untangle2(), by Jenny Bryan

[maurolepore]

Original idea by Luis Verde

https://rstudio-pubs-static.s3.amazonaws.com/287966_3967f466282b4260a0163d9d9acdad57.html

library(dplyr, warn.conflicts = FALSE)
library(tidyr)
library(rlang)

(dat <- tibble(
  jumble = c("Muridae", "diet:seeds", "Sp1", "Sp2", "Sp3",
    "diet:unknown", "Sp4", "Sp5", "Cricetidae", "diet:fruits", "Sp11",
    "Sp32", "Sp113")
))
#> # A tibble: 13 x 1
#>    jumble      
#>    <chr>       
#>  1 Muridae     
#>  2 diet:seeds  
#>  3 Sp1         
#>  4 Sp2         
#>  5 Sp3         
#>  6 diet:unknown
#>  7 Sp4         
#>  8 Sp5         
#>  9 Cricetidae  
#> 10 diet:fruits 
#> 11 Sp11        
#> 12 Sp32        
#> 13 Sp113

untangle2 <- function(df, regex, orig, new) {
  orig <- enquo(orig)
  new <- sym(quo_name(enquo(new)))

  df %>%
    mutate(
      !!new := if_else(grepl(regex, !! orig), !! orig, NA_character_)
    ) %>%
    fill(!! new) %>%
    filter(!grepl(regex, !! orig))
}

dat %>%
  untangle2("dae$", jumble, family) %>%
  untangle2("^diet", jumble, diet)
#> # A tibble: 8 x 3
#>   jumble family     diet        
#>   <chr>  <chr>      <chr>       
#> 1 Sp1    Muridae    diet:seeds  
#> 2 Sp2    Muridae    diet:seeds  
#> 3 Sp3    Muridae    diet:seeds  
#> 4 Sp4    Muridae    diet:unknown
#> 5 Sp5    Muridae    diet:unknown
#> 6 Sp11   Cricetidae diet:fruits 
#> 7 Sp32   Cricetidae diet:fruits 
#> 8 Sp113  Cricetidae diet:fruits

Created on 2018-10-06 by the reprex package (v0.2.1)