = unit),
      adl_ceiling = ceiling_date(adl_time, unit = unit),
      mar_floored = floor_date(mar_time, unit = unit),
      mar_rounded = round_date(mar_time, unit = unit),
      mar_ceiling = ceiling_date(mar_time, unit = unit),
      kat_floored = floor_date(kat_time, unit = unit),
      kat_rounded = round_date(kat_time, unit = unit),
      kat_ceiling = ceiling_date(kat_time, unit = unit)
    ) %>%
    collect()

  compare_local_times <- function(time1, time2) {
    all(year(time1) == year(time1) &
      month(time1) == month(time2) &
      day(time1) == day(time2) &
      hour(time1) == hour(time2) &
      minute(time1) == minute(time2) &
      second(time1) == second(time1))
  }

  base <- shifted_times$utc_rounded
  expect_true(compare_local_times(shifted_times$syd_rounded, base))
  expect_true(compare_local_times(shifted_times$adl_rounded, base))
  expect_true(compare_local_times(shifted_times$mar_rounded, base))
  expect_true(compare_local_times(shifted_times$kat_rounded, base))

  base <- shifted_times$utc_floored
  expect_true(compare_local_times(shifted_times$syd_floored, base))
  expect_true(compare_local_times(shifted_times$adl_floored, base))
  expect_true(compare_local_times(shifted_times$mar_floored, base))
  expect_true(compare_local_times(shifted_times$kat_floored, base))

  base <- shifted_times$utc_ceiling
  expect_true(compare_local_times(shifted_times$syd_ceiling, base))
  expect_true(compare_local_times(shifted_times$adl_ceiling, base))
  expect_true(compare_local_times(shifted_times$mar_ceiling, base))
  expect_true(compare_local_times(shifted_times$kat_ceiling, base))
}

test_that("timestamp rounding takes place in local time", {
  tz_times %>% check_timezone_rounding_vs_lubridate(".001 second")
  tz_times %>% check_timezone_rounding_vs_lubridate("second")
  tz_times %>% check_timezone_rounding_vs_lubridate("minute")
  tz_times %>% check_timezone_rounding_vs_lubridate("hour")
  tz_times %>% check_timezone_rounding_vs_lubridate("day")
  tz_times %>% check_timezone_rounding_vs_lubridate("week")
  tz_times %>% check_timezone_rounding_vs_lubridate("month")
  tz_times %>% check_timezone_rounding_vs_lubridate("quarter")
  tz_times %>% check_timezone_rounding_vs_lubridate("year")

  tz_times %>% check_timezone_rounding_for_consistency("second")
  tz_times %>% check_timezone_rounding_for_consistency("minute")
  tz_times %>% check_timezone_rounding_for_consistency("hour")
  tz_times %>% check_timezone_rounding_for_consistency("day")
  tz_times %>% check_timezone_rounding_for_consistency("week")
  tz_times %>% check_timezone_rounding_for_consistency("month")
  tz_times %>% check_timezone_rounding_for_consistency("quarter")
  tz_times %>% check_timezone_rounding_for_consistency("year")

  tz_times %>% check_timezone_rounding_for_consistency("7 seconds")
  tz_times %>% check_timezone_rounding_for_consistency("7 minutes")
  tz_times %>% check_timezone_rounding_for_consistency("7 hours")
  tz_times %>% check_timezone_rounding_for_consistency("7 months")
  tz_times %>% check_timezone_rounding_for_consistency("7 years")

  tz_times %>% check_timezone_rounding_for_consistency("13 seconds")
  tz_times %>% check_timezone_rounding_for_consistency("13 minutes")
  tz_times %>% check_timezone_rounding_for_consistency("13 hours")
  tz_times %>% check_timezone_rounding_for_consistency("13 months")
  tz_times %>% check_timezone_rounding_for_consistency("13 years")
})

test_that("with_tz() and force_tz() works", {
  timestamps <- as_datetime(c(
    "1970-01-01T00:00:59.123456789",
    "2000-02-29T23:23:23.999999999",
    "2033-05-18T03:33:20.000000000",
    "2020-01-01T01:05:05.001",
    "2019-12-31T02:10:10.002",
    "2019-12-30T03:15:15.003",
    "2009-12-31T04:20:20.004132",
    "2010-01-01T05:25:25.005321",
    "2010-01-03T06:30:30.006163",
    "2010-01-04T07:35:35",
    "2006-01-01T08:40:40",
    "2005-12-31T09:45:45",
    "2008-12-28",
    "2008-12-29",
    "2012-01-01 01:02:03"
  ), tz = "UTC")

  timestamps_non_utc <- force_tz(timestamps, "America/Chicago")

