Grammars for the specification of links between terms
"?" means 0 or 1 times, "*" means 0 to N times, "+" means 1 to N times
Commands := Command ((";"|".") Command)*
Command := TermCommand | CGcommand
CGcommand := CG | "?" CG | "^" CG
TermCommand := TermDecl | "?" Term | "^" Term
TermDecl := Term RelSignature? Annotation? (LinksOf1Kind ("," LinksOf1Kind)*)?
RelSignature := "(" Term "," Term ")"
Annotation := "(^" ... "^)"
LinksOf1Kind := LinkKind (DestTerm | TermPartition)+
LinkKind := ">" | "<" | "!" | "~" | ":" | "_" | Identifier
TermPartition:= "{" DestTerm+ "}"
DestTerm := Term LinkContext?
LinkContext := "(" Creator Community? ")"
Term := Identifier | String
String := "\"" ..."\"" | "'" ..."'" | "$(" ... ")$"
Identifier := TermLetter1 TermLetter*
TermLetter1 := [a-z0-9_] | "#"[a-z0-9_]
TermLetter := [a-z0-9_] | "#" | "-" | "/" | "?" | "&" | "@" | "~" | "'" |"."
| "://" //thus a URL may be a Term
Number := ("+"|"-")? Digit+ ("." Digit* )?
Digit := [0-9]
//Additional notes on the lexical parsing:
- white spaces and the HTML imbreakable space encoding " " are ignored
- Java/C++ comments ("/* ... */" and "//...") are ignored
- HTML tags are ignored but the content of HTML comments is parsed
- annotations are enclosed within "(^" and "^)"
- strings may be single quoted, double quoted or enclosed within "$(" and ")$"
Lex&Yacc grammars
<![CDATA[
%token IDENTIFIER NUMERAL STRING ANNOTATION CG
%start COMMANDS
%%
COMMANDS : COMMAND ';'
| COMMANDS COMMAND ';'
;
COMMAND : TERM_DECL | CG_COMMAND
;
CG_COMMAND : CG
| '?' CG
;
TERM_DECL : TERM _ANNOT _LINKS
;
TERM : TERM_NAME _SIGNATURE
;
_ANNOT :
| ANNOTATION
;
_SIGNATURE :
| '(' R_DOMAIN ',' R_RANGE ')'
;
R_DOMAIN : IDENTIFIER
;
R_RANGE : IDENTIFIER
;
_LINKS : | LINKS;
LINKS : LINKS1_S
| LINKS1_S ',' LINKS
;
LINKS1_S : LINK_KIND LINK_DESTS
;
LINK_DESTS : LINK_DEST
| LINK_DEST LINK_DESTS
;
LINK_DEST : DEST_TERM
| '{' DEST_TERMS '}' _LINK_CTXT
;
DEST_TERMS : DEST_TERM _DEST_TERMS;
_DEST_TERMS: | DEST_TERMS;
DEST_TERM : DEST_NAME _LINK_CTXT
;
DEST_NAME : IDENTIFIER
;
_LINK_CTXT :
| '(' CREATOR _COMMUNITY ')'
;
CREATOR : IDENTIFIER
;
_COMMUNITY :
| IDENTIFIER
;
LINK_KIND : '>'
| '<'
| '='
| '~'
| ':'
| '^'
| IDENTIFIER
;
TERM_NAME : IDENTIFIER
| STRING
;
%%
DIGIT [0-9]
ALPHA [a-z]
ALNUM [a-z0-9]
ANY (.)
S ([ \t ]|" ")+
IDENT0 ({ALNUM}|"_")
IDENT1 ({IDENT0}|"#"{IDENT0})
IDENTn ({IDENT0}|"-"|"/"|"?"|"&"|"@"|"#"|"~"|"'"|"."|"://")
IDENT {IDENT1}({IDENTn})*
NUMBER ("+"|"-")?{DIGIT}+("."{DIGIT}*)?
OTHER_OPER ("=>"|"<=>"|"="|"=<"|">="|">")
IDENT_OPER ({IDENT}|"!="|"<="|"=<"|{OTHER_OPER})
ANNOT_ALL "(^"([^\^]|(\^[^\)]))*"^)"
ANNOT_END ([^\^]|(\^[^\)]))*"^)"
DELIM_STR "$("([^\)]|(\)[^\$]))*")$"
DELIM_END ([^\)]|(\)[^\$]))*")$"
QSTR1 '([^'\\]|(\\{ANY}))*'
QSTR1_END ([^'\\]|(\\{ANY}))*'
QSTR2 \"([^"\\]|(\\{ANY}))*\"
QSTR2_END ([^"\\]|(\\{ANY}))*\"
QUOTED_STR ({QSTR1}|{QSTR2})
ANY_VALUE ({IDENT_OPER}|{QUOTED_STR})
%x ALIAS1 ALIAS2 ANNOT1 DELIM1 C_COMM
%%
"/*" BEGIN(C_COMM);
<C_COMM>"\n" LineNumber++;
<C_COMM>"*/" BEGIN(0);
<C_COMM>. ;
{S} ;
"\n" LineNumber++;
"//"[^\n]* ;/* skip C++ line comments */
"<!--" ;/* no wk_skipHTMLComments(): content parsed */
"-->" ;
"<!" wk_skipTag();
"<"{ALPHA} wk_skipTag();
"(^" BEGIN(ANNOT1);
<ANNOT1>{ANNOT_END} {LineNumber+=nbLinesIn(wktext);
wktext[wkleng-2]='\0';BEGIN(0);return ANNOTATION;}
"$(" BEGIN(DELIM1);
<DELIM1>{DELIM_END} {LineNumber+=nbLinesIn(wktext);
wktext[wkleng-2]='\0';BEGIN(0);return STRING;}
{QUOTED_STR} {LineNumber+=nbLinesIn(wktext); return STRING;}
{NUMBER} return NUMERAL;
{IDENT} return IDENTIFIER;
. return wktext[0];
%%
]]>
Philippe Martin
Last modified: Fri Jul 28 20:42:46 MEST 2000