ISO/IEC JTC 1/SC34 N0267

ISO/IEC JTC 1/SC34

Information Technology --

Document Description and Processing Languages

Title:	The Ontopia Schema Language: Specification
Source:	Lars Marius Garshol
Project:	TMCL
Project editor:	Steve Pepper
Status:	Individual contribution
Action:	For information
Date:	2001-11-27
Summary:	A schema language for topic maps, presented as input for the TMCL process.
Distribution:	SC34 and Liaisons
Refer to:
Supercedes:
Reply to:	Dr. James David Mason (ISO/IEC JTC1/SC34 Chairman) Y-12 National Security Complex Information Technology Services Bldg. 9113 M.S. 8208 Oak Ridge, TN 37831-8208 U.S.A. Telephone: +1 865 574-6973 Facsimile: +1 865 574-1896 E-mailk: mailto:[email protected] http://www.y12.doe.gov/sgml/sc34/sc34oldhome.htm Ms. Sara Hafele, ISO/IEC JTC 1/SC 34 Secretariat American National Standards Institute 25 West 43rd Street New York, NY 10036 Tel: +1 212 642-4937 Fax: +1 212 840-2298 E-mail: [email protected]

Abstract

This document is the specification of the Ontopia Schema Language, providing an exhaustive reference documenting all features of the language. For an introduction to the schema language, see the schema tutorial. It is presented to SC34WG3 as input to the process of defining TMCL. Ontopia does not expect that this schema can be adopted as TMCL as-is, but is of the opinion that it requires further work.

Table of contents

• 1. Introduction
• 2. High-level elements
  • 2.1. The tm-schema element
  • 2.2. The ruleset element
  • 2.3. The topic element
  • 2.4. The baseName element
  • 2.5. The variant element
  • 2.6. The occurrence element
  • 2.7. The playing element
  • 2.8. The association element
  • 2.9. The role element
• 3. Low-level elements
  • 3.1. The ruleref element
  • 3.2. The instanceOf element
  • 3.3. The scope element
  • 3.4. The topicRef element
  • 3.5. The subjectIndicatorRef element
  • 3.6. The internalTopicRef element
  • 3.7. The any element
• 4. The DTD

1. Introduction

The Ontopia Schema Language uses an XML syntax defined in a DTD to specify schemas. Schemas written in the language consist of definitions of classes of topics and associations, defining the allowed characteristics of each. Validating a topic map against a schema is done by finding the class definition for each topic and association, and verifying that the instance matches the class definition.

The Ontopia Schema Language has been designed to have a minimal number of features and a minimum of expressive power. We expect to extend its feature set according to user demand, and eventually to replace it with the Topic Map Constraint Language (TMCL) currently being developed within ISO. We have decided not to implement the TMCL before it is finished in order to have a stable schema language as the basis for our product suite.

The Ontopia Schema Language is intended to support many functions within topic map projects, the most important of which are:

• validation of the structure of topic map instances,

• documentation of the structure of topic maps, and

• guidance of topic map editing applications, in order to make topic map editing easier and more convenient.

2. High-level elements

This section defines the semantics of the high-level elements of the Ontopia Schema Language, element by element.

2.1. The tm-schema element

Summary:	The container element for the schema.
Content model:	(ruleset | topic | association)*
Parents:	None

The tm-schema element must be the document element of all topic map schemas. It has no special semantics, but merely acts as a container for the topic map schema.

The tm-schema element has a single attribute: match, which can have the values loose and strict. When doing loose matching, topics and associations which match none of the specified classes are accepted without complaint. When doing strict matching, such topics and associations are rejected. The default is loose matching.

The allowed children of tm-schema elements are ruleset, topic, and association elements, in any order.

Name	Type	Required	Default	Definition
match	loose, strict	#DEFAULT	loose	Controls whether the schema uses loose or strict matching.

2.2. The ruleset element

Summary:	Contains a set of reusable named constraints.
Content model:	(ruleref* , (baseName | occurrence | playing)* )
Parents:	tm-schema

The ruleset element is used to specify a set of constraints which can be reused by any topic class definition. Its intent is to allow the specification of reused rules in a single place, improving maintainability and readability of topic map schemas.

The ruleset element must have one, and only one attribute, named id, and containing a unique identifier for the rule set, allowing it to be referenced from topic classes.

The ruleset element may contain any number of ruleref elements, followed by any number of baseName, occurrence, and playing elements.

Name	Type	Required	Default	Definition
id	ID	#REQUIRED		The ID of the ruleset, by which it may be referred to.

2.3. The topic element

Summary:	Defines a topic class.
Content model:	(instanceOf , otherClass* , (ruleref | superclass)* , (baseName | occurrence | playing)* )
Parents:	tm-schema

The topic element defines the constraints applying to topics belonging to a particular class. The element may contain constraints on the characteristics on instances of the class, as well as other kinds of restrictions.

