DOCUMENT START
TOKENS
// ------------------------------------------
// Tokens

// Comments and whitespace
<DEFAULT> SKIP : {
" "
| "\t"
| "\n"
| "\r"
| "\f"
}

<DEFAULT> TOKEN : {
<#WS: " " | "\t" | "\n" | "\r" | "\f">
}

<DEFAULT> SPECIAL : {
<SINGLE_LINE_COMMENT: "#" (~["\n","\r"])* ("\n" | "\r" | "\r\n")?>
}

// Main tokens */
<DEFAULT> TOKEN : {
<IRIref: "<" (~[">","<","\"","{","}","^","\\","|","`","\u0000"-" "])* ">">
| <PNAME_NS: (<PN_PREFIX>)? ":">
| <PNAME_LN: <PNAME_NS> <PN_LOCAL>>
| <BLANK_NODE_LABEL: "_:" <PN_LOCAL>>
| <VAR1: "?" <VARNAME>>
| <VAR2: "$" <VARNAME>>
| <LANGTAG: <AT> (<A2Z>)+ ("-" (<A2ZN>)+)*>
| <#A2Z: ["a"-"z","A"-"Z"]>
| <#A2ZN: ["a"-"z","A"-"Z","0"-"9"]>
}

// -------------------------------------------------
// Keyworks : includes operators that are words and should be
// before general things like IDENTIFIER which swallow almost
// anything
<DEFAULT> TOKEN : {
<KW_A: "a">
}

<DEFAULT> TOKEN [IGNORE_CASE] : {
<BASE: "base">
| <PREFIX: "prefix">
| <SELECT: "select">
| <DISTINCT: "distinct">
| <REDUCED: "reduced">
| <DESCRIBE: "describe">
| <CONSTRUCT: "construct">
| <ASK: "ask">
| <LIMIT: "limit">
| <OFFSET: "offset">
| <ORDER: "order">
| <BY: "by">
| <ASC: "asc">
| <DESC: "desc">
| <NAMED: "named">
| <FROM: "from">
| <WHERE: "where">
| <AND: "and">
| <GRAPH: "graph">
| <OPTIONAL: "optional">
| <UNION: "union">
| <FILTER: "filter">
| <BOUND: "bound">
| <STR: "str">
| <DTYPE: "datatype">
| <LANG: "lang">
| <LANGMATCHES: "langmatches">
| <IS_URI: "isURI">
| <IS_IRI: "isIRI">
| <IS_BLANK: "isBlank">
| <IS_LITERAL: "isLiteral">
| <REGEX: "regex">
| <SAME_TERM: "sameTerm">
| <TRUE: "true">
| <FALSE: "false">
}

// -------------------------------------------------
<DEFAULT> TOKEN : {
<#DIGITS: (["0"-"9"])+>
| <INTEGER: <DIGITS>>
| <DECIMAL: <DIGITS> "." (<DIGITS>)* | "." <DIGITS>>
| <DOUBLE: (["0"-"9"])+ "." (["0"-"9"])* <EXPONENT> | "." (["0"-"9"])+ <EXPONENT> | (["0"-"9"])+ <EXPONENT>>
| <INTEGER_POSITIVE: <PLUS> <INTEGER>>
| <DECIMAL_POSITIVE: <PLUS> <DECIMAL>>
| <DOUBLE_POSITIVE: <PLUS> <DOUBLE>>
| <INTEGER_NEGATIVE: <MINUS> <INTEGER>>
| <DECIMAL_NEGATIVE: <MINUS> <DECIMAL>>
| <DOUBLE_NEGATIVE: <MINUS> <DOUBLE>>
| <#EXPONENT: ["e","E"] (["+","-"])? (["0"-"9"])+>
| <#QUOTE_3D: "\"\"\"">
| <#QUOTE_3S: "\'\'\'">
| <ECHAR: "\\" ("t" | "b" | "n" | "r" | "f" | "\\" | "\"" | "\'")>
| <STRING_LITERAL1: "\'" (~["\'","\\","\n","\r"] | <ECHAR>)* "\'">
| <STRING_LITERAL2: "\"" (~["\"","\\","\n","\r"] | <ECHAR>)* "\"">
| <STRING_LITERAL_LONG1: <QUOTE_3S> (("\'" | "\'\'")? (~["\'","\\"] | <ECHAR>))* <QUOTE_3S>>
| <STRING_LITERAL_LONG2: <QUOTE_3D> (("\"" | "\"\"")? (~["\"","\\"] | <ECHAR>))* <QUOTE_3D>>
}

