Tidyeval function, dplyr map mutate statement with xml2 function

[bmar]

I have to work with multiple xml files (and some json - but for now xml) of different depths and variable structure. Much of the information I need is stored in attributes. I am looking for the most elegant solution which can be adapted by users who are not going to go in deep with R programming. They would know the basic xml sections and key attributes and would use the data frame exported to Excel ( or maybe rendered in Shiny or a markdown document).

library(tidyverse)
library(rlang)
library(xml2)
library(reprex)

my_xml_file <- "<results>
          <result name='math' status='pass' person='winkin'>90</result> 
          <result name='lit' status='fail' person='blinkin'>60</result>
          <result name='music' status='incl' person='nod'>absent</result>
          </results>"

my_xml <- read_xml(my_xml_file)
my_xml_list = xml_find_all(my_xml,'//result')
# process xml as in Jenny Bryan example
rows_df <- tibble(row=seq_along(my_xml_list), nodeset=my_xml_list)
# extract info from xml to build more usable dataframe
cells_df <- rows_df %>%
  mutate(name = nodeset %>% map_chr(~xml_attr(.,'name')))

# can (most of) the following be done using tidyeval?
cells_df <- rows_df %>%
  mutate(name = nodeset %>% map_chr(~xml_attr(.,'name'))) %>%
  mutate(status = nodeset %>% map_chr(~xml_attr(.,'status'))) %>%
  mutate(person = nodeset %>% map_chr(~xml_attr(.,'person'))) %>%
  select(-nodeset)

# want a function to build 1 mutate statement to get each enumerated var 
# how much tidyeval to use - is cement() from Lionel Henry webinar useful

my_fun <- function(.data, var){
  var_as_txt <- quo_name(enquo(var))
  .data %>%
    mutate( {{var}} := nodeset %>% map_chr(~xml_attr(., !!var_as_txt ) )) %>%
    select(-nodeset) }

# That works, but would like to it be 
#  1) more elegant - proper use of tidyeval functions and
#  2) dplyr/purrr - take multiple vars and keep results in 1 dataframe
my_fun(rows_df, status)
#> # A tibble: 3 x 2
#>     row status
#>   <int> <chr> 
#> 1     1 pass  
#> 2     2 fail  
#> 3     3 incl
my_fun(rows_df, name)
#> # A tibble: 3 x 2
#>     row name 
#>   <int> <chr>
#> 1     1 math 
#> 2     2 lit  
#> 3     3 music
my_fun(rows_df, person)
#> # A tibble: 3 x 2
#>     row person 
#>   <int> <chr>  
#> 1     1 winkin 
#> 2     2 blinkin
#> 3     3 nod

# looking to have something like this:
# tbl_of_xml_nodesets %>% get_identified_attributes(attribute1, attribute2, attribute3, etc.)