Creating a DTD
Why would you want to create a DTD?
The benefits of DTDs are that it allows you to create numerous documents and make sure that the information contained in them will be comparable. For example, all the information about dates are in tags called <date> rather than <time>, <dates>, <Date> or <DATE>. By creating XML documents that meet a DTD's requirements, you can also share information between institutions.
Here's a "real life" example of using DTDs. The Society of American Archivists and the Library of Congress created Encoded Archival Description (EAD) for the purpose of encoding finding aids. The first version of EAD adhered to SMGL standards; however, with the popularity of XML, newer versions are XML-compliant.
Archival institutions like the Center for American History or Nettie Lee Benson Latin American Collection have created their finding aids using an EAD DTD. Automatic harvesters can use these finding aids to generate an online catalog of what is available at these archives. For an example of this, see the Texas Archival Resources Online or TARO. TARO is a collection of all the finding aids of archival repositories in Texas that are encoded in EAD.
Note: This section shows you how to create an external DTD file. However, DTDs can also be placed internally in an XML document.
Rules for Creating DTDs
When creating a DTD, you need to define all the elements and attributes you'll have in the XML documents. So let's create a DTD for our message XML documents.
Some syntax to remember when creating DTDs are the following:
| Symbol | Meaning | Example |
|---|---|---|
, |
AND |
header (sender,
recipient*, date) |
| |
OR |
message (email | letter) |
() |
Occurs only Once |
(email | letter) |
+ |
must occur at least
once |
(header, subject?, text+)
|
? |
occurs either once or not at all |
(header, recipient* , date?) |
* |
can occur
zero or more times |
(sender,
recipient*, date) |
Elements are declared in the following manner:
<!ELEMENT elementName ( elementParts ) >
Attributes are declared like this:
<!ATTLIST elementName attributeName attributeType attributeDefault >
So when creating a DTD for our message XML files, we would have something like this:
<!ELEMENT message ( email | letter ) >
<!ELEMENT letter ( letterhead, text ) >
<!ELEMENT email (header, subject?, text+) >
<!ATTLIST letter reply ( yes | no ) "no" >
<!ATTLIST email reply ( yes | no ) "no" >
<!ELEMENT header ( sender, recipient*, date?) >
<!ELEMENT subject ( #PCDATA) >
<!ELEMENT letterhead ( sender, recipient*, date ) >
<!ELEMENT sender ( #PCDATA ) >
<!ELEMENT recipient ( #PCDATA ) >
<!ELEMENT date ( #PCDATA ) >
<!ELEMENT text ( #PCDATA | salutation )* >
<!ELEMENT salutation ( #PCDATA ) >
(click on image for larger view)
An important thing to remember when making DTDs is that unless you use | when defining the element parts, the order of the elements you have within that section is required in your XML. So in a letter, the element <letterhead> must occur before the element <text >.
There are some tools that will automatically generate DTDs from XML. HitSoftware, a W3Group member, has created this XML to DTD tool. So now that you've created a DTD, how do you validate your XML against it?
Watch the video
Choose format/speed:
dial-up | broadband
dial-up | broadband
Entire tutorial (with captions)
