SpeexX Ocean annotation XML Description

Namespace

The SpeexX Ocean annotation namespace is


        http://www.speexx.de/ocean/ns/annotation/1.0/#

It is recommended, that the namespace must be declared in the XML root element. It is also recommended, that the namespace is used as base namespace for so called nested annotations with equals namespace and localname.

Example:
Nested Annotations

If such an annotation comes up, the namespace is appended by a count (base 36 encoded), followed by a numer sign (#), followed by the original namespace.

Example:
http://www.speexx.de/ocean/ns/annotation/1.0/#count#original-namespace

The `a:annotation` container element attributes

The root element of all annotations is annotation. This element can contains the following attributs.

a:mimetype - contains the mimetype of the a:source. A a:mimetype attribute without a a:source azzribute is not usefull. (optional)
a:source - contains the URI of the source of the text. (optional)
a:version - contains the version of the serialized format. The default value is 1. (optional)
xml:lang - the global language of the content. (optional)

a:annotation has two nested elements: a:meta and a:text

The `a:meta` element.

The a:meta contains meta information about the document. The elements contains one or more a:property elements.
Note: At this time the elements is not yet supported by the API.

The `a:property` element.

The a:property element contains the value of a meta information.
Note: At this time the element is not supported by the API.

The attributes of `a:property`.

a:key - contains the key of the property. (required)
a:value - contains the value of the property. (required)

The `a:text` element.

The a:text contains the textvalue of the document. The value of xml:space is allways preserve. The a:text could contain ANY other XML elements.

Attributes for `a:text`.

There are three defined attributes which can be used in all annotations.

xml:space - Fixed definied with preserve. If there are additional attributes in other vendor annotations, there value will be ignored.

Common attributes for ANY

There are three defined attributes which can be used in all annotations.

a:id - Defines the internal ID of an annotation. This is used to help storing overlapped annotations. Overlapped annotations are overlaped annotations. This is not supported in XML, only in SGML. With a:id and its counterpart a:ref its possible to handle this overlapping. An a:id is recommended if there is a a:ref attribute in the differend element. (optional)
a:ref - The counterpart of a:id to handle overlapping annotations. (optional)
a:weight - Defines the weight of an annotation. This attribute is vendor dependend of third party namespaces. Any other namespace for annotation can define its own weight attributes. But cause this attribute is defined as a signed 32 bit integer with a normal range from 0 - 100 this attribute is prepared to store as number in a database. 100 is the default value.

An example DTD

This XML DTD is only an schematic representation of how an annotation format maybe looks like.

  <!ELEMENT a:annotation (a:meta, a:text) >
  <!ATTLIST a:annotation a:version CDATA #FIXED '1'
                         a:source CDATA #IMPLIED
                         a:mimetype CDATA #IMPLIED
                         xmlns:a CDATA #FIXED 'http://www.speexx.de/ocean/ns/annotation/1.0/#'
                         xml:lang NMTOKEN #IMPLIED>

  <!ELEMENT a:meta (a:property)* >
  <!ELEMENT a:property (EMPTY) >
  <!ATTLIST a:property a:key CDATA #REQUIRED
                       a:value CDATA #REQUIRED>
  <!ELEMENT a:text (#PCDATA | ANY)* >
  <!ATTLIST a:text xml:space NMTOKEN #FIXED 'preserve' >
  <!ATTLIST ANY a:weight CDATA #IMPLIED
                a:id ID #IMPLIED
                a:ref IDREF #IMPLIED>