The basis for your content items in a CoreMedia CMS system is the XML file config/contentserver/doctypes/sample-doctypes.xml. It contains the root element <DocumentTypeModel> which serves as a container for the <DocType>, <XmlGrammar> and <XmlSchema> elements. It imports the grammar for CoreMedia rich text from the required-doctypes.xml file which is located in the framework/doctypes directory in cap-server.jar. The required-doctypes.xml contains the definitions which are essential for CoreMedia CMS. The next example shows a simple content type definition.

<ImportGrammar Name="coremedia-richtext-1.0"
      From="required-doctypes"/>

<DocType Name="article">
  <StringProperty Name="Author" Length="200"/>
  <StringProperty Name="Headline" Length="200"/>
  <DateProperty Name="Date"/>
  <XmlProperty Name="Summary" 
	Grammar="coremedia-richtext-1.0"/>
  <LinkListProperty Name="Images" LinkType="Image"/> 
</DocType>

<DocType Name="Image">
  <IntProperty Name="Width"/>
  <IntProperty Name="Height"/>
  <BlobProperty Name="Image" MimeType="image/*"/> 
</DocType>

Content type names

Every content type has a unique name, specified by the attribute Name. This name will be used by the Content Management Server to define unique data base identifiers like table names, primary key constraint names, foreign key constraint names and index names. The name has a maximum length of 11 characters, is not case-sensitive and can contain letters, numbers, underscore and dash but no umlauts or other special characters ("§", " &", ...). For a localization of content type and property names see Site Manager Developer Manual and Studio Developer Manual.

Caution

You are not allowed to use the following words for content type names:

• Names of SQL commands.

• Names which are reserved for CoreMedia as database table names:

• BlobCodeTable, BlobData, Blobs,ChangeLog, CmGroups, CmLicenses, CmProcessorUsage,CmRules, CmUserGroup, CmUsers, Dictionary, EditorPreferences, FolderIndex, LinkChangeLog, LinkLists, ObservedValues, ObservedValueChangeLog, MaxIds, MaxMemberIds, MySQLBlobStore, Preferences, Query, QueryIdTable, ReplicatorIdTable, Resources, SgmlData, SgmlGrammar, SgmlText, System, Texts, TrashIdTable, WfDefinitions, WfMaxDeletedId, WfPendingSignals, WfProcessInstances, WfProcesses, WfTaskInstances, WfTasks, WfVariables

• Names corresponding to the name scheme <doctype><digit><digit>"i"<char>*, where <doctype> stands for an already defined content type name, <digit> for a number and i<char>* for a word which begins with the letter "i".

If you have used content type names which exceeded the recommended length of 11 characters (which are longer than 15 characters to be exact) and you switch to another database you might exceed database limitations. This is especially true for IBM DB2. In order to circumvent these problems there exists three attributes for the DocumentType element.

Example:

Your content type name is "ProductAttributes" (<DocumentType name="ProductAttributes">). This name contains 17 characters. The CoreMedia Server will extend this name by up to three characters to obtain unique database identifiers. Unfortunately, the DB2 database has a limitation of 18 characters which will be exceeded by these names. In order to use your content type name furthermore you define unique names for the database using the attributes described above. Your new content type definition will look similar to the following example: <DocumentType name="ProductAttributes" PkName="pk_ProdAttr" ResourceFkName="fk_ProdAttr" IndexName="i_ProdAttr">

Three content types, Dictionary, Query and Preferences, are already defined in the required-doctypes.xml file. All types are essential for some components of CoreMedia CMS and must therefore neither be renamed nor deleted.

Defining the grammar

The <XmlGrammar> element with the attributes Name, Root, Parent and SystemId is used to refer to DTDs from XMLProperty elements. The following example defines the CoreMedia rich text grammar:

<XmlGrammar Name="coremedia-richtext-1.0"
      Root="div"
      PublicId="-//CoreMedia//DTD Rich Text 1.0//EN"
      SystemId="classpath:xml/coremedia-richtext-1.0.dtd"/>

The following <XmlGrammar> elements are predefined:

The latter four belong to the Dictionary, Query, EditorPreferences, and Preferences content types respectively. The grammar coremedia-struct-2008 is also used frequently in CoreMedia Blueprint.

If you want to use XML Schemas instead of DTDs you can specify the <XmlSchema> element with the attributes Name, SchemaLocation, Language and Parent. The following example defines the grammar custom-schema with a schema file custom.xsd located in the package my.pkg, which is available in a JAR file deployed with the Content Server and any other server which needs to validate the XML as for example the Workflow Server if actions access properties using this grammar:

<XmlSchema Name="custom-schema" 
            SchemaLocation="classpath:my/pkg/custom.xsd 
                            http://www.w3.org/1999/xlink.xsd"
            Language="http://www.w3.org/2001/XMLSchema">

You specify a DTD grammar or XML schema once in the content type declaration and can refer to it with the attribute Grammar in the XmlProperty element:

<DocType Name="Article">
    <StringProperty Name="Headline" Length="200"/>
    <XmlProperty    Name="Text"    
     Grammar="coremedia-richtext-1.0"/>
    <XmlProperty    Name="Summary" 
     Grammar="coremedia-richtext-1.0"/>
    <XmlProperty    Name="Comment" 
     Grammar="custom-schema"/>
</DocType>

Importing grammars and document types

You can import a grammar or schema into your content type file using the <ImportGrammar> element. The following example imports the "coremedia-richtext-1.0" grammar defined in the required-doctypes.xml file:

<ImportGrammar Name="coremedia-richtext-1.0"
From="required-doctypes"/>

You can also import content types that are defined in different content type files. This is necessary, if you want to use this content type as a parent or as a target of a LinkList property. The following example would import the type Article from the editorial-types.xml file.

<ImportDocType Name="Article"
From="editorial-types"/>

The attributes have the same meaning as for the ImportGrammar element.