<DEFAULT> TOKEN : {
<LPAREN: "(">
| <RPAREN: ")">
| <NIL: <LPAREN> (<WS> | <SINGLE_LINE_COMMENT>)* <RPAREN>>
| <LBRACE: "{">
| <RBRACE: "}">
| <LBRACKET: "[">
| <RBRACKET: "]">
| <ANON: <LBRACKET> (<WS> | <SINGLE_LINE_COMMENT>)* <RBRACKET>>
| <SEMICOLON: ";">
| <COMMA: ",">
| <DOT: ".">
| <EQ: "=">
| <NE: "!=">
| <GT: ">">
| <LT: "<">
| <LE: "<=">
| <GE: ">=">
| <BANG: "!">
| <TILDE: "~">
| <COLON: ":">
| <SC_OR: "||">
| <SC_AND: "&&">
| <PLUS: "+">
| <MINUS: "-">
| <STAR: "*">
| <SLASH: "/">
| <DATATYPE: "^^">
| <AT: "@">
}

// See XML chars.txt for notes
<DEFAULT> TOKEN : {
<#PN_CHARS_BASE: ["A"-"Z"] | ["a"-"z"] | ["\u00c0"-"\u00d6"] | ["\u00d8"-"\u00f6"] | ["\u00f8"-"\u02ff"] | ["\u0370"-"\u037d"] | ["\u037f"-"\u1fff"] | ["\u200c"-"\u200d"] | ["\u2070"-"\u218f"] | ["\u2c00"-"\u2fef"] | ["\u3001"-"\ud7ff"] | ["\uf900"-"\ufffd"]>
| <#PN_CHARS_U: <PN_CHARS_BASE> | "_">
| <#PN_CHARS: <PN_CHARS_U> | "-" | ["0"-"9"] | "\u00b7" | ["\u0300"-"\u036f"] | ["\u203f"-"\u2040"]>
| <#PN_PREFIX: <PN_CHARS_BASE> ((<PN_CHARS> | ".")* <PN_CHARS>)?>
| <#PN_LOCAL: (<PN_CHARS_U> | ["0"-"9"]) ((<PN_CHARS> | ".")* <PN_CHARS>)?>
| <#VARNAME: (<PN_CHARS_U> | ["0"-"9"]) (<PN_CHARS_U> | ["0"-"9"] | "\u00b7" | ["\u0300"-"\u036f"] | ["\u203f"-"\u2040"])*>
}

// Catch-all tokens.  Must be last.  
// Any non-whitespace.  Causes a parser exception, rather than a
// token manager error (with hidden line numbers).
// Only bad IRIs (e.g. spaces) now give unhelpful parse errors.
<DEFAULT> TOKEN : {
<#UNKNOWN: (~[" ","\t","\n","\r","\f"])+>
}

