method

document_order

v2_1_10 - Show latest stable - 0 notes - Class: REXML::XPathParser

document_order( array_of_nodes ) private

Reorders an array of nodes so that they are in document order It tries to do this efficiently.

FIXME: I need to get rid of this, but the issue is that most of the XPath interpreter functions as a filter, which means that we lose context going in and out of function calls. If I knew what the index of the nodes was, I wouldn’t have to do this. Maybe add a document IDX for each node? Problems with mutable documents. Or, rewrite everything.

Show source

# File lib/rexml/xpath_parser.rb, line 566
    def document_order( array_of_nodes )
      new_arry = []
      array_of_nodes.each { |node|
        node_idx = []
        np = node.node_type == :attribute ? node.element : node
        while np.parent and np.parent.node_type == :element
          node_idx << np.parent.index( np )
          np = np.parent
        end
        new_arry << [ node_idx.reverse, node ]
      }
      #puts "new_arry = #{new_arry.inspect}"
      new_arry.sort{ |s1, s2| s1[0] <=> s2[0] }.collect{ |s| s[1] }
    end

document_order

Related methods