Fixes #5: namespace prefixes are applied on the element/attribute dec…

…laring them

README.md

@@ -32,12 +32,12 @@ local myxml = io.open('my.xml'):read('*all')
 
 -- Specify as many/few of these as you like
 parser = SLAXML:parser{
-  startElement = function(name,nsURI)       end, -- When "<foo" or <x:foo is seen
-  attribute    = function(name,value,nsURI) end, -- attribute found on current element
-  closeElement = function(name,nsURI)       end, -- When "</foo>" or </x:foo> or "/>" is seen
-  text         = function(text)             end, -- text and CDATA nodes
-  comment      = function(content)          end, -- comments
-  pi           = function(target,content)   end, -- processing instructions e.g. "<?yes mon?>"
+  startElement = function(name,nsURI,nsPrefix)       end, -- When "<foo" or <x:foo is seen
+  attribute    = function(name,value,nsURI,nsPrefix) end, -- attribute found on current element
+  closeElement = function(name,nsURI)                end, -- When "</foo>" or </x:foo> or "/>" is seen
+  text         = function(text)                      end, -- text and CDATA nodes
+  comment      = function(content)                   end, -- comments
+  pi           = function(target,content)            end, -- processing instructions e.g. "<?yes mon?>"
 }
 
 -- Ignore whitespace-only text nodes and strip leading/trailing whitespace from text
@@ -86,8 +86,8 @@ The returned table is a 'document' comprised of tables for elements, attributes,
   * <strong>`someAttr.type`</strong> : the string `"attribute"`
   * <strong>`someAttr.name`</strong> : the name of the attribute (without any namespace prefix)
   * <strong>`someAttr.value`</strong> : the string value of the attribute (with XML and numeric entities unescaped)
-  * <strong>`someEl.nsURI`</strong> : the namespace URI for the attribute; `nil` if no namespace is applied
-  * <strong>`someEl.parent`</strong> : reference to the the parent element table
+  * <strong>`someAttr.nsURI`</strong> : the namespace URI for the attribute; `nil` if no namespace is applied
+  * <strong>`someAttr.parent`</strong> : reference to the the owning element table
 * **Text** - for both CDATA and normal text nodes
   * <strong>`someText.type`</strong> : the string `"text"`
   * <strong>`someText.name`</strong> : the string `"#text"`
@@ -157,6 +157,10 @@ In this case no table will have a `parent` attribute, elements will not have the
 
 ## History
 
+### v0.6 2014-Apr-18
++ Fixes Issue #5 (and more): Namespace prefixes defined on element are now properly applied to the element itself and any attributes using them when the definitions appear later in source than the prefix usage.
++ The streaming parser now supplies the namespace prefix for elements and attributes.
+
 ### v0.5.3 2014-Feb-12
 + Fixes Issue #3: The [reserved `xml` prefix](http://www.w3.org/TR/xml-names/#ns-decl) may be used without pre-declaring it. (Thanks David Durkee.)
 

slaxml.lua

@@ -1,26 +1,34 @@
 --[=====================================================================[
-v0.5.3 Copyright © 2013 Gavin Kistner <[email protected]>; MIT Licensed
+v0.6 Copyright © 2013-2014 Gavin Kistner <[email protected]>; MIT Licensed
 See http://github.com/Phrogz/SLAXML for details.
 --]=====================================================================]
 local SLAXML = {
-	VERSION = "0.5.3",
+	VERSION = "0.6",
 	_call = {
 		pi = function(target,content)
 			print(string.format("<?%s %s?>",target,content))
 		end,
 		comment = function(content)
 			print(string.format("<!-- %s -->",content))
 		end,
-		startElement = function(name,nsURI)
-			print(string.format("<%s%s>",name,nsURI and (" ("..nsURI..")") or ""))
+		startElement = function(name,nsURI,nsPrefix)
+			                 io.write("<")
+			if nsPrefix then io.write(nsPrefix,":") end
+			                 io.write(name)
+			if nsURI    then io.write(" (ns='",nsURI,"')") end
+			                 print(">")
 		end,
-		attribute = function(name,value,nsURI)
-			print(string.format("  %s=%q%s",name,value,nsURI and (" ("..nsURI..")") or ""))
+		attribute = function(name,value,nsURI,nsPrefix)
+			io.write('  ')
+			if nsPrefix then io.write(nsPrefix,":") end
+			                 io.write(name,'=',string.format('%q',value))
+			if nsURI    then io.write(" (ns='",nsURI,"')") end
+			io.write("\n")
 		end,
 		text = function(text)
 			print(string.format("  text: %q",text))
 		end,
-		closeElement = function(name,nsURI)
+		closeElement = function(name,nsURI,nsPrefix)
 			print(string.format("</%s>",name))
 		end,
 	}
@@ -42,7 +50,7 @@ function SLAXML:parse(xml,options)
 	local textStart = 1
 	local currentElement={}
 	local currentAttributes={}
-	local currentAttributeCt
+	local currentAttributeCt -- manually track length since the table is re-used
 	local nsStack = {}
 
