Add Redshift lag and lead functions without the default parameter

[hdplsa]

Fixes #548

This pull request removes the default argument from the LAG and LEAD SQL functions. Redshift assumes that the default parameter is NULL, therefore, we added an error in case that the user gives an alternative value for default.

Here is a reprex for the same code in #548 but with the fix applied:

library(dbplyr)
library(DBI)
library(reprex)

con <- dbConnect(RPostgres::Redshift(),
                 host = "",
                 dbname = "",
                 port = ,
                 user = "",
                 password = )

table.sql <- sql("SELECT 1 AS YR, 1 AS MON, 123456 AS SALES
                  UNION ALL
                  SELECT 1 AS YR, 2 AS MON, 234567 AS SALES
                  UNION ALL
                  SELECT 1 AS YR, 3 AS MON, 345678 AS SALES
                  UNION ALL
                  SELECT 1 AS YR, 4 AS MON, NULL AS SALES
                  UNION ALL
                  SELECT 2 AS YR, 1 AS MON, 123456 AS SALES
                  UNION ALL
                  SELECT 2 AS YR, 2 AS MON, 234567 AS SALES
                  UNION ALL
                  SELECT 2 AS YR, 3 AS MON, 345678 AS SALES
                  UNION ALL
                  SELECT 3 AS YR, 1 AS MON, 123456 AS SALES
                  ")

new.table <- con %>% dplyr::tbl(table.sql)

new.table %>%
  dbplyr::window_order(yr, mon) %>%
  dplyr::mutate(LG = lag(sales), LD = lead(sales)) %>%
  dplyr::arrange(yr, mon)
#> # Source:     lazy query [?? x 5]
#> # Database:   postgres
#> #   []
#> # Ordered by: yr, mon, yr, mon
#>      yr   mon  sales     lg     ld
#>   <int> <int>  <int>  <int>  <int>
#> 1     1     1 123456     NA 234567
#> 2     1     2 234567 123456 345678
#> 3     1     3 345678 234567     NA
#> 4     1     4     NA 345678 123456
#> 5     2     1 123456     NA 234567
#> 6     2     2 234567 123456 345678
#> 7     2     3 345678 234567 123456
#> 8     3     1 123456 345678     NA

^{Created on 2020-11-26 by the reprex package (v0.3.0)}

And if the user gives a default value other than NA:

new.table %>%
  dbplyr::window_order(yr, mon) %>%
  dplyr::mutate(LG = lag(sales, default = 2), LD = lead(sales)) %>%
  dplyr::arrange(yr, mon)
#> Error in lag(sales, default = 2): Redshift does not support `default` values in LAG. By design `default` is always NULL.

^{Created on 2020-11-26 by the reprex package (v0.3.0)}

[hadley]

Can you please also add a bullet to the top of NEWS.md? It should briefly describe the change and end with (@yourname, #issuenumber).

[hadley]

I think this would be simpler if you just omitted the default argument all together — then the user will just get the regular unused argument (default = 1) error

[hadley]

This will also need to be updated

[hdplsa]

Hi @hadley, just did the changes you suggested.

[hadley]

Thank you!