XSLT 2.0 and XPath 2.0 Programmer's Reference, 4th Edition (157 page)

It is an error if the resulting sequence contains an attribute or namespace node.

Finally, the nodes in the sequence are attached to the new document node as its children. Officially, this involves making a deep copy of each node: This is because nodes in the data model are immutable, so you cannot change the parent of an existing node. In practice, making a copy at this stage is very rarely necessary, because in most cases the node being attached has only just been created and will never be used independently of its new parent. The only case where it is necessary is where the result sequence contains references to existing nodes, which can be produced using the

instruction:

Even in this case, creating new nodes can be avoided if the result tree is to be immediately serialized.

Validating and Annotating the Document

The
validation
and
type
attributes control whether and how the new document is validated. They are available only if you are using a schema-aware processor. As usual, validation has two effects: it triggers a failure if the result document is invalid according to the schema, and it creates type annotations on the nodes in the tree.

The
validation
attribute has the same four values as on other elements:
strip
,
preserve
,
strict
, and
lax
. If the
validation
attribute is not specified, then the default is provided by the
default-validation
attribute on the

element, which in turn defaults to
strip
.

• strip
  removes all type annotations, replacing them with
  xs:untyped
  for elements and
  xs:untypedAtomic
  for attributes.
  
• preserve
  leaves the type annotations as they are, determined by the
  validation
  and
  type
  attributes on the instructions that created individual elements and attributes.
  
• strict
  and
  lax
  firstly check that the tree represents a well-formed XML document, that is, that the children of the document node comprise exactly one element node, no text nodes, and any number of comments and processing instructions. If not, a failure is reported. Then the top-level element node (the document element) is validated against the schema definitions, with the difference being that
  lax
  validates an element only if a schema definition for that element can be found, while
  strict
  fails if no declaration of this element can be found.
  
• Validation of a document node works in the same way as validation of individual elements but with one important exception: ID/IDREF constraints defined in the schema are checked when validation is done at the document level but not when it is done at element level. This involves checking that
  xs:ID
  values are unique and that
  xs:IDREF
  and
  xs:IDREFS
  values reference an ID value somewhere in the document.