Name repair for duplicated columns inconsistent between read_csv and spec_csv #1387
Description
When trying to create column specifications of a file that contains duplicated variable names, spec_csv() renames the variables differently (e.g "error_1") than read_csv() ("error...1").
Let test.csv be a simple CSV with duplicated variables "error" and one observation:
echo "a,error,b,error,c,error\n1,string1,2,string2,3,string3" >> test.csvreadr::spec_csv("test.csv")cols(
a = col_double(),
error = col_character(),
b = col_double(),
error_1 = col_character(),
c = col_double(),
error_2 = col_character()
)
Warning message:
Duplicated column names deduplicated: 'error' => 'error_1' [4], 'error' => 'error_2' [6]
readr::read_csv("test.csv")New names:
* error -> error...2
* error -> error...4
* error -> error...6
Rows: 1 Columns: 6
── Column specification ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
Delimiter: ","
chr (3): error...2, error...4, error...6
dbl (3): a, b, c
ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
# A tibble: 1 × 6
a error...2 b error...4 c error...6
<dbl> <chr> <dbl> <chr> <dbl> <chr>
1 1 string1 2 string2 3 string3