Zmiana comparp

Wojciech Jaworski
1 parent 84afa6af
Showing 11 changed files with 53 additions and 37 deletions
LCGlexicon/ENIAMcategoriesPL.ml
LCGlexicon/resources/lexicon-pl.dic
exec/ENIAMexec.ml
exec/ENIAMvisualization.ml
lexSemantics/ENIAMadjuncts.ml
lexSemantics/ENIAMlexSemantics.ml
lexSemantics/ENIAMlexSemanticsHTMLof.ml
lexSemantics/ENIAMlexSemanticsStringOf.ml
lexSemantics/ENIAMlexSemanticsTypes.ml
lexSemantics/ENIAMvalence.ml
lexSemantics/ENIAMwalRenderer.ml
@@ -159,7 +159,9 @@ let num_nsem lemma =
 let part_set = StringSet.of_list ["się"; "nie"; "by"; "niech"; "niechaj"; "niechże"; "niechajże"; "czy"; "gdyby"]
-let clarify_categories proper cat coerced snode = function
+let snode = SelectorMap.find selector_values SNode
+
+let clarify_categories proper cat coerced (*snode*) = function
     lemma,"subst",[numbers;cases;genders] ->
       let numbers = expand_numbers numbers in
       let cases = expand_cases cases in
 @PHRASE_NAMES
   lex infp np prepnp adjp ip cp ncp advp padvp
-  adja prepadjp comprepnp compar measure num aglt aux-fut
+  adja prepadjp comprepnp comparp measure num aglt aux-fut
   aux-past aux-imp qub interj hyphen int
   rparen rparen2 rquot rquot2 rquot3 inclusion
   day-interval day-lex day-month-interval date-interval
@@ -153,8 +153,8 @@ lemma=w,pos=prep,case=loc:      prepnp*lemma*case{\(1+advp*T),/(day-month+day+ye
 # komparatywy
 # FIXME: trzeba poprawić comparnp i comparpp w walencji
-pos=compar: QUANT[case=nom&gen&dat&acc&inst] compar*lemma*case{\(1+advp*T),/(np*T*case*T*T+adjp*T*case*T)}{\(1+qub),/(1+inclusion)};
-pos=compar: QUANT[case=postp] compar*lemma*case{\(1+advp*T),/(prepnp*T*T+prepadjp*T*T)}{\(1+qub),/(1+inclusion)};
+pos=compar: QUANT[case=nom&gen&dat&acc&inst] comparp*lemma*case{\(1+advp*T),/(np*T*case*T*T+adjp*T*case*T)}{\(1+qub),/(1+inclusion)};
+pos=compar: QUANT[case=postp] comparp*lemma*case{\(1+advp*T),/(prepnp*T*T+prepadjp*T*T)}{\(1+qub),/(1+inclusion)};
 # frazy przymiotnikowe
 # FIXME: let grad = match grads with [grad] -> grad | _ -> failwith "make_adjp: grad" in
@@ -64,16 +64,16 @@ let rec translate_text = function
   | ENIAMsubsyntaxTypes.AltText l -> AltText(Xlist.map l (fun (mode,text) ->
       translate_mode mode, translate_text text))
-let clarify_categories cats snode token =
+let clarify_categories cats (*snode*) token =
   match token.ENIAMtokenizerTypes.token with
     ENIAMtokenizerTypes.Lemma(lemma,pos,interp) ->
       List.flatten (Xlist.map interp (fun interp -> List.flatten (Xlist.map cats (fun (cat,coerced) ->
         (* Printf.printf "lemma=%s pos=%s cat=%s coerced=%s\n%!" lemma pos cat (String.concat "," coerced); *)
-        ENIAMcategoriesPL.clarify_categories false cat coerced snode (lemma,pos,interp)))))
+        ENIAMcategoriesPL.clarify_categories false cat coerced (*snode*) (lemma,pos,interp)))))
   | ENIAMtokenizerTypes.Proper(lemma,pos,interp,senses2) ->