NON-TERMINALS
	QueryUnit	:=	Query <EOF>
	Query	:=	Prologue ( SelectQuery | ConstructQuery | DescribeQuery | AskQuery )
	Prologue	:=	( BaseDecl )? ( PrefixDecl )*
	BaseDecl	:=	<BASE> IRI_REF
	PrefixDecl	:=	<PREFIX> <PNAME_NS> IRI_REF
// ---- Query type clauses
	SelectQuery	:=	Project ( DatasetClause )* WhereClause SolutionModifier
	SubSelect	:=	Project WhereClause SolutionModifier
	Project	:=	<SELECT> ( <DISTINCT> | <REDUCED> )? ( ( ( Var ) )+ | <STAR> )
	ConstructQuery	:=	<CONSTRUCT> ConstructTemplate ( DatasetClause )* WhereClause SolutionModifier
	DescribeQuery	:=	<DESCRIBE> ( ( VarOrIRIref )+ | <STAR> ) ( DatasetClause )* ( WhereClause )? SolutionModifier
	AskQuery	:=	<ASK> ( DatasetClause )* WhereClause
// ----
	DatasetClause	:=	<FROM> ( DefaultGraphClause | NamedGraphClause )
	DefaultGraphClause	:=	SourceSelector
	NamedGraphClause	:=	<NAMED> SourceSelector
	SourceSelector	:=	IRIref
	WhereClause	:=	( <WHERE> )? GroupGraphPattern
	SolutionModifier	:=	( OrderClause )? ( LimitOffsetClauses )?
	OrderClause	:=	<ORDER> <BY> ( OrderCondition )+
	OrderCondition	:=	( ( ( <ASC> | <DESC> ) BrackettedExpression ) | ( Constraint | Var ) )
	LimitOffsetClauses	:=	( LimitClause ( OffsetClause )? | OffsetClause ( LimitClause )? )
	LimitClause	:=	<LIMIT> <INTEGER>
	OffsetClause	:=	<OFFSET> <INTEGER>
// ---- SPARQL/Update
// ---- General Graph Pattern 
	GroupGraphPattern	:=	<LBRACE> GroupGraphPatternSub <RBRACE>
	GroupGraphPatternSub	:=	( TriplesBlock )? ( ( GraphPatternNotTriples | Filter ) ( <DOT> )? ( TriplesBlock )? )*
// -- TriplesBlock
// Two versions - for SPARQL 1.0 and SPARQL 1.1 (with paths)
	TriplesBlock	:=	TriplesSameSubject ( <DOT> ( TriplesBlock )? )?
// -----
	GraphPatternNotTriples	:=	( OptionalGraphPattern | GroupOrUnionGraphPattern | GraphGraphPattern )
// ---- Definitions of each pattern element
	OptionalGraphPattern	:=	<OPTIONAL> GroupGraphPattern
	GraphGraphPattern	:=	<GRAPH> VarOrIRIref GroupGraphPattern
// G (union G)* can be a single group pattern
// or a group pattern as part of an union.
	GroupOrUnionGraphPattern	:=	GroupGraphPattern ( <UNION> GroupGraphPattern )*
	Filter	:=	<FILTER> Constraint
	Constraint	:=	( BrackettedExpression | BuiltInCall | FunctionCall )
	FunctionCall	:=	IRIref ArgList
	ArgList	:=	( <NIL> | <LPAREN> Expression ( <COMMA> Expression )* <RPAREN> )
// -------- Construct patterns
	ConstructTemplate	:=	<LBRACE> ( ConstructTriples )? <RBRACE>
	ConstructTriples	:=	TriplesSameSubject ( <DOT> ( ConstructTriples )? )?
// -------- Triple lists with property and object lists
// -------- Without paths: entry: TriplesSameSubject
	TriplesSameSubject	:=	VarOrTerm PropertyListNotEmpty
		|	TriplesNode PropertyList
	PropertyListNotEmpty	:=	Verb ObjectList ( <SEMICOLON> ( Verb ObjectList )? )*
	PropertyList	:=	( PropertyListNotEmpty )?
	ObjectList	:=	Object ( <COMMA> Object )*
	Object	:=	GraphNode
	Verb	:=	( VarOrIRIref | <KW_A> )
// -------- BGPs with paths.
// -------- Entry point: TriplesSameSubjectPath
// -------- Paths
// -------- Triple expansions

