Copyright © 2008 W3C® (MIT, ERCIM, Keio), All Rights Reserved. W3C liability, trademark and document use rules apply.
This document, developed by the Rule Interchange Format (RIF) Working Group, specifies RIF-Core, a common subset of RIF-BLD and RIF-PRD based on RIF-DTB 1.0. The RIF-Core presentation syntax and semantics are specified as restrictions on RIF-BLD. The XML serialization syntax of RIF-Core is specified via a mapping from the presentation syntax. A normative XML schema is also provided.
This section describes the status of this document at the time of its publication. Other documents may supersede this document. A list of current W3C ®publications and the latest revision of this technical report can be found in the W3C technical reports index at http://www.w3.org/TR/.
This document is being published as one of a set of 5 documents:
The Rule Interchange Format (RIF) Working Group seeks
public feedback on these Working Drafts. Please send your
comments to public-rif-comments@w3.org ( MIT , ERCIM , Keio ), All Rights Reserved.public archive). If possible, please offer
specific changes to the text that would address your
concern. You may also wish to check the Wiki
Version of this document for internal-review comments and changes being
drafted which may address your concerns.
Publication as a Working Draft does not imply endorsement by the W3C liability , trademarkMembership. This is a draft document and may be updated, replaced or obsoleted by other documents at any time. It is inappropriate to cite this document use rules apply.as other than work in progress.
This document was produced by a group operating under the 5 February 2004 W3C Patent Policy. W3C maintains a public list of any patent disclosures made in connection with the deliverables of the group; that page also includes instructions for disclosing a patent. An individual who has actual knowledge of a patent which the individual believes contains Essential Claim(s) must disclose the information in accordance with section 6 of the W3C Patent Policy.
Contents |
This specification develops RIF-Core (the Core of the Rule Interchange Format). From a theoretical perspective, RIF-Core corresponds to the language of definite Horn rules without function symbols ('Datalog') and a standard first-order semantics. RIF-Core thus is a subset of RIF-BLD [RIF-BLD]. At the same time, RIF-Core is a language of production rules permitting only assert actions. RIF-Core thus also is a subset of RIF-PRD [RIF-PRD]. Moreover, RIF-Core is based on the built-ins of RIF-DTB 1.0 [RIF-DTB]. The common subset of RIF-BLD and RIF-PRD is specified based on RIF-DTB 1.0.
Syntactically, RIF-Core has a number of Datalog extensions to support features such as objects and frames as in F-logic [KLW95], internationalized resource identifiers (or IRIs, defined by [RFC-3987]) as identifiers for concepts, and XML Schema datatypes [XML-SCHEMA2]. In addition, RIF RDF and OWL Compatibility [RIF-RDF+OWL] defines the syntax and semantics of integrated RIF-Core/RDF and RIF-Core/OWL languages. These features make RIF-Core a Web-aware language. However, it should be kept in mind that RIF is designed to enable interoperability among rule languages in general, and its uses are not limited to the Web.
RIF-Core is defined as a specialization of RIF-BLD (hence of [RIF-FLD] which is part of the RIF extensibility framework). It is a syntactic subset of RIF-BLD, so that a well-formed RIF-Core formula (including documents and condition formula) is also a well-formed RIF-BLD formula. The semantics of RIF-Core is defined to be identical to that of RIF-BLD.
RIF-Core is also a syntactic subset of RIF-PRD. It is intended that a RIF-PRD consumer can treat a RIF-Core ruleset as if it were a RIF-PRD rule set and it would conform to the normative RIF-Core first order semantics. However, due to the presence of builtin functions and predicates there are RIF-Core rulesets which are unsafe and do not reach a stable fixed-point under RIF-PRD semantics. We define the conformance to RIF-Core so as to only require conformance over a safe subset of rules, in this way we permit RIF-PRD processors to be safely conformant while allowing RIF-Core documents to contain unsafe rules. Producers of RIF-Core who require maximum interchange are advised to restrict themselves to safe rules. These notions of safeness and safe conformance are defined formally in section 5 Conformance and Safety.