-      List.flatten (Xlist.map interp (fun interp -> List.flatten (Xlist.map cats (fun (cat,coerced) -> ENIAMcategoriesPL.clarify_categories true cat coerced snode (lemma,pos,interp)))))
+      List.flatten (Xlist.map interp (fun interp -> List.flatten (Xlist.map cats (fun (cat,coerced) -> ENIAMcategoriesPL.clarify_categories true cat coerced (*snode*) (lemma,pos,interp)))))
   | ENIAMtokenizerTypes.Interp lemma ->
-      List.flatten (Xlist.map cats (fun (cat,coerced) -> ENIAMcategoriesPL.clarify_categories false cat coerced snode (lemma,"interp",[])))
+      List.flatten (Xlist.map cats (fun (cat,coerced) -> ENIAMcategoriesPL.clarify_categories false cat coerced (*snode*) (lemma,"interp",[])))
   | _ -> []
 let create_chart rules tokens lex_sems paths last =
@@ -85,8 +85,8 @@ let create_chart rules tokens lex_sems paths last =
       ENIAM_LCGrenderer.reset_variable_names ();
       ENIAM_LCGrenderer.add_variable_numbers ();
       (* if s.ENIAMlexSemanticsTypes.schemata = [] then failwith ("create_chart: no schema for token=" ^ t.ENIAMtokenizerTypes.orth ^ " lemma=" ^ ENIAMtokens.get_lemma t.ENIAMtokenizerTypes.token) else *)
-      Xlist.fold s.ENIAMlexSemanticsTypes.schemata chart (fun chart (selectors,cats,snode,schema) ->
-        let cats = clarify_categories cats snode t in
+      Xlist.fold s.ENIAMlexSemanticsTypes.schemata chart (fun chart (selectors,cats,(*snode,*)schema) ->
+        let cats = clarify_categories cats (*snode*) t in
       (* let chart = ENIAM_LCGchart.add_inc_list chart lnode rnode s.ENIAMlexSemanticsTypes.lex_entries 0 in *)
         let l = ENIAM_LCGlexicon.create_entries rules id t.ENIAMtokenizerTypes.orth cats [selectors,schema] s.ENIAMlexSemanticsTypes.lex_entries in
         ENIAM_LCGchart.add_inc_list chart lnode rnode l 0)) in
@@ -115,8 +115,8 @@ let create_dep_chart dep_rules tokens lex_sems paths =
       let s = ExtArray.get lex_sems id in
       ENIAM_LCGrenderer.reset_variable_names ();
       ENIAM_LCGrenderer.add_variable_numbers ();
-      Xlist.fold s.ENIAMlexSemanticsTypes.schemata nodes (fun nodes (selectors,cats,snode,schema) ->
-        let cats = clarify_categories ["X",["X"]] snode t in
+      Xlist.fold s.ENIAMlexSemanticsTypes.schemata nodes (fun nodes (selectors,cats,(*snode,*)schema) ->
+        let cats = clarify_categories ["X",["X"]] (*snode*) t in
       (* let chart = ENIAM_LCGchart.add_inc_list chart lnode rnode s.ENIAMlexSemanticsTypes.lex_entries 0 in *)
         let l = ENIAM_LCGlexicon.create_entries dep_rules id t.ENIAMtokenizerTypes.orth cats [selectors,schema] s.ENIAMlexSemanticsTypes.lex_entries in
         IntMap.add_inc nodes i l (fun l2 -> l @ l2))) in
@@ -797,9 +797,21 @@ let rec extract_pos_cat_internal vars = function
 let rec extract_pos_cat vars = function
   | Tensor [] -> failwith "extract_pos_cat: ni"
   | Tensor [pos] -> extract_pos_cat_internal vars pos
+  | Tensor [pos;_] -> extract_pos_cat_internal vars pos
+  | Tensor [pos;_;_] -> extract_pos_cat_internal vars pos
   | Tensor (Atom "num" :: _) -> "Number"
+  | Tensor (Atom "aglt" :: _) -> "Aglt"
