Version: 0.5 2017.09.15 AET
"EEX" is acronym for EDM EXPRESS Extension Schema, and is also the file extension used for such files.
EXPRESS_EXTENSION_SCHEMA Declaration
EXPRESS_EXTENSION_SCHEMA extensionSchemaName FOR expressSchemaName; <body> END_EXPRESS_EXTENSION_SCHEMA;
EXPRESS_EXTENSION_SCHEMA extensionSchemaName {‘GUID’} FOR expressSchemaName;<body> END_EXPRESS_EXTENSION_SCHEMA;
...
Example: exporting a population including “extension schema data” on a step file package. When importing such step file package, the GUID is used to ensure that the extension schemata in the importing DB is exactly the same as in the exporting DB. There will be an option in write/read step file package to include exporting/defining of the extension schemata.
ENTITY Declaration
ENTITY declarations in an extension schema have the same syntax as in an Express schema except the following:
...
Code Block | ||||
---|---|---|---|---|
| ||||
-- extend the IFC schema with a window suitable for cars ENTITY xifcWindShield SUBTYPE of IfcWindow; fitsMakes : LIST OF xifcVehicle; -- list of car makes that this window will fit in END_ENTITY; ENTITY xifcVehicle; brand : STRING; -- i.e. 'Ford' model : STRING; -- i-e 'T' year : STRING; -- i.e '1921-1924' END_ENTITY; |
ENTITY_EXTENSION Declaration
The ENTITY_EXTENSION – END_ENTITY_EXTENSION as earlier agreed, plus the same extensions for attribute declarations as for ENTITY.
...
We can discuss how many of the ”extension schema rules checking” we implement in the compiler and how many we implement in the edmCloseExtensionSchemaCompilerModel() (similar as edmCloseCompilerModel()).
Supported EXPRESS Declarations
The following Express declarations are supported:
- TYPE - without the WHERE
- ENTITY – with the above listed restrictions and extensions
- Subtype_constraint – to be decided???
- CONSTANT
- FUNCTION
- PROCEDURE
Not supported EXPRESS Declarations
The following Express declarations are not supported:
- SCHEMA
- RULE (Rules can be written in separate RULE_SCHEMA)
New tokens (reserved words):
ALL_CROSS_MODEL_REFS_TO_ME
ALL_LOCAL_REFS_TO_ME
END_ENTITY_EXTENSION
ENTITY_EXTENSION
END_EXPRESS_EXTENSION_SCHEMA
EXPRESS_EXTENSION_SCHEMA SUBTYPES
Compiler checking
In addition to the “standard compilation time checking” during defining (compilation) of an extension schema, the following will be checked:
...
We can discuss how many of the ”extension schema rules checking” we implement in the compiler and how many we implement in the edmCloseExtensionSchemaCompilerModel() (similar as edmCloseCompilerModel()).
Syntax Illustration 0.
...
5
Code Block |
---|
EXPRESS_EXTENSION_SCHEMA ExtensionSchemaName FOR IFC2X3; ENTITY_EXTENSION extToIfcxx1 FOR ifcxx1; (* Attribute name must be unique within the extended entity (ifcxx1), may be same as (** Check basic elements for EXTENSION_SCHEMA Should compile with no errors *) EXPRESS_EXTENSION_SCHEMA EEX_1011 FOR IFC4; GLOBAL gGeneric : GENERIC; gGenericEntity : GENERIC_ENTITY; geexEntity : eexEntity; END_GLOBAL; CONSTANT cInteger : INTEGER := 1; END_CONSTANT; ENTITY eexEntity; explicitAttr1 attribute: namesINTEGER; in super type Entity. When so, this extended attributeexplicitAttr2 must be : GENERIC_ENTITY; DERIVE derivedAttr1 : STRING := foo1(SELF); derivedAttr2 : accessedSTRING by name using "full qualified attribute name". The attribute domains can be any domain in the actual schema or domain (instance type) defined in the current extension schema *):= foo2(SELF); INVERSE cross_references : SET OF GENERIC_ENTITY FOR ALL_CROSS_MODEL_REFERENCES_TO_ME; local_refs: SET OF GENERIC_ENTITY FOR ALL_LOCAL_REFS_TO_ME; END_ENTITY; ENTITY eexTask SUBTYPE OF(ifcTask); explicitAttr1 : INTEGER; END_ENTITY; ENTITY_EXTENSION extToIfcSpace extendedExplicitAttr1FOR : STRINGIfcSpace; (* Attribute name --must anybe simpleunique datawithin typethe canextended beentity used(ifcxx1), notmay definedbe typessame as extendedExplicitAttr2 : GENERIC_ENTITY; attribute --sdaiINSTANCEnames ofin anysuper type Entity. canWhen be used as link to instance in any other model.so, this extended attribute must be accessed by extendedExplicitAttr3name :using