To give a preview, here is a simple complete RIF-Core example deriving a ternary relation from its inverse.
Example 1 (An introductory RIF-Core example).
A rule can be written in English to derive the buy relationships (rather than store them) from the sell relationships that are stored as facts (e.g., as exemplified by the English statement below):
The fact Mary buys LeRif from John can be logically derived by a modus ponens argument. Assuming Web IRIs for the predicates buy and sell, as well as for the individuals John, Mary, and LeRif, the above English text can be represented in RIF-Core Presentation Syntax as follows.
Document( Prefix(cpt http://example.com/concepts#) Prefix(ppl http://example.com/people#) Prefix(bks http://example.com/books#) Group ( Forall ?Buyer ?Item ?Seller ( cpt:buy(?Buyer ?Item ?Seller) :- cpt:sell(?Seller ?Item ?Buyer) ) cpt:sell(ppl:John bks:LeRif ppl:Mary) ) )
For the interchange of such rule (and fact) documents, an equivalent RIF-Core XML Syntax is given in this specification. To formalize their meaning, a RIF-Core Semantics is specified.
RIF-BLD [RIF-BLD] specifies a presentation syntax and an XML syntax. The presentation syntax is normative, but is not intended to be a concrete syntax. Since RIF is an interchange format, it uses XML as its concrete syntax.
RIF-Core is a syntactic subset of RIF-BLD and in this section we define it as a restriction on the RIF-BLD syntax.
The alphabet of the presentation language of RIF-Core is the alphabet of the RIF-BLD presentation language with the exclusion of the symbol ## and the set of symbols ArgNames.
Editor's Note: The status of membership (#) and subclass (##) formulas within Core is under debate in the working group. While there is a notion of membership within PRD it is restricted compared to that in BLD. This current draft for Core includes membership (#) but restricts its use to solely within the RIF Core Condition Formulas. Future drafts may extend its use, also include ## or may omit both entirely.
The Terms of RIF-Core are the terms of RIF-BLD with the exclusion of terms with named arguments, subclass terms, and external terms of the form External(fr), where fr is a frame term.
Editor's Note: This presumes that the At Risk #1 feature External frames is retained in RIF-BLD. If that is not the case the above will need updating.
The Formulas of the RIF-Core are the formulas of RIF-BLD with the following modifications.
Editor's Note: Built-in functions and predicates present problems for many bottom-up and top-down reasoners, which lack constraint-solving capabilities. This is because these reasoners can typically make use of a built-in only if certain arguments are bound. For instance, a bottom-up reasoner might have trouble with evaluating a condition formula And(p(?X) External(foo(?Y,?X))), where p is a database predicate and foo a builtin predicate, if the reasoner can invoke foo only when the first argument ?Y is bound. The working group is currently discussing the implications of all this for RIF-Core compliance.
RIF-Core allows every term and formula to be optionally annotated in the same way as RIF-BLD. The frame formulas that are allowed as part of an annotation must be syntactically correct for RIF-Core. In particular, no function symbols are allowed in such a formula.
A syntactically correct RIF-Core formula that passes the well-formedness test for RIF-BLD is also a well-formed RIF-Core formula. In this case, it simply means that
For rif:local symbols, the first three of the above clauses apply only within the scope of a single document. That is, two different occurrences of the same rif:local symbol in different documents may have different arities, may occur both as predicates and individuals, etc. For other symbols, the above restrictions apply not only to the document that contains the particular occurrence of the symbol, but also to all documents that are imported by that document.
Until now, we have used mathematical English to specify the syntax of RIF-Core as a restriction on RIF-BLD. Tool developers, however, may prefer EBNF notation, which provides a more succinct overview of the syntax. However, note that EBNF is unable to express all of the above well-formedness conditions, in particular the requirement that each symbol appear in only one context. For this reason this section is not Normative.
The EBNF for the RIF-Core presentation syntax is given as follows, showing the entire (top-down) context of its three parts for rules, conditions, and annotations.