  nonexistent <- as_datetime(c(
    "2015-03-29 02:30:00",
    "2015-03-29 03:30:00"
  ), tz = "UTC")

  ambiguous <- as_datetime(c(
    "2015-10-25 02:30:00",
    "2015-10-25 03:30:00"
  ), tz = "UTC")

  compare_dplyr_binding(
    .input %>%
      mutate(
        timestamps_with_tz_1 = with_tz(timestamps, "UTC"),
        timestamps_with_tz_2 = with_tz(timestamps, "America/Chicago"),
        timestamps_with_tz_3 = with_tz(timestamps, "Asia/Kolkata"),
        timestamps_force_tz_1 = force_tz(timestamps, "UTC"),
        timestamps_force_tz_2 = force_tz(timestamps, "America/Chicago"),
        timestamps_force_tz_3 = force_tz(timestamps, "Asia/Kolkata")
      ) %>%
      collect(),
    tibble::tibble(timestamps = timestamps)
  )

  compare_dplyr_binding(
    .input %>%
      mutate(
        timestamps_with_tz_1 = with_tz(timestamps, "UTC"),
        timestamps_with_tz_2 = with_tz(timestamps, "America/Chicago"),
        timestamps_with_tz_3 = with_tz(timestamps, "Asia/Kolkata")
      ) %>%
      collect(),
    tibble::tibble(timestamps = timestamps_non_utc)
  )

  # We can match some roll_dst behaviour for nonexistent times
  compare_dplyr_binding(
    .input %>%
      mutate(
        timestamps_with_tz_1 = force_tz(
          timestamps,
          "Europe/Brussels",
          roll_dst = c("boundary", "post")
        )
      ) %>%
      collect(),
    tibble::tibble(timestamps = nonexistent)
  )

  # We can match all roll_dst behaviour for ambiguous times
  compare_dplyr_binding(
    .input %>%
      mutate(
        # The difference is easier to see if we transform back to UTC
        # because both pre and post will display as 02:30 otherwise
        timestamps_with_tz_pre = with_tz(
          force_tz(
            timestamps,
            "Europe/Brussels",
            roll_dst = c("boundary", "pre")
          ),
          "UTC"
        ),
        timestamps_with_tz_post = with_tz(
          force_tz(
            timestamps,
            "Europe/Brussels",
            roll_dst = c("boundary", "post")
          ),
          "UTC"
        )
      ) %>%
      collect(),
    tibble::tibble(timestamps = ambiguous)
  )

  # non-UTC timezone to other timezone is not supported in arrow's force_tz()
  expect_warning(
    tibble::tibble(timestamps = timestamps_non_utc) %>%
      arrow_table() %>%
      mutate(timestamps = force_tz(timestamps, "UTC")) %>%
      collect(),
    "`time` with a non-UTC timezone not supported in Arrow"
  )

  # We only support some roll_dst values
  expect_warning(
    tibble::tibble(timestamps = nonexistent) %>%
      arrow_table() %>%
      mutate(timestamps = force_tz(
        timestamps,
        "Europe/Brussels",
        roll_dst = "post"
      )) %>%
      collect(),
    "roll_dst` value must be 'error' or 'boundary' for nonexistent times"
  )

  expect_warning(
    tibble::tibble(timestamps = nonexistent) %>%
      arrow_table() %>%
      mutate(timestamps = force_tz(
        timestamps,
        "Europe/Brussels",
        roll_dst = c("boundary", "NA")
      )) %>%
      collect(),
    "`roll_dst` value must be 'error', 'pre', or 'post' for nonexistent times"
  )

  # Raise error when the timezone falls into the DST-break
  expect_error(
    record_batch(timestamps = nonexistent) %>%
      mutate(nonexistent_roll_false = force_tz(timestamps, "Europe/Brussels")) %>%
      collect(),
    "Timestamp doesn't exist in timezone 'Europe/Brussels'"
  )
})

test_that("with_tz() and force_tz() can add timezone to timestamp without timezone", {
  timestamps <- Array$create(1L:10L, int64())$cast(timestamp("s"))

  expect_equal(
    arrow_table(timestamps = timestamps) %>%
      mutate(timestamps = with_tz(timestamps, "America/Chicago")) %>%
      compute(),
    arrow_table(timestamps = timestamps$cast(timestamp("s", "America/Chicago")))
  )

  expect_equal(
    arrow_table(timestamps = timestamps) %>%
      mutate(timestamps = force_tz(timestamps, "America/Chicago")) %>%
      compute(),
    arrow_table(
      timestamps = call_function("assume_timezone", timestamps, options = list(timezone = "America/Chicago"))
    )
  )
})