在R的递归函数中向data.frame添加新行

Question

I am trying to parse XML in R using the XML package by Duncan Temple Lang. The code I have is as follows:

library(XML)
retrieveStructureInfo <- function(node, tableData) {  
    tableD <- data.frame(path = NA, node = NA, value = NA)

    for (i in 1 : xmlSize(xmlAttrs(node))) {
      tableD <- rbind(tableD, c("path", "node", "value"))  
      tableData <<- rbind(tableData, tableD)    
    }

    #children is the no. of nodes within a node
    for (i in 1 : children) {
      #recursive function call
      retrieveStructureInfo(node[[i]], tableD) 
    }
}

#parse xml document
#xmlfile is the file path
doc <- xmlParse(xmlfile)
r <- xmlRoot(doc)
tableData <- data.frame(path = NA, node = NA, value = NA)
retrieveStructureInfo(r, tableData)
tableData

I am having issues adding rows to the data.frame because it is being done in a recursive function. For the XML given below, only the last two attribute values are added to the data.frame ie Source="b" and Available="true". I created a main table called tableData and try to update it with a local table within the function called tableD but it doesn't work.

<CATALOG>
   <PLANT>
      <COMMON Source="a" Available="false">Bloodroot</COMMON>
   </PLANT>
   <PLANT>
      <COMMON Source="b" Available="true">Columbine</COMMON>
   </PLANT>
</CATALOG>

I forgot to add that I am aiming to create a function that reads any xml (that is why I went with the idea of recursion) and gives an output:

                   path                 node                  value parent      type
  CATALOG/PLANT/COMMON               Source                    a    PLANT  attribute
  CATALOG/PLANT/COMMON            Available                  false  PLANT  attribute
  CATALOG/PLANT/COMMON               COMMON              Bloodroot  PLANT       text

Answer 1

Usually the answer to xml and recursion is to use xpath. You may be able to create a table with a few xpath queries or one of the helper functions like xmlToList.

x<- '<CATALOG>
   <PLANT>
      <COMMON Source="a" Available="false">Bloodroot</COMMON>
   </PLANT>
   <PLANT>
      <COMMON Source="b" Available="true">Columbine</COMMON>
   </PLANT>
</CATALOG>'

doc <- xmlParse(x)

xpathSApply(doc, "//COMMON", xmlValue)
[1] "Bloodroot" "Columbine"
xpathSApply(doc, "//COMMON", xmlGetAttr, "Source")
[1] "a" "b"

y <- xmlToList(doc)
data.frame(path=names(unlist(y)),value=unlist( y) )
                           path     value
1             PLANT.COMMON.text Bloodroot
2    PLANT.COMMON..attrs.Source         a
3 PLANT.COMMON..attrs.Available     false
4             PLANT.COMMON.text Columbine
5    PLANT.COMMON..attrs.Source         b
6 PLANT.COMMON..attrs.Available      true

library(plyr)
ldply(y, data.frame)  #OR 
ldply( y, function(x) data.frame(x, names(x$COMMON$.attrs)  ) )
    .id COMMON.text COMMON..attrs names.x.COMMON..attrs.
1 PLANT   Bloodroot             a                 Source
2 PLANT   Bloodroot         false              Available
3 PLANT   Columbine             b                 Source
4 PLANT   Columbine          true              Available