Rule Language:
Document ::= IRIMETA? 'Document' '(' Base? Prefix* Import* Group? ')' Base ::= 'Base' '(' IRI ')' Prefix ::= 'Prefix' '(' Name IRI ')' Import ::= IRIMETA? 'Import' '(' IRICONST PROFILE? ')' Group ::= IRIMETA? 'Group' '(' (RULE | Group)* ')' RULE ::= (IRIMETA? 'Forall' Var+ '(' CLAUSE ')') | CLAUSE CLAUSE ::= Implies | ATOMIC Implies ::= IRIMETA? ATOMIC ':-' FORMULA PROFILE ::= TERM
Condition Language:
FORMULA ::= IRIMETA? 'And' '(' FORMULA* ')' | IRIMETA? 'Or' '(' FORMULA* ')' | IRIMETA? 'Exists' Var+ '(' FORMULA ')' | ATOMIC | IRIMETA? Equal | IRIMETA? Member | IRIMETA? 'External' '(' Atom ')' ATOMIC ::= IRIMETA? (Atom | Frame ) Atom ::= UNITERM UNITERM ::= Const '(' (TERM* ')' Equal ::= TERM '=' ( TERM | 'External' '(' FUNC ')' ) FUNC ::= Const '(' (GENERAL_TERM* ')' GENERAL_TERM ::= IRIMETA? (Const | Var | FUNC) Member ::= TERM '#' TERM Frame ::= TERM '[' (TERM '->' TERM)* ']' TERM ::= IRIMETA? (Const | Var) Const ::= '"' UNICODESTRING '"^^' SYMSPACE | CONSTSHORT Name ::= UNICODESTRING Var ::= '?' UNICODESTRING SYMSPACE ::= ANGLEBRACKIRI | CURIE
Annotations:
IRIMETA ::= '(*' IRICONST? (Frame | 'And' '(' Frame* ')')? '*)'
The following subsections explain and exemplify the Condition Language and Rule Language parts; the Annotations part is unchanged from RIF-BLD.
The RIF-Core Condition Language represents formulas that can be used in the premises of RIF-Core rules (also called rule bodies). The EBNF grammar for a superset of the RIF-Core condition language is shown in the above conditions part.
This is a subset of the EBNF for the RIF-BLD condition language specified in RIF-BLD conditions part reflecting the syntax restrictions on RIF-Core described normatively in sections 2.1 through 2.5 above.
The Example 2 from the RIF-BLD document illustrates a RIF-Core condition.
The presentation syntax for RIF-Core rules is based on the syntax in Section EBNF for RIF-Core Condition Language with the productions shown in the above rules part.
Again, this is a subset of the EBNF for the RIF-BLD rule language specified in RIF-BLD rules part reflecting the syntax restrictions on RIF-Core described normatively in sections 2.1 through 2.5 above.
Example 3 from the RIF-BLD document illustrates a set of RIF-Core rules and Example 4 a RIF-Core document that contains an annotated group formula. In contrast, the Example 6 from the RIF-BLD document shows a formula that is not in RIF-Core because it includes terms with named arguments, which are not allowed in this dialect.
RIF-Core is a syntactic subset of RIF-BLD and the semantics of RIF-Core is identical that of RIF-BLD.
The XML syntax of RIF-Core is a subset of the XML syntax of RIF-BLD. All XML tags of RIF-BLD (except Subclass, sub and super) are supported, but the XML schema of RIF-Core restricts their context with respect to what is allowed by the XML Schema of RIF-BLD. The semantics of the XML syntax for RIF-Core is defined through the same RIF-BLD XML-to-presentation syntax mapping.
Editor's Note: As noted earlier the status of membership (#) and subclass (##) formulas are under discussion in the working group. Depending on the resolution of this issue the tags relating to subclass may be reinstated or those relating to membership may be removed.
Example 5, "A RIF condition and its XML serialization," in the RIF-BLD document illustrates XML serialization for a RIF-Core condition formula. XML serialization of a complete RIF-Core document appears in the RIF-BLD document as Example 7.
RIF-Core is a syntactic subset of both RIF-BLD and RIF-PRD. The semantics of a RIF-Core formula is the same as in RIF-BLD.
