/* Prolog syntax rules for turning EBNF+annotations into RDF statements using a grammar vocabulary. The EBNF dialect here is that used by the XML 1.0 Recommendation, as extracted by: lynx -source http://www.w3.org/TR/2000/REC-xml-20001006 > REC-xml.html lynx -dump -nolist -width=65535 http://www.w3.org/TR/2000/REC-xml-20001006 | grep -v '^symbol' | grep '::=' | cut -b11- | sed 's;^ *;;' > xml-grammar OOPS -- misses some with explicit brs? --> should start on ::= line and go on until blank line or another ::= line. perl? When this works, you should be able to feed it that xml-grammar file and it will return a large set of RDF triples which can in turn be used to parse XML files. Like yacc, etc, we extend the EBNF with annotations. Any rule may be followed by a logical expression in curly-braces. That expression should be in language [...which one?...] and refers to elements in the rule by number -- but it can refer to the value or the text. Also, the default rules for things like X* are reasonable, making it a list. a ::= b{2,4} == a : b b { $0 = ($1, $2) } | b b b { $0 = ($1, $2, $3) } | b b b b { $0 = ($1, $2, $3, $4) } */ % auto_table breaks IO! % :- auto_table. :- table ruleset/3. :- table opgap/2. :- table gap/2. :- table word/2. :- table string_contents1/2. :- table string_contents2/2. :- table charset_parts/2. :- table not_star_slash/2. :- table pattern/3. :- import append/3 from basics. :- import member/2 from basics. % ruleset --> opgap; rule, opgap, [10], opgap, ruleset. ruleset([]) --> opgap. ruleset(S) --> opgap, rule(R), opgap, ruleset(SB), {append(SB, [R], S)}. rule(R) --> word(WordText), opgap, "::=", opgap, pattern(P), {R=rule(WordText, P)}, {write('Parsed Rule: '), name(N, WordText), writeln(N)}. pattern(P) --> atom(P) ; expr(P). atom(A) --> word(Text), { A=class(Text) } ; char ; string ; charset ; "(", opgap, pattern(A), opgap, ")" . expr(E) --> seq(E) ; alt(E) ; except(E) ; repeat(E) . %opgap --> [] ; gap. opgap --> [] ; opgap, whitespace. gap --> whitespace ; whitespace, gap. whitespace([Head|Tail], Tail) :- isspace(Head). % whitespace --> [10]; [8]; [13]; " ". lower --> "a"; "b"; "c"; "d"; "e"; "f"; "g"; "h"; "i"; "j"; "k"; "l"; "m"; "n"; "o"; "p"; "q"; "r"; "s"; "t"; "u"; "v"; "w"; "x"; "y"; "z". upper --> "A"; "B"; "C"; "D"; "E"; "F"; "G"; "H"; "I"; "J"; "K"; "L"; "M"; "N"; "O"; "P"; "Q"; "R"; "S"; "T"; "U"; "V"; "W"; "X"; "Y"; "Z". letter --> lower ; upper. digit --> "1"; "2"; "3"; "4"; "5"; "6"; "7"; "8"; "9"; "0". hexdigit --> "1"; "2"; "3"; "4"; "5"; "6"; "7"; "8"; "9"; "0"; "A"; "B"; "C"; "D"; "E"; "F". normal_punctuation --> " "; "`"; "="; "~"; "!"; "@"; "$"; "%"; "&"; "("; ")"; "_"; "+"; "|"; "["; "{"; "}"; ";"; ","; "."; "/"; "<"; ">"; "?". special_punctuation --> ":", "'"; "\"; """"; "-"; "]"; "^"; "#"; "*"; "/". not_special --> letter; digit; normal_punctuation. not_apos --> not_special; ":"; "\"; """"; "-"; "]"; "^"; "#"; "*"; "/". not_quote --> not_special; ":"; "'"; "\"; "-"; "]"; "^"; "#"; "*"; "/". word --> letter, ([] ; word). word(Text, Li, Lo) :- word(Li, Lo), append(Text, Lo, Li). char --> "#x", hexdigit. char --> "#x", hexdigit, hexdigit. char --> "#x", hexdigit, hexdigit, hexdigit. char --> "#x", hexdigit, hexdigit, hexdigit, hexdigit. string --> "'", string_contents1, "'". string_contents1 --> []; not_apos, string_contents1. string --> """", string_contents2, """". string_contents2 --> []; not_quote, string_contents2. seq(E) --> pattern(E1), gap, pattern(E2), { E=seq(E1,E2) }. alt(E) --> pattern(E1), opgap, "|", opgap, pattern(E2), { E=or(E1,E2) }. except(E) --> pattern(E1), opgap, "-", opgap, pattern(E2), { E=and(E1,not(E2)) }. repeat(E) --> pattern(E1), opgap, "+", { E=repeat(E1, 1, max) } ; pattern(E1), opgap, "?", { E=repeat(E1, 0, 1) } ; pattern(E1), opgap, "*", { E=repeat(E1, 0, max) }. % what elements can occur inside the [...] char-set construct? % omit: "-"; "]"; "^"; "#"; ":" if you want them use is #x construct setchar --> char; not_special; "'"; "\"; """"; "*"; "/". charset --> "[", charset_parts , "]" % normal ; "[-", charset_parts, "]" % with dash ; "[", charset_parts, "-]" % with dash ; "[^", charset_parts, "]" % inverted ; "[^-", charset_parts, "]" % inverted with dash ; "[^", charset_parts, "-]". % inverted with dash charset_parts --> [] ; setchar, charset_parts ; setchar, "-", setchar, charset_parts. % disjoint from charset because of the ":" constraint --> "[", opgap, "VC:", constraint_expression, opgap, "]" ; "[", opgap, "WFC:", constraint_expression, opgap, "]". % this is so ugly because we're not using charset inverses % we're not doing nesting here. comment --> "/*", not_star_slash, "*/". not_star_or_slash --> not_special; ":", "'"; "\"; """"; "-"; "]"; "^"; "#". not_star_slash --> [] ; not_star_or_slash, not_star_slash ; "*", not_star_or_slash, not_star_slash ; "/", not_star_slash. :- import file_open/3, file_close/2, file_getbuf_list/4 from file_io. :- import length/2 from basics. do(Filename) :- read_string(Filename, Contents), writeln(''), writeln('------Parsing-------'), name(X, Contents), writeln(X), writeln('------Results--------'), ruleset(R, Contents, ""), writeln(R), writeln('---------------------'). xxread_string(Filename, Contents) :- seeing(Previous), see(Filename), % get0(A), % get0(B), % get0(C), % get0(D), % Contents=[A,B,C,D], getall(Contents), see(Previous). getall(L) :- writeln('getall1'), get0(N), writeln('getall2'), !, (N = -1, L=[], writeln('EOF.')); (write('Not done, read: '), writeln(N), getall(Rest), append([N], Rest, L)). read_string(Filename, Contents) :- file_open(Filename, 0, IOPort), getall(IOPort, Contents), !. % file_close(IOPort). getall(IOPort, Contents) :- (file_read_line_list(IOPort, Line), %name(X, Line), %write('Read line: '), %writeln(X), getall(IOPort, Rest), append(Line, Rest, Contents)) %writeln('Appended')) ; %writeln('Done'), Contents = [].