   | Tensor (Atom "prepnp" :: _) -> "Prep"
-  | Tensor (pos :: cat :: _) -> (*extract_pos_cat_internal vars pos ^ "*" ^*) extract_pos_cat_internal vars cat
+  | Tensor (Atom "comparp" :: _) -> "Compar"
+  | Tensor (Atom "cp" :: _) -> "Comp"
+  | Tensor [_;cat;_;_] -> extract_pos_cat_internal vars cat
+  | Tensor [_;_;cat;_;_] -> extract_pos_cat_internal vars cat
+  | Tensor [_;_;_;cat;_;_] -> extract_pos_cat_internal vars cat
+  | Tensor [_;_;_;_;cat;_;_] -> extract_pos_cat_internal vars cat
+  | Tensor [_;_;_;_;_;cat;_;_] -> extract_pos_cat_internal vars cat
+  | Tensor [_;_;_;_;_;_;cat;_;_] -> extract_pos_cat_internal vars cat
+  (* | Tensor (pos :: cat :: _) -> (*extract_pos_cat_internal vars pos ^ "*" ^*) extract_pos_cat_internal vars cat *)
+  | Tensor _ as t -> print_endline ("Unknown symbol " ^ ENIAM_LCGstringOf.grammar_symbol 0 t); "Unknown"
   | Plus l -> failwith "extract_pos_cat: ni"
   | Imp(s,d,t2) -> extract_pos_cat vars s
   | One -> failwith "extract_pos_cat: ni"
@@ -822,7 +834,8 @@ let omited = StringSet.of_list [&quot;&lt;subst&gt;&quot;;&quot;&lt;depr&gt;&quot;;&quot;&lt;ppron12&gt;&quot;;&quot;&lt;ppron3&gt;&quot;;&quot;&lt;sieb
   "<email>";"<obj-id>";"<adj>";"<apron>";"<adjc>";"<adjp>";"<adja>";"<adv>";"<ger>";"<pact>";
   "<ppas>";"<fin>";"<bedzie>";"<praet>";"<winien>";"<impt>";"<imps>";"<pred>";"<aglt>";"<inf>";
   "<pcon>";"<pant>";"<qub>";"<comp>";"<compar>";"<conj>";"<interj>";"<sinterj>";"<burk>";
-  "<interp>";"<part>";"<unk>";"<building-number>";"<html-tag>";"<list-item>";"<numcomp>";"<phone-number>";"<postal-code>"]
+  "<interp>";"<part>";"<unk>";"<building-number>";"<html-tag>";"<list-item>";"<numcomp>";
+  "<phone-number>";"<postal-code>";"<sentence>";"<paragraph>"]
 let cat_tokens_sequence text_fragments g =
   let _,_,l = ENIAM_LCGchart.fold g (0,0,[]) (fun (m,n,l) (symbol,node1,node2,sem,layer) ->
@@ -253,19 +253,19 @@ let simplify_schemata lexemes pos pos2 lemma schemata =
               "{" ^ String.concat ";" (PhraseSet.fold morfs [] (fun l m -> ENIAMwalStringOf.phrase m :: l)) ^ "}")))); *)
       schemata
-let add_adjuncts preps compreps compars pos2 (selectors,cat,has_context,schema) =
+let add_adjuncts preps compreps compars pos2 (selectors,cat,(*has_context,*)schema) =
   let compreps = Xlist.rev_map compreps ENIAMwalRenderer.render_comprep in
   let prepnps = Xlist.rev_map preps (fun (prep,cases) -> ENIAMwalRenderer.render_prepnp prep cases) in
   let prepadjps = Xlist.rev_map preps (fun (prep,cases) -> ENIAMwalRenderer.render_prepadjp prep cases) in
   let compars = Xlist.rev_map compars ENIAMwalRenderer.render_compar in
   match pos2 with
-    "verb" -> [selectors,cat,has_context,schema @ ENIAMwalRenderer.verb_adjuncts_simp @ prepnps @ prepadjps @ compreps @ compars]
+    "verb" -> [selectors,cat,(*has_context,*)schema @ ENIAMwalRenderer.verb_adjuncts_simp @ prepnps @ prepadjps @ compreps @ compars]
   | "noun" -> [
