The join operations causes R to crash, when tibble column names are messed up

[JLYJabc]

this chunk of code gives a nice error message

df_a <- data.frame(A = 1:3, B = c("a", "b", "c"))
names(df_a) <- c("AA")
df_b <- data.frame(A = 2:4, C = c("aa", "bb", "cc"))
names(df_b) <- c("AA")
df_ab <- dplyr::full_join(df_a, df_b, by = "AA")

Error: Column 2 must be named

wheras these lines causes R to crash

df_a <- tibble::tibble(A = 1:3, B = c("a", "b", "c"))
names(df_a) <- c("AA")
df_b <- tibble::tibble(A = 2:4, C = c("aa", "bb", "cc"))
names(df_b) <- c("AA")
df_ab <- dplyr::full_join(df_a, df_b, by = "AA")

(It also crashes with base::data.frame())

[batpigandme]

Could you please turn this into a self-contained reprex (short for minimal reproducible example)? It will help us help you if we can be sure we're all working with/looking at the same stuff.

If you've never heard of a reprex before, you might want to start by reading the tidyverse.org help page. The reprex dos and don'ts are also useful.

[JLYJabc]

I've updated the post, let me know if it is ok now?
(first time doing this)

[batpigandme]

Hi @JLYJabc,

OK, so I sent you on a bit of a goose chase, because (since it causes a crash, or, for me, an endless process), it's un-reprex-able.

So, here's a lengthy reprex, printing all but that final step (which isn't in the reprex because, well, see above).

df_a <- data.frame(A = 1:3, B = c("a", "b", "c"))
df_a
#>   A B
#> 1 1 a
#> 2 2 b
#> 3 3 c
names(df_a) <- c("AA")
df_a
#>   AA NA
#> 1  1  a
#> 2  2  b
#> 3  3  c
df_b <- data.frame(A = 2:4, C = c("aa", "bb", "cc"))
df_b
#>   A  C
#> 1 2 aa
#> 2 3 bb
#> 3 4 cc
names(df_b) <- c("AA")
df_b
#>   AA NA
#> 1  2 aa
#> 2  3 bb
#> 3  4 cc
df_ab <- dplyr::full_join(df_a, df_b, by = "AA")
#> Error: Column 2 must be named

df_a <- tibble::tibble(A = 1:3, B = c("a", "b", "c"))
names(df_a) <- c("AA")
df_a
#> # A tibble: 3 x 2
#>      AA `NA` 
#>   <int> <chr>
#> 1     1 a    
#> 2     2 b    
#> 3     3 c
df_b <- tibble::tibble(A = 2:4, C = c("aa", "bb", "cc"))
names(df_b) <- c("AA")
df_b
#> # A tibble: 3 x 2
#>      AA `NA` 
#>   <int> <chr>
#> 1     2 aa   
#> 2     3 bb   
#> 3     4 cc

Created on 2018-03-13 by the reprex package (v0.2.0).

After this, the line below ⇨ wonkiness…

df_ab <- dplyr::full_join(df_a, df_b, by = "AA")

I don't have a solution yet, just reproducing for now!

[romainfrancois]

I can reproduce the endlessness.

[romainfrancois]

➡️ https://github.com/tidyverse/dplyr/blob/master/src/join_exports.cpp#L83

[krlmlr]

I'm working on it as part of #3307.

[krlmlr]

In 0.7.4 I'm seeing dplyr caught in an endless loop with the original example. This means we can safely raise an error here without breaking "compatibility".