konwersja leksykonu do pdf'a

Wojciech Jaworski
1 parent be1d1198
Showing 10 changed files with 76 additions and 31 deletions
LCGlexicon/.gitignore
LCGlexicon/ENIAM_LCGlexicon.ml
LCGlexicon/ENIAM_LCGlexiconLatexOf.ml
LCGlexicon/ENIAM_LCGlexiconPL.ml
LCGlexicon/ENIAM_LCGlexiconParser.ml
LCGlexicon/ENIAM_LCGlexiconTypes.ml
LCGlexicon/ENIAM_LCGrendererPL.ml
LCGlexicon/ENIAMcategoriesPL.ml
LCGlexicon/TODO
LCGlexicon/makefile
 test
+print_lexicon
 results/*
@@ -22,25 +22,26 @@ open ENIAM_LCGtypes
 open ENIAM_LCGlexiconTypes
 open ENIAMcategoriesPL
  
-let grammar = load_lexicon "data/grammar.sh"
-
-type rule2 =
-    Basic of string
-  | Quant of (selector * string) list * string
-  | Raised of (selector * string) list * string * selector list
-  | Quot of (selector * string) list * string
-  | Inclusion of string
-  | Conj of (selector * string) list * string
-  | Bracket of string
-
-
 let rec find_selector s = function
     (t,Eq,x :: _) :: l -> if t = s then x else find_selector s l
   | (t,_,_) :: l -> if t = s then failwith "find_selector 1" else find_selector s l
   | [] -> failwith "find_selector 2"
  
+let rec get_syntax rev = function
+    Syntax syntax :: rule -> syntax, (List.rev rev) @ rule
+  | t :: rule -> get_syntax (t :: rev) rule
+  | [] -> failwith "get_syntax"
+
+let rec get_quant rev = function
+    Quant quant :: rule -> quant, (List.rev rev) @ rule
+  | t :: rule -> get_quant (t :: rev) rule
+  | [] -> [],  List.rev rev
+
+let rec get_bracket rev = function
+    Bracket :: rule -> true, (List.rev rev) @ rule
+  | t :: rule -> get_bracket (t :: rev) rule
+  | [] -> false, List.rev rev
  
-(* FIXME: sprawdzić, czy to nie wycina Ctype *)
 let merge_quant pos_quants quants =
   let map = Xlist.fold quants SelectorMap.empty (fun map (k,v) -> SelectorMap.add map k v) in
   let l,map = Xlist.fold pos_quants ([],map) (fun (l,map) (cat,v) ->
@@ -48,6 +49,34 @@ let merge_quant pos_quants quants =
       else (cat,v) :: l, map) in
   List.rev (SelectorMap.fold map l (fun l cat v -> (cat,v) :: l))
  
+let assign_quantifiers (selectors,rule,weight) =
+  let pos = find_selector Pos selectors in
+  let categories =
+    try StringMap.find pos_categories pos
+    with Not_found -> failwith ("assign_quantifiers: " ^ pos) in
+  let categories = Xlist.map categories (fun s -> s,Top) in
+  let syntax,rule = get_syntax [] rule in
+  let quant,rule = get_quant [] rule in
+  let bracket,rule = get_bracket [] rule in
+  let quant = merge_quant categories quant in
+  selectors, (bracket,quant,syntax),(rule,weight)
+
+let _ =
+  let lexicon = ENIAM_LCGlexiconParser.load_lexicon "resources/lexicon-pl.dic" in
+  List.rev (Xlist.rev_map lexicon assign_quantifiers)
+
+(***
+type rule2 =
+    Basic of string
+  | Quant of (selector * string) list * string
+  | Raised of (selector * string) list * string * selector list
+  | Quot of (selector * string) list * string
+  | Inclusion of string
+  | Conj of (selector * string) list * string
+  | Bracket of string
+
+
+
 let parse_quants_range quant =
   Xlist.map quant (fun (cats,v) -> cats, parse_quant_range (cats,v))
 (**
@@ -285,3 +314,4 @@ let create (paths,last) tokens lex_sems =
   let chart = IntIntSet.fold set chart (fun chart (i,j) -> LCGchart.make_unique chart i j) in
   chart
 *)
+  ***)
@@ -17,6 +17,10 @@
  *  along with this program.  If not, see <http://www.gnu.org/licenses/>.
  *)
  
+open ENIAM_LCGtypes
+open ENIAM_LCGlexiconTypes
+open ENIAMcategoriesPL
+
 let rec add_quantifiers t = function
     [] -> t
   | (cat,s) :: l ->  add_quantifiers (WithVar(string_of_selector cat,s,"",t)) l
@@ -89,15 +93,19 @@ let latex_of_selectors selectors =
       let rel = if rel = Eq then "=" else "!=" in
       ENIAMcategoriesPL.string_of_selector cat ^ rel ^ (String.concat "|" l)))
  
