sort attributes at Element creation time to make their creation order…

… predictable

CHANGES.txt

@@ -40,6 +40,13 @@ Bugs fixed
 Other changes
 --------------
 
+* Element attributes passed in as an ``attrib`` dict or as keyword
+  arguments are now sorted by (namespaced) name before being created
+  to make their order predictable for serialisation and iteration.
+  Note that adding or deleting attributes afterwards does not take
+  that order into account, i.e. setting a new attribute appends it
+  after the existing ones.
+
 * Several classes that are for internal use only were removed
   from the ``lxml.etree`` module dict:
   ``_InputDocument, _ResolverRegistry, _ResolverContext, _BaseContext,

doc/objectify.txt

@@ -1122,8 +1122,8 @@ XSI type name:
     >>> print(objectify.dump(root))
     root = None [ObjectifiedElement]
         x = '5' [StringElement]
-          * py:pytype = 'str'
           * myattr = 'someval'
+          * py:pytype = 'str'
 
     >>> root.x = objectify.DataElement(5, _xsi="integer")
     >>> print(objectify.dump(root))

src/lxml/apihelpers.pxi

@@ -292,7 +292,7 @@ cdef _initNodeAttributes(xmlNode* c_node, _Document doc, attrib, extra):
             attrib.update(extra)
     if attrib:
         is_html = doc._parser._for_html
-        for name, value in attrib.items():
+        for name, value in sorted(attrib.items()):
             attr_ns_utf, attr_name_utf = _getNsTag(name)
             if not is_html:
                 _attributeValidOrRaise(attr_name_utf)