library(XML)


#Read an xml file into an R list which is formatted like a tree
xmlTreeParse("~/Stat133/lectures/XML/matrix.xml")

x = xmlTreeParse("~/Stat133/lectures/XML/matrix.xml")
# Note that what prints on the screen is an xml document,
# but what is stored internally is a list of class XMLDocument

class(x)
mode(x)
names(x)

# Explore a bit to find that the main document is 
x$doc
x$dtd

names(x$doc)

x$doc["file"]
x$doc["version"]
x$doc["children"]

# This function is very handy for pulling out 
# the root of the xml document

xtree = xmlRoot(x)
class(xtree)

# There are several functions provided by the XML package
# These operate on XMLNode objects

xmlName(xtree)
xmlAttrs(xtree)
xmlChildren(xtree)

class(xmlChildren(xtree))
xmlSize(xmlChildren(xtree))

# There is a special apply function in the XML package
# that is for XMLNodes
xmlSApply(xtree, xmlSize)


xtree[[1]]
class(xtree[[1]])
xmlSize(xtree[[1]])
xmlName(xtree[[1]])
xmlChildren(xtree[[1]])

xmlValue(xtree[[1]][[1]])

xmlApply(xtree, function(node) 
       if(xmlName(node) == "row") xmlAttrs(node)[["id"]])

rn = xmlApply(xtree, function(node) if(xmlName(node) == "row") 
       xmlAttrs(node)[["id"]])

rn = unlist(rn)


cn  = xmlSApply(xtree[["colNames"]], xmlValue)
dm = xmlAttrs(xtree)
numrow = as.integer(dm[["nrow"]])
numcol = as.integer(dm[["ncol"]])

 
xtree[[2]]
xmlSapply(xtree[[2]], xmlValue)

# Note we have to use sapply here because xtree[-1] is not an XMLNode
matData = sapply(xtree[-1], function(node) { if (xmlName(node) =="row") xmlSApply(node, xmlValue)})

matData
as.numeric(matData)