-let print_latex_grammar grammar =
-  Printf.printf "grammar size: %d\n" (Xlist.size grammar);
-  Xlatex.latex_file_out "results/" "grammar" "a0" false (fun file ->
-      Xlist.iter grammar (fun (selectors,(bracket,quant,syntax,semantics),weight) ->
+let print_latex_lexicon lexicon =
+  Printf.printf "lexicon size: %d\n" (Xlist.size lexicon);
+  Xlatex.latex_file_out "results/" "lexicon" "a0" false (fun file ->
+      Xlist.iter lexicon (fun (selectors,(bracket,quant,syntax),semantics) ->
           let syntax = add_quantifiers_simple syntax (List.rev quant) in
           Printf.fprintf file "%s\\\\\n$\\begin{array}{l}%s\\end{array}$\\\\\\;\\\\\\;\\\\\n" (latex_of_selectors selectors) (latex_of_grammar_symbol 0 syntax)));
-  Xlatex.latex_compile_and_clean "results/" "grammar"
-
+  Xlatex.latex_compile_and_clean "results/" "lexicon"
  
-(* let grammar = parse_grammar grammar *)
  
-(* let _ = print_latex_grammar grammar *)
+let _ =
+  if Array.length Sys.argv < 3 then
+    print_endline "missing argument\nUsage: print_lexicon <input-file> <output-file>"
+  else
+    let lexicon = ENIAM_LCGlexiconParser.load_lexicon Sys.argv.(1) in
+    let lexicon = List.rev (Xlist.rev_map lexicon ENIAM_LCGlexicon.assign_quantifiers) in
+    print_latex_lexicon lexicon
@@ -18,8 +18,8 @@
  *)
  
 open Xstd
-open ENIAM_LCGlexiconTypes
 open ENIAM_LCGtypes
+open ENIAM_LCGlexiconTypes
 open ENIAMcategoriesPL
  
 let rec get_first n = function
@@ -282,12 +282,6 @@ let parse_quantifiers tokens =
       | t :: _ -> failwith ("parse_quantifiers: unexpected token '" ^ t ^ "'")
       | [] -> failwith "parse_quantifiers: no token")
  
-type rule =
-    Bracket
-  | Quant of (selector * internal_grammar_symbol) list
-  | Raised of string list
-  | Syntax of grammar_symbol
-
 let rec parse_rule atoms = function
     "BRACKET" :: tokens -> Bracket :: parse_rule atoms tokens
   | "QUANT" :: "[" :: tokens ->
@@ -46,6 +46,13 @@ type rule_sem =
  
 type selector_relation = Eq | Neq (*| StrictEq*)
  
+type rule =
+    Bracket
+  | Quant of (selector * ENIAM_LCGtypes.internal_grammar_symbol) list
+  | Raised of string list
+  | Syntax of ENIAM_LCGtypes.grammar_symbol
+
+
 (* x="s" oznacza, że żeby reguła została użyta token musi mieć "s" jako jedną z wartości atrybutu x, reguła zostanie wykonana dla x z usuniętymi pozostałymi wartościami *)
 (* x!="s" oznacza, że żeby reguła została użyta token musi mieć jako jedną z wartości atrybutu x symbol inny od "s", reguła zostanie wykonana dla x z usuniętą wartością "s" *)
 (* x=="s" oznacza, że żeby reguła została użyta token musi mieć "s" jako jednyną z wartość atrybutu x *)
@@ -526,4 +526,4 @@ let pos_categories = Xlist.fold [
     "burk",[Lemma;];
     "interp",[Lemma;];
     "unk",[Lemma;Number;Case;Gender;Person;];
-  ] StringMap.empty (fun map (k,l) -> StringMap.add map k (Xlist.map l (fun v -> v,"T")))
+  ] StringMap.empty (fun map (k,l) -> StringMap.add map k l)
 - unumber jako atrybut
  
 "Można było" - brakuje uzgodnienia rodzaju przymiotnika w przypadku predykatywnym, i ogólnie kontroli składniowej
+
+- dodać do wyświetlania leksykonu informacje o bracetach i semantyce
@@ -27,6 +27,10 @@ test: test.ml $(SOURCES)
 	mkdir -p results
 	$(OCAMLOPT) -o test $(OCAMLOPTFLAGS) $(SOURCES) test.ml
  
+print_lexicon: ENIAM_LCGlexiconLatexOf.ml $(SOURCES)
+	mkdir -p results
+	$(OCAMLOPT) -o print_lexicon $(OCAMLOPTFLAGS) $(SOURCES) ENIAM_LCGlexiconLatexOf.ml
+
 .SUFFIXES: .mll .mly .ml .mli .cmo .cmi .cmx
  
 .mll.ml:
@@ -48,4 +52,4 @@ test: test.ml $(SOURCES)
 	$(OCAMLOPT) $(OCAMLOPTFLAGS) -c $<
  
 clean:
-	rm -f *~ *.cm[aoix] *.o *.so *.cmxa *.a test
+	rm -f *~ *.cm[aoix] *.o *.so *.cmxa *.a test print_lexicon