SET"full OF GENERIC_ENTITY; qualified attribute name". DERIVE extendedDeriveAttr : SET OF GENERIC_ENTITY := ExtFunc(SELF); INVERSE ALL_CROSS_MODEL_REFERENCES_TO_ME : SET OF GENERIC_ENTITY; -- reserved name, keeps all references from other models ALL_LOCAL_REFS_TO_ME : SET OF GENERIC_ENTITY; The attribute domains can be any domain in the actual schema or domain (instance type) defined in the current extension schema *) extendedExplicitAttr1 : STRING; -- reserved name, keeps all references from other models -- any simple data type can extendedInverseAttr : SET OF GENERIC_ENTITY FOR (entityx.explicitAttr1, entityx.explicitAttr2, ...entityz.explicitAttr5); END_ENTITY_EXTENSION; ENTITY_EXTENSION extToifcxx2 FOR SUBTYPES ifcxx2; (* same as ENTITY_EXTENSION except that all extensions will be valid for all subtypes of the extended entity, i.e., all subtypes of ifcxx2 *) extendedExplicitAttr1 : STRINGbe used, not defined types extendedExplicitAttr2 : GENERIC_ENTITY; --sdaiINSTANCE of any type. can be used as link to instance in any other model. extendedExplicitAttr3 : SET OF GENERIC_ENTITY; DERIVE extendedDeriveAttr : SET OF GENERIC_ENTITY := foo4(SELF); INVERSE cross_references : SET -- any simple data type can be used, not defined types extendedExplicitAttr2 : GENERIC_ENTITY; --sdaiINSTANCE of any type. can be used as link to instance in any other model. extendedExplicitAttr3 : SET OF GENERIC_ENTITY; INVERSE ALL_CROSS_MODEL_REFERENCES_TO_ME : SET OF GENERIC_ENTITY; -- reserved name, keeps all references from other models ALL_LOCAL_REFS_TO_ME : SET OF GENERIC_ENTITY; -- reserved name, keeps all references from other models extendedInverseAttr : SET OF GENERIC_ENTITY FOR (entityx.explicitAttr1, entityx.explicitAttr2, ...entityz.explicitAttr5); END_ENTITY_EXTENSION; ENTITY_EXTENSION name; -- no FOR clause means it is a "new" entity explicitAttr1 : INTEGER; explicitAttr2 : GENERIC_ENTITY; -- ........ INVERSE -- ....... ALL_CROSS_MODEL_REFERENCES_TO_ME : SET OF GENERIC_ENTITY; -- reserved name, keeps all references from other models END_ENTITY_EXTENSION; ENTITY_EXTENSION name SUBTYPE OF (FOR entity1, entity2, ... entityN); -- subtype of any entity in main schema (FOR xxx) or "extended entity" in schema explicitAttr1 : INTEGER; explicitAttr2 : GENERIC_ENTITY; -- ........ INVERSE -- ....... ALL_CROSS_MODEL_REFERENCES_TO_ME : SET OF GENERIC_ENTITY; -- reserved name, keeps all references from other models ALL_LOCAL_REFS_TO_ME : SET OF GENERIC_ENTITY; -- reserved name, keeps all references from other models END_ENTITY_EXTENSION; OF GENERIC_ENTITY FOR ALL_CROSS_MODEL_REFERENCES_TO_ME; local_refs : SET OF GENERIC_ENTITY FOR ALL_LOCAL_REFS_TO_ME; -- extendedInverseAttr : SET OF GENERIC_ENTITY FOR (entityx.explicitAttr1, entityx.explicitAttr2, ...entityz.explicitAttr5); END_ENTITY_EXTENSION; ENTITY_EXTENSION extToIfcElementSubtypes FOR SUBTYPE IfcElement; (* same as ENTITY_EXTENSION except that all extensions will be valid for all subtypes of the extended entity, i.e., all subtypes of IfcElement *) extendedExplicitAttr1 : STRING; -- any simple data type can be used, not defined types -- INVERSE -- extendedInverseAttr : SET OF GENERIC_ENTITY FOR (entityx.explicitAttr1, entityx.explicitAttr2, ...entityz.explicitAttr5); END_ENTITY_EXTENSION; FUNCTION foo1(arg : GENERIC) : STRING; RETURN('typename'); END_FUNCTION; FUNCTION foo2(arg : GENERIC_ENTITY) : STRING; RETURN('typename'); END_FUNCTION; FUNCTION foo4(arg : GENERIC_ENTITY) : SET OF GENERIC_ENTITY; LOCAL result : SET OF GENERIC_ENTITY := []; END_LOCAL; RETURN(?); END_FUNCTION; PROCEDURE poo1(arg1 : GENERIC;VAR arg2: GENERIC); arg1.explicitAttr1 := 2; arg2.explicitAttr1 := 2; END_PROCEDURE; PROCEDURE poo2(arg1 : GENERIC_ENTITY;VAR arg2: GENERIC_ENTITY); arg1.explicitAttr1 := 2; arg2.explicitAttr1 := 2; END_PROCEDURE; END_EXPRESS_EXTENSION_SCHEMA; |
Historical:
Expand | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||
Arnes adjusted proposal out of Hans Karsten initial spec:
I think this approach also allows for using only "ENTITY" in stead of "ENTITY_EXTENSION", but that is maybe going too far? Or maybe not, ref RULE_SCHEMA. |
...