I have been looking at the following specification to try and get a better understanding of how the mime type for binary content is specified in the schema and in the actual document transmitted: -
My interpretation is that the 'expectedContentTypes' attribute is an annotation which is added to the XML schema to list the mime types that can be represented by the binary content.
My interpretation of the 'contentType' is that this is used on the document instance itself (not the schema) to specify what is the actual mime type of the binary data being transmitted.
So in the schema 'expectedContentTypes' can be used to list a set of mime types and then in the document 'contentType' is used to specify what is being transmitted in this document instance.
However in the user guide and in some test cases we seem to be using 'contentType' in the place of 'expectedContentTypes' or in addition to 'expectedContentTypes' in the schema.
This seems incorrect to me, should it really be possible to use 'contentType' in the schema?
You are correct. Also the expectedContentType value, if present is ignored, since the xerces schema API can't represent it.