Updated d-t-m lib, with optimizations and bugfixes

h/lib/dom_text_mapper.coffee

@@ -93,7 +93,7 @@ class window.DomTextMapper
     startTime = @timestamp()
     @saveSelection()
     @path = {}
-    @traverseSubTree @pathStartNode
+    @traverseSubTree @pathStartNode, @getDefaultPath()
     t1 = @timestamp()
 #    console.log "Phase I (Path traversal) took " + (t1 - startTime) + " ms."
 
@@ -148,7 +148,7 @@ class window.DomTextMapper
         delete @path[p]        
         
 #      console.log "Done. Collecting new path info..."
-      @traverseSubTree node
+      @traverseSubTree node, path
 
 #      console.log "Done. Updating mappings..."
 
@@ -369,28 +369,34 @@ class window.DomTextMapper
       when "#cdata-section" then return "cdata-section()"
       else return nodeName
 
+  getNodePosition: (node) ->
+    pos = 0
+    tmp = node
+    while tmp
+      if tmp.nodeName is node.nodeName
+        pos++
+      tmp = tmp.previousSibling
+    pos
+
+  getPathSegment: (node) ->
+    name = @getProperNodeName node
+    pos = @getNodePosition node
+    name + (if pos > 1 then "[#{pos}]" else "")
+
   getPathTo: (node) ->
     xpath = '';
     while node != @rootNode
-      pos = 0
-      tempitem2 = node
-      while tempitem2
-        if tempitem2.nodeName is node.nodeName
-          pos++
-        tempitem2 = tempitem2.previousSibling
-
-      xpath = (@getProperNodeName node) +
-          (if pos>1 then "[" + pos + ']' else "") + '/' + xpath
+      xpath = (@getPathSegment node) + '/' + xpath
       node = node.parentNode
     xpath = (if @rootNode.ownerDocument? then './' else '/') + xpath
     xpath = xpath.replace /\/$/, ''
     xpath
 
   # This method is called recursively, to traverse a given sub-tree of the DOM.
-  traverseSubTree: (node, invisible = false, verbose = false) ->
+  traverseSubTree: (node, path, invisible = false, verbose = false) ->
     # Step one: get rendered node content, and store path info,
     # if there is valuable content
-    path = @getPathTo node
+    @underTraverse = path
     cont = @getNodeContent node, false
     @path[path] =
       path: path
@@ -414,7 +420,8 @@ class window.DomTextMapper
     # A: I seem to remember that the answer is yes, but I don't remember why.
     if node.hasChildNodes()
       for child in node.childNodes
-        @traverseSubTree child, invisible, verbose
+        subpath = path + '/' + (@getPathSegment child)
+        @traverseSubTree child, subpath, invisible, verbose
     null
 
   getBody: -> (@rootWin.document.getElementsByTagName "body")[0]
@@ -455,7 +462,9 @@ class window.DomTextMapper
 
   # Select the given node (for visual identification),
   # and optionally scroll to it
-  selectNode: (node, scroll = false) ->  
+  selectNode: (node, scroll = false) ->
+    unless node?
+      throw new Error "Called selectNode with null node!"
     sel = @rootWin.getSelection()
 
     # clear the selection
@@ -506,12 +515,18 @@ class window.DomTextMapper
           # If this is the case, then it's OK.
           unless USE_EMPTY_TEXT_WORKAROUND and @isWhitespace node
             # No, this is not the case. Then this is an error.
-            throw exception
+            console.log "Warning: failed to scan element @ " + @underTraverse
+            console.log "Content is: " + node.innerHTML
+            console.log "We won't be able to properly anchor to any text inside this element."
+#            throw exception
     if scroll
       sn = node
-      while not sn.scrollIntoViewIfNeeded?
+      while sn? and not sn.scrollIntoViewIfNeeded?
         sn = sn.parentNode
-      sn.scrollIntoViewIfNeeded()
+      if sn?
+        sn.scrollIntoViewIfNeeded()
+      else
+        console.log "Failed to scroll to element. (Browser does not support scrollIntoViewIfNeeded?)"
     sel
 
   # Read and convert the text of the current selection.
@@ -661,4 +676,13 @@ class window.DomTextMapper
 
   # Decides whether a given node is a text node that only contains whitespace
   isWhitespace: (node) ->
-    node.nodeType is Node.TEXT_NODE and WHITESPACE.test node.data
+    result = switch node.nodeType
+      when Node.TEXT_NODE
+        WHITESPACE.test node.data
+      when Node.ELEMENT_NODE
+        mightBeEmpty = true
+        for child in node.childNodes
+          mightBeEmpty = mightBeEmpty and @isWhitespace child
+        mightBeEmpty
+      else false
+    result