DTDdriven

DTDdriven denotes an approach in which the structure and constraints of XML data are defined by a Document Type Definition (DTD) and used as the primary reference for data modeling, validation, and processing. In a DTDdriven workflow, the DTD serves as the contract that producers and consumers of XML documents agree to follow.

Originally developed in the SGML ecosystem and carried into XML, DTDs specify element declarations, attribute lists, entities, and simple content models. Validation against the DTD ensures that documents conform to the defined structure before they are parsed by applications. DTDdriven practices contrast with approaches that rely on XML Schema, Relax NG, or other schema languages that provide richer typing and namespace support.

In practice, DTDdriven development may involve publishing a DTD to define the allowed document structure, generating test data from the DTD, and configuring validators to reject nonconforming documents. It is common in legacy integration scenarios, document-centric data exchange, or environments where simple XML with minimal typing suffices.

Advantages include simplicity, broad tool support, and a self-contained validation mechanism that does not require additional schema tooling. Limitations include limited data typing (generally only CDATA in attributes and elements), lack of robust namespace support in traditional DTDs, and difficulty expressing complex constraints or evolving schemas; many modern XML pipelines favor XML Schema or RELAX NG.

Example: a minimal DTD fragment for a library catalog could be:

<!DOCTYPE library [

<!ELEMENT library (book*)>

<!ELEMENT book (title, author, year?)>

<!ELEMENT title (#PCDATA)>

<!ELEMENT author (#PCDATA)>

<!ELEMENT year (#PCDATA)>

]>

This snippet shows how the DTD defines the allowed structure and content of the documents in a DTDdriven system.