xml processing

xmlprocess.R

library(xml2)
library(tidyverse)

## Extract any number of xml files
list.files(pattern="*.xml") %>% 
  map(read_xml) %>% 
  ## Convert to a list to process
  map(as_list) %>%
  ## Simplify list structure
  map(flatten) %>%
  ## Turn into one big data frame
  map_df(as_data_frame) %>% 
  ## Handle missing request info
  mutate(Request=ifelse(Request=="NULL",NA,
                        Request)) ->
  highleveldata

## Convert embedded XML into XML nodes
xml_handler<-function(x) {
  as_xml_document(
    paste0("<node>",ifelse(is.na(x),"<CustomerId>NA</CustomerId>",x),"</node>"))
}

highleveldata %>% 
  ## Extract Requests as a vector
  pluck("Request") %>% 
  map(xml_handler) %>% 
  map(as_list) %>%
  map(flatten) %>%
  map_df(as_data_frame)  ->
  requestdata

## If the same number of rows matches i.e. didn't screw anything up, combine the two dataframes
if(nrow(highleveldata)==nrow(requestdata)) {
  alldata<-cbind(highleveldata, requestdata)
}  

## Et voila
View(alldata)