-      [Nsyn,Eq,["proper"]] @ selectors,cat,has_context,ENIAMwalRenderer.proper_noun_adjuncts_simp @ prepnps @ compreps @ compars;
-      [Nsyn,Eq,["common"];Nsem,Eq,["measure"]] @ selectors,cat,has_context,ENIAMwalRenderer.measure_noun_adjuncts_simp @ prepnps @ compreps @ compars;
-      [Nsyn,Eq,["common"];Nsem,Neq,["measure"]] @ selectors,cat,has_context,ENIAMwalRenderer.common_noun_adjuncts_simp @ prepnps @ compreps @ compars]
-  | "adj" -> [selectors,cat,has_context,schema @ ENIAMwalRenderer.adj_adjuncts_simp @ compars]
-  | "adv" -> [selectors,cat,has_context,schema @ ENIAMwalRenderer.adv_adjuncts_simp @ compars]
+      [Nsyn,Eq,["proper"]] @ selectors,cat,(*has_context,*)ENIAMwalRenderer.proper_noun_adjuncts_simp @ prepnps @ compreps @ compars;
+      [Nsyn,Eq,["common"];Nsem,Eq,["measure"]] @ selectors,cat,(*has_context,*)ENIAMwalRenderer.measure_noun_adjuncts_simp @ prepnps @ compreps @ compars;
+      [Nsyn,Eq,["common"];Nsem,Neq,["measure"]] @ selectors,cat,(*has_context,*)ENIAMwalRenderer.common_noun_adjuncts_simp @ prepnps @ compreps @ compars]
+  | "adj" -> [selectors,cat,(*has_context,*)schema @ ENIAMwalRenderer.adj_adjuncts_simp @ compars]
+  | "adv" -> [selectors,cat,(*has_context,*)schema @ ENIAMwalRenderer.adv_adjuncts_simp @ compars]
   | _ -> []
 open ENIAMlexSemanticsTypes
@@ -23,7 +23,7 @@ open ENIAMlexSemanticsTypes
 open ENIAMwalTypes
 open Xstd
-let snode_values = ENIAM_LCGlexiconTypes.SelectorMap.find ENIAMcategoriesPL.selector_values ENIAM_LCGlexiconTypes.SNode
+(*let snode_values = ENIAM_LCGlexiconTypes.SelectorMap.find ENIAMcategoriesPL.selector_values ENIAM_LCGlexiconTypes.SNode*)
 let find_sense m =
   try
@@ -305,9 +305,9 @@ let assign_valence tokens lex_sems group =
       let schemata = ENIAMadjuncts.simplify_schemata lexemes pos pos2 lemma schemata1 in
       (* Printf.printf "C %s |schemata|=%d\n" lemma (Xlist.size schemata); *)
       let schemata = Xlist.rev_map schemata (fun (selectors,schema) ->
-          selectors,["X",["X"]],snode_values,ENIAMwalRenderer.render_simple_schema schema) in
+          selectors,["X",["X"]],(*snode_values,*)ENIAMwalRenderer.render_simple_schema schema) in
       let schemata = List.flatten (Xlist.rev_map schemata (ENIAMadjuncts.add_adjuncts preps compreps compars pos2)) in
-      let schemata = if schemata = [] then [[],["X",["X"]],snode_values,[]] else schemata in
+      let schemata = if schemata = [] then [[],["X",["X"]],(*snode_values,*)[]] else schemata in
       (* Printf.printf "D %s |schemata|=%d\n" lemma (Xlist.size schemata); *)
       let entries = List.flatten (Xlist.rev_map entries (ENIAMvalence.transform_lex_entry pos lemma)) in
       let entries = Xlist.map entries (fun (selectors,entry) ->
@@ -60,10 +60,10 @@ let html_of_lex_sems tokens lex_sems =
     let core = Printf.sprintf "%3d %s %s" id orth lemma  in
     let lex_entries = Xlist.map t.lex_entries (fun (selectors,s) ->
         "&emsp;&emsp;[" ^ ENIAMcategoriesPL.string_of_selectors selectors ^ "] " ^ ENIAM_LCGstringOf.grammar_symbol 0 s) in
