`tiledb_query_import_buffer()` is not working with `Date` and `POSIXct`

[cgiachalis]

tiledb_query_import_buffer() works for nanosecond resolution but not for Date and POSIXct objects.

library(tiledb)

uri <- tempfile()

d1 <- data.frame(date32 = Sys.Date(),
                 datetime = Sys.time())

nms <- colnames(d1)

# Map to TileDB as DATETIME_DAY, DATETIME_MS
tiledb::fromDataFrame(d1, uri)


arr <- tiledb::tiledb_array(uri)
qry <- tiledb::tiledb_query(arr, "WRITE")

# Date
na1 <- nanoarrow::as_nanoarrow_array(d1[[1]], schema = arrow::date32())
qry <- tiledb::tiledb_query_import_buffer(qry, nms[1], na1)
#> Error: [TileDB-Arrow]: Unknown or unsupported Arrow format string 'tdD'

# Posixct
na2 <- nanoarrow::as_nanoarrow_array(d1[[2]], schema = arrow::timestamp("ms", "UTC"))
qry <- tiledb::tiledb_query_import_buffer(qry, nms[2], na2)
#> Error: [TileDB-Arrow]: Unknown or unsupported Arrow format string 'tsm:UTC'

The above error has been raised in arrow_type_to_tiledb where there is no mapping for tdD and tsm https://github.com/TileDB-Inc/TileDB-R/blob/f7792f1a90d0c1ea5d20c3e53da0266742ad8dfe/src/updated_arrow_io_impl.h#L307-L317

nanosecond example works OK:

library(tiledb)
uri <- tempfile()

d1 <- data.frame(index = nanotime::as.nanotime(1), var = "a")

# Map to TileDB as DATETIME_NS
tiledb::fromDataFrame(d1, uri, col_index = 1)

arr <- tiledb::tiledb_array(uri)
qry <- tiledb::tiledb_query(arr, "WRITE")

d2 <- data.frame(index = nanotime::as.nanotime(2:3), var = c("b", "c"))

# we create an arrow Array object as nanoarrow::as_nanoarrow_array doesn't handle
# int64->timestamp mapping; see https://github.com/apache/arrow-nanoarrow/issues/811

na <- arrow::arrow_array(bit64::as.integer64(d2[[1]]))$cast(arrow::timestamp("ns"))
na <- nanoarrow::as_nanoarrow_array(na)

qry <- tiledb::tiledb_query_import_buffer(qry, "index", na)

na2 <- nanoarrow::as_nanoarrow_array(d2[[2]])
qry <- tiledb::tiledb_query_import_buffer(qry, "var", na2)

tiledb::tiledb_query_set_layout(qry, "UNORDERED")
tiledb::tiledb_query_submit(qry)
tiledb::tiledb_query_finalize(qry)

arr <- tiledb::tiledb_array(uri, return_as = "data.table")
arr[]
#>                                  index    var
#>                             <nanotime> <char>
#> 1: 1970-01-01T00:00:00.000000001+00:00      a
#> 2: 1970-01-01T00:00:00.000000002+00:00      b
#> 3: 1970-01-01T00:00:00.000000003+00:00      c