Syntax of Query Language Dialect 2

The following EBNF production rules define the syntax of the extensible Indexing Service query language, Query Dialect 2. Terminal symbols are case insensitive: or, OR, Or, and oR are equivalent. The production rules give terminal symbols in lowercase only.

Query =                                                (* start symbol *)

( [ Query ( or { · }- | | )] AndTerm )

| ( not { · }- | ! ) Query ;

AndTerm =

NearTerm

| ( ( not { · }- | ! ) PropTerm )

| ( AndTerm ( and { · }- | & ) [ ( not { · }- | ! ) ] NearTerm ) ;

NearTerm =

CoerceTerm

| ( NearTerm ( {near { · }- | ~ ) CoerceTerm )

| ( NearTerm NearElement } CoerceTerm ) ;

PropTerm =

( Property [ = ] Regex [ {/prop} ] )

| ( Property Operator Value [ {/prop} ] ) ;

CoerceTerm =

( [ Generate ] NestTerm [ {/generate} ] )

‹| ( [ Generate ] {/coerce} [ Generate ] NestTerm [ {/generate} ] )

›| ( [ Generate ] Weight [ Generate ] NestTerm [ {/generate} ] ) ;

NearElement =

White

| ,

| ( dist { · } = { · } Digits )

| ( unit { · } = { · } ‹( ›word ‹| sent | par | chap ) ›) ;

Property =

[ LongPropName | ShortPropName ] ;

(* The following rule only prepares for the query engine, which does *)

(* the final parsing. See Pattern-Matching Queries for the full syntax. *)

Regex =

( {regex} " [ AnyExceptQuote ] " {/regex} )

| ( {regex} [ AnyExceptCurlyOpen ] {/regex} )

| ( {regex} [ [ AnyExceptNewline ] | [ AnyExceptNewline ] ] {/regex} ) ;

Operator =

= | != | > | >= | < | <= | ^a | ^s

| ( = { · }^s | ^s { · } = )

| ( != { · } ^s | ^s { · } != )

| ( > { · } ^s | ^s { · } > )

| ( >= { · } ^s | ^s { · } >= )

| ( < { · } ^s | ^s { · } < )

| ( <= { · } ^s | ^s { · } <= )

| ( ^s { · } ^a )

| ( ^s { · } ^s )

| ( ^a { · } = )

| ( ^a { · } != )

| ( ^a { · } > )

| ( ^a { · } >= )

| ( ^a { · } < )

| ( ^a { · } <= )

| ( ^a { · } ^a )

| ( ^a { · } ^s ) ;

Value =

Phrase

| Freetext

| VectorValue ;

Generate =

( {generate { · }- method { · } = { · } prefix { · } } )

| ( {generate { · }- method { · } = { · } inflect { · } } ) ;

NestTerm =

VectorTerm

| Term

| ( ( Query ) ) ;

Weight =

{weight { · }- value { · } = { · } ( 0 | 1 | ( 0. { DecimalDigit } ) | ( 1. { 0 } ) | ( . { DecimalDigit }- ) ) ;

LongPropName =

( {prop { · }- name { · } = { · } " [ AnyExceptQuote ] " { · } } )

| ( {prop { · }- name { · } = { · } AnyExceptQuoteCurlyCloseSpace [ AnyExceptCurlyClose ] } ) ;

ShortPropName =

( ( @ | $ | # ) " AnyExceptQuote " )

| ( ( @ | $ | # ) AnyExceptExcluded ) ;

Phrase =

( LongPhrase | ShortPhrase ) [ * | ** ] ;

Freetext =

( LongFreetext | ShortFreetext ) [ * | ** ] ;

VectorValue =

EmptyVector

| SingletVector

| MultiVector ) ;

LongPhrase =

( {phrase} { · } " [ AnyExceptQuote ] " { · } {/phrase} )

| ( {phrase} [ AnyExceptCurlyOpen ] {/phrase} ) ;

ShortPhrase =

( " [ AnyExceptQuote ] " )

| ( AnyExceptMany [ AnyExceptOp ] [ · ] ) ;

LongFreetext =

( {freetext} { · } " [ AnyExceptQuote ] " { · } {/freetext} )

| ( {freetext} [ AnyExceptCurlyOpen ] {/freetext} ) ;

ShortFreetext =

AnyExceptMany [ AnyExceptOp ] [ · ] ;

VectorTerm =

VectorSpec {/vector} ;

Term =

PropTerm

| ( Property [ contains { · }- ] Phrase [ {/prop} ] )

| ( Property [ contains { · }- ] Generate Phrase [ {/generate} ] [ {/prop} ) ]

| ( Property [ contains { · }- ] Query )

| ( Property [ contains { · }- ] Freetext [ {/prop} ] ) ;

EmptyVector =

( ) ;

SingletVector =

( ( Phrase ) )

| ( ( Freetext ) ) ;

MultiVector =

( VectorElement VectorElement )

| ( MultiVector VectorElement ) ;

VectorSpec =

( VectorMethod {ve} Query )

| ( VectorSpec {ve} Query ) ;

VectorElement =

White

| ;

| ( " [ AnyExceptQuote ] " )

| ( AnyExceptVector [ AnyExceptSemiClose ] ( ; | ) ) ) ;

VectorMethod =

( {vector { · }- rankmethod { · } = { · } " [ AnyExceptQuote ] " { · } } )

| ( {vector { · }- rankmethod { · } = [ AnyExceptCurlyClose ] } ) ;

Digits =

{ DecimalDigit }- ;

DecimalDigit =

0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 ;

White =

· | \t | \n | \f | \r ;

AnyExceptQuote =

? Any character sequence ? - " ;        (* quoted " ( = "" ) is OK *)

AnyExceptNewline =

? Any character sequence ? - \n ;

AnyExceptCurlyOpen =

? Any character sequence ? - { ;

AnyExceptCurlyClose =

? Any character sequence ? - } ;

AnyExceptSemiClose =

? Any character sequence ? - ( ; | ) ) ;

AnyExceptQuoteCurlyCloseSpace =

? Any character sequence ? - ( " | } | · ) ;

AnyExceptVector =

? Any character sequence ? - ( " | ; | · ) ;

AnyExceptOp =

? Any character sequence ? - ( & | ~ | | | { | ) | · ) ;

AnyExceptExcluded =

? Any character sequence ? - ( " | · | < | > | = | ! | & | | | ~ | ^ ) ;

AnyExceptMany =

? Any character sequence ? - ( # | $ | @ | ~ | & | | | < | > | = | ! | ^| * | " | ( | ) | { | · ) ;