-    let schemata = Xlist.map t.schemata (fun (selectors,cat,snode,l) ->
+    let schemata = Xlist.map t.schemata (fun (selectors,cat,(*snode,*)l) ->
         "&emsp;&emsp;[" ^ ENIAMcategoriesPL.string_of_selectors selectors ^ "]" ^
         String.concat "," (Xlist.map cat (fun (m,l) -> m ^ "[" ^ String.concat "," l ^ "]")) ^
-        String.concat "|" snode ^
+        (*String.concat "|" snode ^*)
         " {" ^ String.concat ", " (Xlist.map l (fun (d,s) ->
             ENIAM_LCGstringOf.direction d ^ ENIAM_LCGstringOf.grammar_symbol 0 s)) ^ "}") in
     (* let frames = Xlist.map t.frames (fun (selectors,senses,schema) -> FIXME
@@ -40,10 +40,10 @@ let string_of_lex_sems tokens lex_sems =
     let core = Printf.sprintf "%3d %s %s" id orth lemma  in
     let lex_entries = Xlist.map t.lex_entries (fun (selectors,s) ->
         "&[" ^ ENIAMcategoriesPL.string_of_selectors selectors ^ "] " ^ ENIAM_LCGstringOf.grammar_symbol 0 s) in
-    let schemata = Xlist.map t.schemata (fun (selectors,cat,snode,l) ->
+    let schemata = Xlist.map t.schemata (fun (selectors,cat,(*snode,*)l) ->
         "[" ^ ENIAMcategoriesPL.string_of_selectors selectors ^ "]" ^
         String.concat "," (Xlist.map cat (fun (m,l) -> m ^ "[" ^ String.concat "," l ^ "]")) ^
-        String.concat "|" snode ^
+        (*String.concat "|" snode ^*)
         " {" ^ String.concat "," (Xlist.map l (fun (d,s) ->
             ENIAM_LCGstringOf.direction d ^ ENIAM_LCGstringOf.grammar_symbol 0 s)) ^ "}") in
     let frames = Xlist.map t.frames (fun f ->
@@ -31,18 +31,18 @@ type frame = {
   agf: string;
   sem_args: string list;
   rev_hipero: bool;
-  snode: string list;
+  (*snode: string list;*)
   sopinion: ENIAMwalTypes.opinion;
   fopinion: ENIAMwalTypes.opinion;
   }
-let empty_frame = {selectors=[]; senses=[]; cats=["X",["X"]]; positions=[]; arole=""; arole_attr=""; arev=false; agf=""; sem_args=[]; rev_hipero=false; snode=[];
+let empty_frame = {selectors=[]; senses=[]; cats=["X",["X"]]; positions=[]; arole=""; arole_attr=""; arev=false; agf=""; sem_args=[]; rev_hipero=false; (*snode=[];*)
   sopinion=ENIAMwalTypes.Nieokreslony; fopinion=ENIAMwalTypes.Nieokreslony}
 type lex_sem = {
   schemata: ((ENIAM_LCGlexiconTypes.selector * ENIAM_LCGlexiconTypes.selector_relation * string list) list *
              (string * string list) list * (* sensy *)
-             string list * (* has_context *)
+             (*string list **) (* has_context *)
              (ENIAM_LCGtypes.direction * ENIAM_LCGtypes.grammar_symbol) list) list;
   lex_entries: ((ENIAM_LCGlexiconTypes.selector * ENIAM_LCGlexiconTypes.selector_relation * string list) list *
                 ENIAM_LCGtypes.grammar_symbol) list;
@@ -465,7 +465,8 @@ let transform_ger_schema lemma negation schema = (* FIXME: zakładam, że ger ze
                     | phrase -> transform_pers_phrase lemma negation "gerundial" phrase))})
 let transform_ppas_schema lemma negation mood schema =