All RIF-Core documents are syntactically also RIF-PRD documents. However, some formulas may be unsafe and cannot be executed under the RIF-PRD operational semantics. Thus, in order to allow production rule systems and logic programming systems to interchange rulesets via RIF-Core, we define a safe subset of RIF-Core and the notion of safely conformant RIF-Core consumers or producers. A safely conformant RIF-Core consumer can reject non-safe RIF-Core documents even if they are otherwise well-formed. The notion of (general) conformance allows RIF-Core producers and consumers to produce and consume unsafe rulesets. For safe Core documents, the logical semantics of RIF-BLD and operational fixed-point semantics of RIF-PRD coincide. In the absence of the data types and built-ins, this follows from [Vianu97].
Definition (Safeness).
RIF-Core does not require or expect conformant systems to implement the RIF-Core presentation syntax. Instead, conformance is described in terms of semantics-preserving transformations.
Let Τ be a set of datatypes that includes the datatypes specified in [RIF-DTB], and suppose Ε is a set of external predicates and functions that includes the built-ins listed in [RIF-DTB]. We say that a formula φ is a CoreΤ,Ε formula iff
A RIF processor is a conformant CoreΤ,Ε consumer iff it implements a semantics-preserving mapping from the set of all CoreΤ,Ε formulas to the language L of the processor.
A RIF processor is a conformant CoreΤ,Ε producer iff it implements a semantics-preserving mapping from a subset of the language L of the processor to a set of CoreΤ,Ε formulas.
A conformant document is one that conforms to all the syntactic constraints of RIF-Core, including ones that cannot be checked by an XML Schema validator.
In addition:
Feature At Risk #3: Strictness Requirement
Note: This feature is "at risk" and may be removed from this specification based on feedback. Please send feedback to public-rif-comments@w3.org.
The two preceding clauses are features AT RISK. In particular, the "strictness" requirement is under discussion.
This normative section describes RIF-Core by specializing RIF-PRD . The reader is assumed to be familiar with RIF-PRD as described in [RIF-PRD]. The reader who is not interested in how RIF-Core is derived from PRD can skip this section.
RIF-PRD and RIF-BLD share essentially the same presentation syntax and XML syntax.
The differences between the two dialects are summarized below:
For context, the table below lists the constructs which are in the maximal common subset (intersection) of BLD and PRD, marks certain rows as not being in Core, or annotates them w.r.t. restrictions making them part of Core.
Presentation syntax | XML syntax | |
---|---|---|
TERM ::= | ||
Core |
[Const | Var | External] | |
Const ::= | ||
Core |
'"' UNICODESTRING '"^^' SYMSPACE |
<Const type=xsd:anyURI [xml:lang=xsd:language]? > Any Unicode string </Const> |
Var ::= | ||
Core |
'?' Any Unicode string |
<Var> Any Unicode string </Var> |
External (as TERM) ::= | Remark: only used for external built-in calls; no logical functions | |
Core |
'External' '(' Expr ')' |
<External> <content> Expr </content> </External> |
ATOMIC ::= | ||
Core |
[Atom | Frame] | |
Atom ::= | ||
Core |
Const '(' (TERM* ')' |
<Atom> <op> Const </op> <args rif:ordered="yes"> TERM* </args>? </Atom> |
Equal ::= | ||
Core |
TERM = TERM |
<Equal> <left> TERM </left> <right> TERM </right> </Equal> |
Member ::= | Remark: under discussion; might be not part of Core | |
Core |
TERM # TERM |
<Member> <instance> TERM </instance> <class> TERM </class> </Member> |
