Fix incorrect ValueError introduced in #537c645.

The underlying cause is the poor API of NBT: rather than having multiple factory methods, there is just one __init__ which takes a plethora of arguments, and tries to figure out how to instantiate an object (from value, from buffer, etc.). The additional logic required to make this work is causing too much code and too much bugs.

nbt/nbt.py

@@ -358,8 +358,8 @@ def __init__(self, type=None, value=None, name=None, buffer=None):
         self.tags = []
         if buffer:
             self._parse_buffer(buffer)
-        if self.tagID == None:
-            raise ValueError("No type specified for list: %s" % (name))
+        # if self.tagID == None:
+        #     raise ValueError("No type specified for list: %s" % (name))
 
     #Parsers and Generators
     def _parse_buffer(self, buffer):
@@ -447,7 +447,7 @@ def _parse_buffer(self, buffer):
                 try:
                     tag = TAGLIST[type.value]()
                 except KeyError:
-                    raise ValueError("Unrecognised tag type")
+                    raise ValueError("Unrecognised tag type %d" % (type.value))
                 tag.name = name
                 self.tags.append(tag)
                 tag._parse_buffer(buffer)