Would this be the proper method of declaring an XML element Foo that cannot have a child element or other content?

<xs:element name="Foo" type="xs:string" fixed="" nillable="true" />

The valid examples of this element in an XML document would be:

<Foo></Foo>

and

<Foo/>

Anything else would be invalid, e.g.:

<Foo>stuff</Foo>

It would be nice, though not essential, that the following be valid

<Foo>
</Foo>

• xml
• xsd
• schema

I combined a few options into a single schema and attempted to use it to validate some test elements.

Sources: http://www.w3schools.com/schema/schema_complex_empty.asp

http://www.herongyang.com/XML-Schema/complexType-Empty-Element-Declaration.html

The schema:

<?xml version="1.0" encoding="ISO-8859-1" ?>
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema">
  <xs:element name="root">
    <xs:complexType>
      <xs:choice maxOccurs="unbounded">
        <xs:element name="foo">
          <xs:complexType></xs:complexType>
        </xs:element>
        <xs:element name="y1">
          <xs:simpleType>
            <xs:restriction base="xs:string">
              <xs:length value="0" />
            </xs:restriction>
          </xs:simpleType>
        </xs:element>
        <xs:element name="y2">
          <xs:complexType>
            <xs:complexContent>
              <xs:restriction base="xs:anyType" />
            </xs:complexContent>
          </xs:complexType>
        </xs:element>
      </xs:choice>
    </xs:complexType>
  </xs:element>
</xs:schema>

Test cases:

<?xml version="1.0" encoding="UTF-8"?>
<root xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
xsi:noNamespaceSchemaLocation="NewXMLSchema.xsd">
 <!--  These six are all valid -->
  <foo />
  <y1 />
  <y2 />
  <foo></foo>
  <y1></y1>
  <y2></y2>
  <!-- These three are all invalid! -->
  <foo>
  </foo>
  <y1>
  </y1>
  <y2>
  </y2>
  <!--  These are invalid too. -->
  <foo>invalid</foo>
  <y1>invalid</y1>
  <y2>invalid</y2>
</root>

It looks like any of the three element declarations would do what you want, except for the extra blank line feature.

I wouldn't include type="xs:string", fixed="", or nillable="true" because these are all semantically different from an empty tag. An empty tag wouldn't really have a type (like string) because it's empty. It also wouldn't have a fixed value of the null string, because that's different from being empty. A nillable element is semantically different from an empty tag because it is a semantically equivalent to the tag being absent (Source: XSLT 2.0 and XSLT 2.0 Programmer's Reference by Michael Kay, p. 182.). An empty tag like <br/> does not mean that, in this case, the line break is absent. It means that the line break is present but has no content or attriubtes.

There are two approaches to defining a type that enforces empty content: one is to use a simple type of xs:string which is constrained using an enumeration or pattern facet (or using fixed) to be zero-length; the other is to use a complex type defined as (say) an empty sequence of elements. For validation purposes, there's not much to choose between them; but if you're interested in doing more advanced things with your schema, like defining type hierarchies and substitution groups, or mapping the schema to a Java object model, it could make a difference which you choose.

<xs:element name="Foo">
  <xs:complexType>
    <xs:attribute name="Bar" type="xs:positiveInteger"/>
  </xs:complexType>
</xs:element>

You could leave out the attribute tag to leave it totally empty...