Each topic element must have an instanceOf child, specifying what class of topic is constrained by this rule. Only topics which match this instanceOf element must follow the constraints specified by this topic element.

topic elements may have a match attribute, which may be set to loose or strict. If set to loose base names, occurrences, and roles not matching any of the constraints in the topic class definition are accepted. If set to strict such characteristics will be rejected. The default is strict matching.

The otherClasses element can be used to specify what other classes instances of this topic class may be instances of. If the element is not present instances of this class may not be instances of any other class. If present, instances of this class may only be instances of the classes specified by the element, and not any other classes. [Remark: Need a separate section, I guess.]

The baseName, occurrence, and playing elements specify what characteristics instances of this class may have.

Name	Type	Required	Default	Definition
id	ID	#IMPLIED		The ID of the class by which it may be referred to.
match	loose, strict	#DEFAULT	strict	Whether matching of constraints is loose or strict.

2.4. The baseName element

Summary:	A constraint on the base names of a topic.
Content model:	(scope , variant*)
Parents:	ruleset, topic

baseName elements constrain the base names a topic may have. If a base name has scope matching that specified in the scope child element of a baseName element it is validated against that baseName rule.

baseName elements may have min and max attributes specifying the minimum and maximum number of matching base names each topic may have. The default is that any number of matches is allowed.

The variant child element is used to specify what variants a base name may have. Matching of variant names is always strict, so variant names that do not match a rule in the schema are rejected.

Name	Type	Required	Default	Definition
max	CDATA	#DEFAULT	Inf	The maximum number of matches allowed on this constraint for each instance of the class.
min	CDATA	#DEFAULT	0	The minimum number of matches allowed on this constraint for each instance of the class.

2.5. The variant element

Summary:	A constraint on the variant names of a base name.
Content model:	(scope)
Parents:	baseName

The variant element is used to constrain the allowed variants of base names. If a variant name has scope matching that specified in the scope child element of a variant element it is validated against that variant rule.

variant elements may have min and max attributes specifying the minimum and maximum number of matching variant names each base name may have. The default is that any number of matches is allowed.

Name	Type	Required	Default	Definition
max	CDATA	#DEFAULT	Inf	The maximum number of matches allowed on this constraint for each instance of the class.
min	CDATA	#DEFAULT	0	The minimum number of matches allowed on this constraint for each instance of the class.

2.6. The occurrence element

Summary:	A constraint on the occurrences of a topic.
Content model:	(instanceOf , scope?)
Parents:	ruleset, topic

The occurrence element is used to constrain the possible occurrences a topic may have. If an occurrence has a type matching that specified by the instanceOf child it is validated against that occurrence rule.

The occurrence element may also have min and max attributes. The internal attribute may be set to yes (meaning that matching occurrences must be internal), no (meaning that matching occurrences must be external), or either, meaning that they may be either.

If the occurrence element has a scope child element the scope of each occurrence is matched against that specified in the scope child.

Name	Type	Required	Default	Definition
internal	yes, no, either	#DEFAULT	either	Whether the occurrence must be internal or external, or whether it can be both.
max	CDATA	#DEFAULT	Inf	The maximum number of matches allowed on this constraint for each instance of the class.
min	CDATA	#DEFAULT	0	The minimum number of matches allowed on this constraint for each instance of the class.

2.7. The playing element

Summary:	Defins a constraint on the roles a topic may play in associations.
Content model:	(instanceOf , in?)
Parents:	ruleset, topic

The playing element is used to constrain what roles a topic may play in associations. If an association role has a type matching that specified in the instanceOf child it is validated against the playing rule.

The playing element may have min and max attributes constraining the cardinality of matching association roles.

The playing element may have an in child, which specifies what types of associations the association roles may be part of. The in element may contain any number of instanceOf elements. As long as the containing association matches one of those elements the association role is accepted.

Name	Type	Required	Default	Definition
max	CDATA	#DEFAULT	Inf	The maximum number of matches allowed on this constraint for each instance of the class.
min	CDATA	#DEFAULT	0	The minimum number of matches allowed on this constraint for each instance of the class.

2.8. The association element

Summary:	Defines an association class.
Content model:	(instanceOf , scope? , role+)
Parents:	tm-schema

The association element is used to define the structure of a class of associations. Any association which matches the instanceOf child of an association element is validated against the rule specified by that element.

The scope element can be used to define what scopes are allowed for associations of this class. If the element is not present, all scopes are allowed. If present, the scope of all associations must match that specified.

The association roles in the association will be validated against the role children of the association element. Roles that do not match any role element will be rejected.

The association element has no attributes.

2.9. The role element

Summary:	Contains a constraint on the association roles in an association.
Content model:	(instanceOf , player*)
Parents:	association

The role element is used to constrain the association roles that may appear inside an association. The instanceOf child is used to specify the type matched by the role element, and all association roles that match the instanceOf element will be matched against the role element.