-  if not (Xlist.fold schema false (fun b p -> if p.gf = OBJ then true else b)) then raise Not_found else
+  if not (Xlist.fold schema false (fun b p -> if p.gf = OBJ then true else b)) then
+    (*failwith ("transform_ppas_schema: attempt to make ppas schema for lemma " ^ lemma ^ "without OBJ arg")*)raise Not_found else
   Xlist.map schema (fun s ->
         let morfs = List.flatten (Xlist.map s.morfs (transform_comps negation mood)) in
         let morfs = List.flatten (Xlist.map morfs transform_preps) in
@@ -53,7 +53,7 @@ let render_pos_entry = function
   | "ppas" ->  [Atom "ppas"; AVar "number"; AVar "case"; AVar "gender"; AVar "negation"]
   | "inf" ->  [Atom "inf"; AVar "aspect"; AVar "negation"]
   | "qub" ->  [Atom "qub"]
-  | "compar" ->  [Atom "compar"; AVar "case"]
+  | "compar" ->  [Atom "comparp"; AVar "case"]
   | "comp" ->  [Atom "comp"; AVar "ctype"]
   | "fin" ->  [Atom "pers"; AVar "negation"]
   | "praet" ->  [Atom "pers"; AVar "negation"]
@@ -130,7 +130,7 @@ let render_phrase = function
     | PrepNumP(_,prep,Case case) -> Tensor[Atom "prepnump"; Atom prep; Atom case] *)
 (*      | ComprepNP("") -> Tensor[Atom "comprepnp"; Top]*)
     | ComprepNP(prep) -> Tensor[Atom "comprepnp"; Atom prep]
-    | ComparP(prep,Case case) -> Tensor[Atom "compar"; Atom prep; Atom case]
+    | ComparP(prep,Case case) -> Tensor[Atom "comparp"; Atom prep; Atom case]
     (* | ComparPP(_,prep) -> Tensor[Atom "comparpp"; Atom prep] *)
     (* | IP -> Tensor[Atom "ip";Top;Top;Top] *)
     | CP (ctype,Comp comp) -> Tensor[Atom "cp"; arg_of_ctype ctype; Atom comp]
@@ -209,7 +209,7 @@ let render_phrase_cat cat role node = function
     | PrepNumP(_,prep,Case case) -> Tensor[Atom "prepnump"; Atom prep; Atom case] *)
 (*      | ComprepNP("") -> Tensor[Atom "comprepnp"; Top; Atom cat; Atom role; Atom node]*)
     | ComprepNP(prep) -> Tensor[Atom "comprepnp"; Atom prep; Atom cat; Atom role; Atom node]
-    | ComparP(prep,Case case) -> Tensor[Atom "compar"; Atom prep; Atom case; Atom cat; Atom role; Atom node]
+    | ComparP(prep,Case case) -> Tensor[Atom "comparp"; Atom prep; Atom case; Atom cat; Atom role; Atom node]
     (* | ComparPP(_,prep) -> Tensor[Atom "comparpp"; Atom prep; Atom cat; Atom role; Atom node] *)
     (* | IP -> Tensor[Atom "ip";Top;Top;Top; Atom cat; Atom role; Atom node] *)
     | CP (ctype,Comp comp) -> Tensor[Atom "cp"; arg_of_ctype ctype; Atom comp; Atom cat; Atom role; Atom node]
@@ -367,9 +367,9 @@ let render_connected_prepadjp prep cases =
   adjunct (postp @ (Xlist.map cases (fun case ->
       Tensor[Atom "prepadjp"; Atom prep; Atom case])))
-let render_compar prep = Both,Plus[One;Tensor[Atom "compar"; Atom prep; Top]]
+let render_compar prep = Both,Plus[One;Tensor[Atom "comparp"; Atom prep; Top]]
-let render_connected_compar prep = adjunct [Tensor[Atom "compar"; Atom prep; Top]]
+let render_connected_compar prep = adjunct [Tensor[Atom "comparp"; Atom prep; Top]]
 let verb_adjuncts_simp = [
   Both, Plus[One;Tensor[Atom "advp"; Atom "pron"]];