Improve separate() error when no sep specified

[pgensler]

This is not very informative, other than the underlying logic failed

data2 <- c(
  'Order/ID : IN-2013-77878',
  'Order/Date : 2/5/13',
  'Ship/Mode: Second Class',
  'product/id : TEC-AC-10003033',
  '   ',
  'Order/ID : IN-2013-77878',
  'Order/Date : 2/8/13',
  'Ship/Mode: Second Class',
  'product/id : FUR-CH-10004050'
)


> data <- tibble::enframe(data2) %>%
+   select(value) %>%
+   filter(., str_detect(value,':')) %>%
+   separate(., value, into = c('var','val')) # use tidyr's seperate to split data
Warning message:
Too many values at 8 locations: 1, 2, 3, 4, 5, 6, 7, 8 

It fails due to no sep argument, so maybe outputting an error message stating "Hey are you missing a separator argument?)

This works:

data <- tibble::enframe(data2) %>%
  select(value) %>%
  filter(., str_detect(value,':')) %>%
  separate(., value, into = c('var','val'), sep = ':')

[hadley]

The default splits at non-alphanumeric symbols, which I think is useful. But maybe the error message could be more informative here