The role element may have min and max attributes, which specify the number of matches that are allowed within each association.

The player child element is used to specify what classes of topics are allowed to play this role in an association. Topics are matched against the player element in the same way they are matched against the instanceOf element. (See the section on that element for details.)

Name	Type	Required	Default	Definition
max	CDATA	#DEFAULT	Inf	The maximum number of matches allowed on this constraint for each instance of the class.
min	CDATA	#DEFAULT	0	The minimum number of matches allowed on this constraint for each instance of the class.

3. Low-level elements

This section defines the semantics of the low-level elements of the Ontopia Schema Language.

3.1. The ruleref element

Summary:	A reference to a ruleset that is to be included in a topic class.
Content model:	EMPTY
Parents:	ruleset, topic

The ruleref element is used to refer to ruleset elements in the same schema. The element has a rule attribute, which contains the ID of the ruleset element being referenced. All the constraints contained in that ruleset element are then included in the set of constraints defined by the ruleset or topic element that contains the ruleref element.

It is an error for the ruleref element to refer to an ID that does not exist in the schema.

Name	Type	Required	Default	Definition
rule	IDREF	#REQUIRED		The ID of the ruleset being referred to.

3.2. The instanceOf element

Summary:	Specifies the type of a topic map object.
Content model:	(topicRef | subjectIndicatorRef | internalTopicRef | any)?
Parents:	association, playing, scope, in, occurrence, role, topic

The instanceOf element is used to match topic map objects by their type. The contents of the element specify what topic is allowed as the type of the objects to be matched. If the element is empty it means that the object must have no type in order to match.

The subclasses attribute is used to control whether instances of subclasses should match or not. If it is set to yes they do match, and if set to no they do not. The default is that subclasses do match.

The topicRef, subjectIndicatorRef, internalTopicRef, and any child elements are used to specify the type of the topic map object.

Name	Type	Required	Default	Definition
subclasses	yes, no	#DEFAULT	yes	Whether instances of subclasses are accepted.

3.3. The scope element

Summary:	Controls the allowed scope of a topic characteristic.
Content model:	(topicRef | subjectIndicatorRef | internalTopicRef | any | instanceOf)*
Parents:	association, occurrence, variant, baseName

The scope element is used to constrain (or match) the scopes of topic map objects. The contents of the element specify what topics are allowed as themes in the scope. If the element is empty it means that the object must be in the unconstrained scope to match.

The match attribute is used to specify how the scope of the topic map object is matched against that specified in this element. If set to exact every topic mentioned in the scope element must be present, and no topics that are not specified are allowed. If set to superset some or all of the topics specified may be left out, but no extra topics are allowed. If set to subset no topics may be left out, but the ones specified must at least be present.

The topicRef, subjectIndicatorRef, internalTopicRef, and any child elements are used to specify the themes that may appear in the scope. The instanceOf element is also allowed as a child of the scope element, and if present it states that all topics that match the instanceOf element are allowed as themes in the scope.

Name	Type	Required	Default	Definition
match	subset, superset, exact	#DEFAULT	exact	Controls how the specified scope is matched against the scope of actual topic characteristics.

3.4. The topicRef element

Summary:	Identifies a topic by its source locator.
Content model:	EMPTY
Parents:	otherClass, scope, player, instanceOf

The topicRef element is used to match topics by their source locators. The href attribute contains a URI (resolved relative to that of the current entity), and topics which have that URI as their source locator will match, while no other topics will.

Name	Type	Required	Default	Definition
href	CDATA	#REQUIRED		The URI that is the source locator of the topic.

3.5. The subjectIndicatorRef element

Summary:	Identifies a topic by its subject indicator.
Content model:	EMPTY
Parents:	otherClass, scope, player, instanceOf

The subjectIndicatorRef element is used to match topics by their subject indicators. The href attribute contains a URI (resolved relative to that of the current entity), and topics which have that URI as their subject indicator will match, while no other topics will.

Name	Type	Required	Default	Definition
href	CDATA	#REQUIRED		The URI that is the subject indicator.

3.6. The internalTopicRef element

Summary:	Identifies a topic by its source locator.
Content model:	EMPTY
Parents:	otherClass, scope, player, instanceOf

The internalTopicRef element is used to match topics by their source locators. The href attribute contains a URI (resolved relative to the base address of the topic map), and topics which have that URI as their source locator will match, while no other topics will.

Name	Type	Required	Default	Definition
href	CDATA	#REQUIRED		The URI that is the source locator of the topic.

3.7. The any element

Summary:	Any topic.
Content model:	EMPTY
Parents:	otherClass, scope, player, instanceOf

The any element matches any topic and is used as a wildcard in matching.

The any element has no attributes.

