seq_linter should lint on 1:n() [dplyr] and 1:.N [data.table]

[MichaelChirico]

Both of these are flavor-specific versions of 1:nrow(.) that we should recognize and lint