// Anything that can stand in a node slot and which is
// a number of triples
	TriplesNode	:=	Collection
		|	BlankNodePropertyList
	BlankNodePropertyList	:=	<LBRACKET> PropertyListNotEmpty <RBRACKET>
// ------- RDF collections
	Collection	:=	<LPAREN> ( GraphNode )+ <RPAREN>
// -------- Nodes in a graph pattern or template
	GraphNode	:=	VarOrTerm
		|	TriplesNode
	VarOrTerm	:=	( Var | GraphTerm )
// Property (if no bNodes) + DESCRIBE
	VarOrIRIref	:=	( Var | IRIref )
	Var	:=	( <VAR1> | <VAR2> )
	GraphTerm	:=	IRIref
		|	RDFLiteral
		|	NumericLiteral
		|	BooleanLiteral
		|	BlankNode
		|	<NIL>
// -------- Constraint syntax
	Expression	:=	ConditionalOrExpression
	ConditionalOrExpression	:=	ConditionalAndExpression ( <SC_OR> ConditionalAndExpression )*
	ConditionalAndExpression	:=	ValueLogical ( <SC_AND> ValueLogical )*
	ValueLogical	:=	RelationalExpression
	RelationalExpression	:=	NumericExpression ( <EQ> NumericExpression | <NE> NumericExpression | <LT> NumericExpression | <GT> NumericExpression | <LE> NumericExpression | <GE> NumericExpression )?
	NumericExpression	:=	AdditiveExpression
	AdditiveExpression	:=	MultiplicativeExpression ( <PLUS> MultiplicativeExpression | <MINUS> MultiplicativeExpression | ( NumericLiteralPositive | NumericLiteralNegative ) ( ( <STAR> UnaryExpression ) | ( <SLASH> UnaryExpression ) )? )*
	MultiplicativeExpression	:=	UnaryExpression ( <STAR> UnaryExpression | <SLASH> UnaryExpression )*
	UnaryExpression	:=	<BANG> PrimaryExpression
		|	<PLUS> PrimaryExpression
		|	<MINUS> PrimaryExpression
		|	PrimaryExpression
	PrimaryExpression	:=	( BrackettedExpression | BuiltInCall | IRIrefOrFunction | RDFLiteral | NumericLiteral | BooleanLiteral | Var )
	BrackettedExpression	:=	<LPAREN> Expression <RPAREN>
	BuiltInCall	:=	<STR> <LPAREN> Expression <RPAREN>
		|	<LANG> <LPAREN> Expression <RPAREN>
		|	<LANGMATCHES> <LPAREN> Expression <COMMA> Expression <RPAREN>
		|	<DTYPE> <LPAREN> Expression <RPAREN>
		|	<BOUND> <LPAREN> Var <RPAREN>
		|	<SAME_TERM> <LPAREN> Expression <COMMA> Expression <RPAREN>
		|	<IS_IRI> <LPAREN> Expression <RPAREN>
		|	<IS_URI> <LPAREN> Expression <RPAREN>
		|	<IS_BLANK> <LPAREN> Expression <RPAREN>
		|	<IS_LITERAL> <LPAREN> Expression <RPAREN>
		|	RegexExpression
	RegexExpression	:=	<REGEX> <LPAREN> Expression <COMMA> Expression ( <COMMA> Expression )? <RPAREN>
// The case of "q:name()" or just "q:name"
// by expanding out FunctionCall()
	IRIrefOrFunction	:=	IRIref ( ArgList )?
	RDFLiteral	:=	String ( ( <LANGTAG> ) | ( <DATATYPE> IRIref ) )?
	NumericLiteral	:=	( NumericLiteralUnsigned | NumericLiteralPositive | NumericLiteralNegative )
	NumericLiteralUnsigned	:=	<INTEGER>
		|	<DECIMAL>
		|	<DOUBLE>
	NumericLiteralPositive	:=	<INTEGER_POSITIVE>
		|	<DECIMAL_POSITIVE>
		|	<DOUBLE_POSITIVE>
	NumericLiteralNegative	:=	<INTEGER_NEGATIVE>
		|	<DECIMAL_NEGATIVE>
		|	<DOUBLE_NEGATIVE>
	BooleanLiteral	:=	<TRUE>
		|	<FALSE>
	String	:=	( <STRING_LITERAL1> | <STRING_LITERAL2> | <STRING_LITERAL_LONG1> | <STRING_LITERAL_LONG2> )
	IRIref	:=	IRI_REF
		|	PrefixedName
	PrefixedName	:=	( <PNAME_LN> | <PNAME_NS> )
	BlankNode	:=	<BLANK_NODE_LABEL>
		|	<ANON>
	IRI_REF	:=	<IRIref>

DOCUMENT END