 	local entityMap  = { ["lt"]="<", ["gt"]=">", ["amp"]="&", ["quot"]='"', ["apos"]="'" }
@@ -94,13 +102,14 @@ function SLAXML:parse(xml,options)
 		anyElement = true
 		first, last, match1 = find( xml, '^<([%a_][%w_.-]*)', pos )
 		if first then
-			currentElement[2] = nil
+			currentElement[2] = nil -- reset the nsURI, since this table is re-used
+			currentElement[3] = nil -- reset the nsPrefix, since this table is re-used
 			finishText()
 			pos = last+1
 			first,last,match2 = find(xml, '^:([%a_][%w_.-]*)', pos )
 			if first then
 				currentElement[1] = match2
-				currentElement[2] = nsForPrefix(match1)
+				currentElement[3] = match1 -- Save the prefix for later resolution
 				match1 = match2
 				pos = last+1
 			else
@@ -137,12 +146,12 @@ function SLAXML:parse(xml,options)
 					nsStack[#nsStack][name] = match2
 				else
 					currentAttribute[1] = name
-					currentAttribute[3] = nsForPrefix(prefix)
+					currentAttribute[4] = prefix
 				end
 			else
 				if match1=='xmlns' then
 					nsStack[#nsStack]['!'] = match2
-					currentElement[2] = match2
+					currentElement[2]      = match2
 				end
 			end
 			currentAttributeCt = currentAttributeCt + 1
@@ -169,8 +178,15 @@ function SLAXML:parse(xml,options)
 			pos = last+1
 			textStart = pos
 
+			-- Resolve namespace prefixes AFTER all new/redefined prefixes have been parsed
+			if currentElement[3] then currentElement[2] = nsForPrefix(currentElement[3])    end
 			if self._call.startElement then self._call.startElement(unpack(currentElement)) end
-			if self._call.attribute then for i=1,currentAttributeCt do self._call.attribute(unpack(currentAttributes[i])) end end
+			if self._call.attribute then
+				for i=1,currentAttributeCt do
+					if currentAttributes[i][4] then currentAttributes[i][3] = nsForPrefix(currentAttributes[i][4]) end
+					self._call.attribute(unpack(currentAttributes[i]))
+				end
+			end
 
 			if match1=="/" then
 				pop(nsStack)

test/files/namespace_declare_and_use.xml

@@ -1,3 +1,14 @@
 <r>
-	<z:el xmlns:z="zoo">declaring and using a namespace on the same element is legal</z:el>
+	<cat xmlns="cat" xmlns:a="dog">
+		<cat />
+		<a:dog>
+			<cat />
+			<a:hog a:hog="yes" xmlns:a="hog" xmlns:b="bog" b:bog="yes">
+				<a:hog />
+				<b:bog />
+			</a:hog>
+			<a:dog />
+		</a:dog>
+	</cat>
+	<c:cog xmlns:c="cog" />
 </r>

test/test.lua

@@ -137,9 +137,38 @@ end
 
 function test_xml_namespace_immediate_use()
 	local doc = SLAXML:dom(XML['namespace_declare_and_use'])
-	assertEqual(#doc.root.el,1)
-	local z = doc.root.el[1]
-	assertEqual(z.nsURI,'zoo')
+	local cat1 = doc.root.el[1]
+	assertEqual(cat1.name,'cat')
+	assertEqual(cat1.nsURI,'cat')
+	local cat2 = cat1.el[1]
+	assertEqual(cat2.name, 'cat')
+	assertEqual(cat2.nsURI,'cat')
+	local dog1 = cat1.el[2]
+	assertEqual(dog1.name, 'dog')
+	assertEqual(dog1.nsURI,'dog')
+	local cat3 = dog1.el[1]
+	assertEqual(cat3.name, 'cat')
+	assertEqual(cat3.nsURI,'cat')
+	local hog1 = dog1.el[2]
+	assertEqual(hog1.name, 'hog')
+	assertEqual(hog1.nsURI,'hog')
+	for _,attr in ipairs(hog1.attr) do
+		if attr.value=='yes' then
+			assertEqual(attr.nsURI,attr.name)
+		end
+	end
+	local hog2 = hog1.el[1]
+	assertEqual(hog2.name, 'hog')
+	assertEqual(hog2.nsURI,'hog')
+	local bog1 = hog1.el[2]
+	assertEqual(bog1.name, 'bog')
+	assertEqual(bog1.nsURI,'bog')
+	local dog2 = dog1.el[3]
+	assertEqual(dog2.name, 'dog')
+	assertEqual(dog2.nsURI,'dog')
+	local cog2 = doc.root.el[2]
+	assertEqual(cog2.name, 'cog')
+	assertEqual(cog2.nsURI,'cog')
 end
 
 function test_dom_namespaces()