read_csv: Simultaneous selection and renaming of columns

[nbsmokee]

When reading data with read_csv (and probably other read_* functions), I sometimes need to read only some columns and also rename them. When I try to do this with one call to read_csv, I would expect that when I use both the col_select and col_names keywords, I have to provide only new names for the selected columns, but instead, in order for the function to produce the desired output, I have to provide names for all columns in the file, which is somewhat annoying, at least in my particular use case.

In the below example, I think it is much more natural to use the 1. test case, however the result I expect is produced only by the 2. function call, the 1st one giving an error.

library("readr")

# Create test file
readr::write_csv(data.frame(array(1:24, c(6, 4))), "test.csv")

# TEST CASES:
# 1. Attempt (and fail) to select AND remane simultaneously
readr::read_csv("test.csv", col_select = 2:3, skip = 1, col_names = paste0("col", 2:3))

# 2. Working (but unpleasant) solution: provide new names for ALL columns
readr::read_csv("test.csv", col_select = 2:3, skip = 1, col_names = paste0("col", 1:4))