mutate fails with empty name-value pair

[davidlockhart]

test.df <- data.frame(x = 1:2, y=3:4)
mutate(test.df,
z = x+y,
w = x-y,
)

produces the error message:
Error in mutate_impl(.data, dots) :
attempt to use zero-length variable name

It would be really nice if this were just gracefully parsed away, perhaps with a warning, or if necessary gave a more informative error. When I googled the message I found lots of suggestions of empty dataframe column names due to ingestion errors, which became rather a snipe hunt, but no suggestions that it could be an empty new variable name, which would have pointed me in a better direction.

[hadley]

Duplicate of #1039