spec_csv errors on one-line literal data

[nicolaspayette]

I would expect spec_csv to work on literal data, just like read_csv, but when the data passed to it is just a single line, it tries to interpret the string as a filename:

library(readr) # v2.1.5

# `spec_csv` works with a literal two-line string:
spec_csv(file = I("1,x\n2,y"), col_names = c("a", "b"))
#> cols(
#>   a = col_double(),
#>   b = col_character()
#> )

# but doesn't work with a one-line literal string:
spec_csv(file = I("1,x"), col_names = c("a", "b"))
#> Error: '1,x' does not exist in current working directory ('/tmp/RtmptpBMP2/reprex-276b5141a9194f-jolly-bass').

# while `read_csv` itself can handle it just fine
spec(read_csv(file = I("1,x"), col_names = c("a", "b")))
#> Rows: 1 Columns: 2
#> ── Column specification ────────────────────────────────────────────────────────
#> Delimiter: ","
#> chr (1): b
#> dbl (1): a
#> 
#> 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.
#> cols(
#>   a = col_double(),
#>   b = col_character()
#> )

^{Created on 2024-03-04 with reprex v2.1.0}

[jennybc]

To get you unstuck in the here and now, I advise that you end with an explicit newline like so:

library(readr)

spec_csv(file = I("1,x\n"), col_names = c("a", "b"))
#> cols(
#>   a = col_double(),
#>   b = col_character()
#> )
spec_csv(file = "1,x\n", col_names = c("a", "b"))
#> cols(
#>   a = col_double(),
#>   b = col_character()
#> )

(The spec_*() functions are generated in a fairly exotic way and would be good for a refresh/refactor.)