Avoid duplicated columns in count, tally, and summarise

[eutwt]

This fixes the following errors (and instead produces results matching dplyr)

library(dtplyr)
library(dplyr, warn.conflicts = FALSE)
dt <- lazy_dt(data.frame(a = 2, n = 3))

dt %>% 
  count(a, name = 'a') %>% 
  as_tibble()
#> Error: Column name `a` must not be duplicated.
#> Use .name_repair to specify repair.

dt %>% 
  group_by(a) %>% 
  summarise(a = 1) %>% 
  as_tibble()
#> Error: Column name `a` must not be duplicated.
#> Use .name_repair to specify repair.

dt %>% 
  count(n) %>% 
  as_tibble()
#> Error: Column name `n` must not be duplicated.
#> Use .name_repair to specify repair.

^{Created on 2021-09-04 by the reprex package (v2.0.1)}

[eutwt]

I gave a name to an input in one of the pivot_wider tests because earlier the snapshot of the expression was showing a different name, e.g. setnames(dcast(`_DT17`... instead of setnames(dcast(`_DT12`..., causing the test to fail.

[hadley]

Thanks!