Frame ::= | ||
Core |
TERM ' [ ' (TERM ' -> ' TERM)* ' ] ' |
<Frame> <object> TERM </object> <slot rif:ordered="yes"> TERM TERM </slot>* </Frame> |
FORMULA ::= | ||
Core |
[ATOMIC | External | And | Or | Exists] | |
External (as FORMULA) ::= | ||
Core |
'External '(' [Atom | Frame] ')' |
<External> <content> [ Atom | Frame ] </content> </External> |
And ::= | ||
Core |
'And' '(' FORMULA* ')' |
<And> <formula> FORMULA </formula>* </And> |
Or ::= | ||
Core |
'Or' '(' FORMULA* ')' |
<Or> <formula> FORMULA </formula>* </Or> |
Exists ::= | ||
Core |
'Exists' Var+ '(' FORMULA ')' |
<Exists> <declare> Var </declare>+ <formula> FORMULA </formula> </Exists> |
RULE ::= | Remark: There is an implicit assert assumption for production rules | |
Core |
[ Forall | Implies | ATOMIC ] | |
Forall ::= | Remark: There is an implicit assert assumption for production rules | |
Core |
' Forall ' Var+ ' ( ' [Implies | ATOMIC] ' ) ' |
<Forall> <declare> Var </declare>+ <formula> [Implies | ATOMIC ] </formula> </Forall> |
Implies ::= | Remark: There is an implicit assert assumption for production rules; | |
Core |
ATOMIC '=>' FORMULA |
<Implies> <if> FORMULA </if> <then> ATOMIC </then> </Implies> |
Group ::= | ||
Core |
METADATA? 'Group' '(' ([RULE | Group])* ')' |
<Group> <sentence> [ RULE | Group ] </sentence>* </Group> |
Core |
METADATA? 'Document' '(' Group? ')' |
<Document> <payload> Group </payload>? </Document> |
METADATA ::= | ||
Core |
'(*' Const? (Frame | 'And' '(' Frame* ')')? '*)' |
<AnyClassTag> <id> Const </declare>? <meta> [ Frame | <And> <formula> Frame </formula> </And> ] </meta>? other content of AnyClassTag </AnyClassTag> |
This document is the product of the Rules Interchange Format (RIF) Working Group (see below) whose members deserve recognition for their time and commitment. The editors extend special thanks to: ***, for their thorough reviews and insightful discussions; the working group chairs, Chris Welty and Christian de Sainte-Marie, for their invaluable technical help and inspirational leadership; and W3C staff contact Sandro Hawke, a constant source of ideas, help, and feedback.
The regular attendees at meetings of the Rule Interchange Format (RIF) Working Group at the time of the publication were:
Adrian Paschke (REWERSE),
Axel Polleres (DERI),
Chris Welty (IBM),
Christian de Sainte Marie (ILOG),
Dave Reynolds (HP),
Gary Hallmark (ORACLE),
Harold Boley (NRC),
Hassan Aït-KaciAït-Kaci (ILOG),
Igor Mozetic (JSI),
John Hall (OMG),
Jos de Bruijn (FUB),
Leora Morgenstern (IBM),
Michael Kifer (Stony Brook),
Mike Dean (BBN),
Sandro Hawke (W3C/MIT), and
Stella Mitchell (IBM).
We would also like to thank *** past members of the working group, ***.
The namespace of RIF is http://www.w3.org/2007/rif#.
XML schemas for the RIF-Core sublanguages are defined below and are also available here with additional examples.
<?xmlversion="1.0"encoding="UTF-8"?>version="1.0" encoding="UTF-8"?>
<?xmlversion="1.0"encoding="UTF-8"?>Retrievedfrom"http://www.w3.org/2005/rules/wiki/Core"ViewsPageDiscussionViewsourceHistoryPersonaltoolsLogin/createaccountLoginwithOpenIDif(window.isMSIE55)fixalpha();RIFWorkingGroupHomeEventsMeetingsArchivesBlogTrackerResolutionsCharterOldWikiTeleconRegretsHelpRecentchangesDraftsBLDFLDRDF+OWLDTBUCRPRDSearchToolboxWhatlinkshereRelatedchangesUploadfileSpecialpagesPrintableversionPermanentlinkThispagewaslastmodified14:19,25November2008.Thispagehasbeenaccessed493times.PrivacypolicyAboutRIFDisclaimersif(window.runOnloadHook)runOnloadHook();version="1.0" encoding="UTF-8"?>