4. The DTD

<!-- ...................................................................... -->
<!-- Ontopia Schema Language DTD  ......................................... -->
<!-- ...................................................................... -->
<!-- 

    This is OSL, an XML interchange syntax for the Ontopia Schema
    Language. The copyright in this DTD is held by Ontopia. The DTD
    may be used and distributed freely, as long as it is not changed.

    Version:  1.0.
    Revision: $Revision: 1.3 $
    Date:     $Date: 2001/09/21 12:09:34 $

    Recommended public identifier:
      '+//IDN ontopia.net//DTD Ontopia Schema Language (1.0)//EN'
    More information at:
      <URL: http://www.ontopia.net/ >

-->


<!-- %ref: Topic identification elements ........................ -->

<!ENTITY % ref "topicRef | subjectIndicatorRef | internalTopicRef | any">


<!-- tm-schema: Topic map schema element ........................ -->

<!ELEMENT tm-schema (ruleset | topic | association)*>
<!ATTLIST tm-schema match (loose | strict) "loose">


<!-- ruleset: Named rule set element ............................ -->

<!ELEMENT ruleset (ruleref*, (baseName | occurrence | playing)*)>
<!ATTLIST ruleset id ID #REQUIRED>


<!-- topic: Topic class definition element ...................... -->

<!ELEMENT topic (instanceOf, otherClass*, (ruleref | superclass)*, 
                 (baseName | occurrence | playing)*)>
<!ATTLIST topic
  match (loose | strict) "strict"
  id    ID               #IMPLIED>


<!-- otherClass: Definition of allowed classes of topics ........ -->

<!ELEMENT otherClass (%ref;)>
<!-- other classes of which topics of this class are allowed to be instances.
     if this element is empty or not given topics of this class can be
     instances of no other class. -->


<!-- ruleref: Reference to named rule set ....................... -->

<!ELEMENT ruleref EMPTY>
<!ATTLIST ruleref rule IDREF #REQUIRED>


<!-- superclass: Reference to superclass of class ............... -->

<!ELEMENT superclass EMPTY>
<!ATTLIST superclass ref IDREF #REQUIRED>


<!-- instanceOf: Reference to class from instance ............... -->

<!ELEMENT instanceOf (%ref;)?>
<!ATTLIST instanceOf
  subclasses (yes | no) "yes">


<!-- topicRef: Direct reference to <topic> element .............. -->

<!ELEMENT topicRef EMPTY>
<!-- refers to an external topic which defines the class which this
     element constrains -->

<!ATTLIST topicRef
  href CDATA #REQUIRED>


<!-- subjectIndicatorRef: Refers to topic by subject indicator .. -->

<!ELEMENT subjectIndicatorRef EMPTY>
<!ATTLIST subjectIndicatorRef
  href CDATA #REQUIRED>


<!-- internalTopicRef: Refers to <topic> relative to TM doc ..... -->

<!ELEMENT internalTopicRef EMPTY>
<!ATTLIST internalTopicRef
  href CDATA #REQUIRED>


<!-- any: Refers to any topic ................................... -->

<!ELEMENT any EMPTY>


<!-- baseName: Constraint for base names ........................ -->

<!ELEMENT baseName (scope, variant*)>
<!ATTLIST baseName 
  min CDATA "0"
  max CDATA "Inf">
  
<!-- min: minimum number of base names according to this spec (0 - )
     max: maximum number of base names according to this spec (0 - , Inf) -->


<!-- scope: Defines allowed scope ............................... -->

<!ELEMENT scope (%ref; | instanceOf)*>
<!ATTLIST scope
  match (subset | superset | exact) "exact">


<!-- variant: Variant name constraint ........................... -->

<!ELEMENT variant (scope)>
<!ATTLIST variant
  min CDATA "0"
  max CDATA "Inf">


<!-- occurrence: Occurrence constraint .......................... -->

<!ELEMENT occurrence (instanceOf, scope?)>
<!ATTLIST occurrence
  min      CDATA       	       "0"
  max      CDATA       	       "Inf"
  internal (yes | no | either) "either">


<!-- playing: Association role of topic constraint .............. -->

<!ELEMENT playing (instanceOf, in?)>
<!ATTLIST playing
  min      CDATA       	       "0"
  max      CDATA       	       "Inf">


<!-- in: Specifies association type of role ..................... -->

<!ELEMENT in (instanceOf*)>


<!-- association: Association class definition .................. -->

<!ELEMENT association (instanceOf, scope?, role+)>


<!-- role: Association role constraint .......................... -->

<!ELEMENT role (instanceOf, player*)>
<!ATTLIST role
  min CDATA "0"
  max CDATA "Inf">


<!-- player: Definition of role player type ..................... -->

<!ELEMENT player (%ref;)?>
<!ATTLIST player
  subclasses (yes | no) "yes">