Commit 2667fa91bfa4c7761e59650b82eceea9f2eb209c
1 parent
8f0e1464
rozpoczęcie poprawiania interfejsu do Słowosieci w lexSemantics
Showing
3 changed files
with
67 additions
and
7 deletions
lexSemantics/ENIAMlexSemanticsTypes.ml
| ... | ... | @@ -77,3 +77,15 @@ let subst_uncountable_lexemes_filename2 = resource_path ^ "/Walenty/subst_uncoun |
| 77 | 77 | let subst_container_lexemes_filename = resource_path ^ "/Walenty/subst_container.dat" |
| 78 | 78 | let subst_numeral_lexemes_filename = resource_path ^ "/Walenty/subst_numeral.dat" |
| 79 | 79 | let subst_time_lexemes_filename = resource_path ^ "/Walenty/subst_time.dat" *) |
| 80 | + | |
| 81 | +(*let rzeczownik_filename = resource_path ^ "/plWordnet/rzeczownik.tab" | |
| 82 | +let czasownik_filename = resource_path ^ "/plWordnet/czasownik.tab" | |
| 83 | +let przymiotnik_filename = resource_path ^ "/plWordnet/przymiotnik.tab" | |
| 84 | +let synsets_filename = resource_path ^ "/plWordnet/synsets.tab" | |
| 85 | +let hipero_filename = resource_path ^ "/plWordnet/hipero.tab" | |
| 86 | +let predef_filename = resource_path ^ "/plWordnet/predef_prefs.tab" | |
| 87 | +let proper_classes_filename = resource_path ^ "/plWordnet/proper_classes.tab"*) | |
| 88 | + | |
| 89 | +let lu_filename = resource_path ^ "/plWordnet/lu.tab" | |
| 90 | +let ex_hipo_filename = resource_path ^ "/plWordnet/ex_hipo.tab" | |
| 91 | +let syn_filename = resource_path ^ "/plWordnet/syn.tab" | |
| ... | ... |
lexSemantics/ENIAMplWordnet.ml
0 → 100644
| 1 | +(* | |
| 2 | + * ENIAMlexSemantics is a library that assigns tokens with lexicosemantic information. | |
| 3 | + * Copyright (C) 2016-2017 Wojciech Jaworski <wjaworski atSPAMfree mimuw dot edu dot pl> | |
| 4 | + * Copyright (C) 2016-2017 Institute of Computer Science Polish Academy of Sciences | |
| 5 | + * | |
| 6 | + * This library is free software: you can redistribute it and/or modify | |
| 7 | + * it under the terms of the GNU Lesser General Public License as published by | |
| 8 | + * the Free Software Foundation, either version 3 of the License, or | |
| 9 | + * (at your option) any later version. | |
| 10 | + * | |
| 11 | + * This library is distributed in the hope that it will be useful, | |
| 12 | + * but WITHOUT ANY WARRANTY; without even the implied warranty of | |
| 13 | + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
| 14 | + * GNU Lesser General Public License for more details. | |
| 15 | + * | |
| 16 | + * You should have received a copy of the GNU Lesser General Public License | |
| 17 | + * along with this program. If not, see <http://www.gnu.org/licenses/>. | |
| 18 | + *) | |
| 19 | + | |
| 20 | +open Xstd | |
| 21 | +open ENIAMlexSemanticsTypes | |
| 22 | + | |
| 23 | +let load_lu filename = | |
| 24 | + File.fold_tab filename IntMap.empty (fun lumap -> function | |
| 25 | + [lu_id; lemma; variant; syn_id] -> IntMap.add lumap (int_of_string lu_id) (lemma,variant,int_of_string syn_id) | |
| 26 | + | l -> failwith ("load_lu: " ^ String.concat "\t" l)) | |
| 27 | + | |
| 28 | +let load_syn filename = | |
| 29 | + File.fold_tab filename IntMap.empty (fun synmap -> function | |
| 30 | + syn_id :: pos :: lu_ids -> | |
| 31 | + let lu_ids = Xlist.map lu_ids int_of_string in | |
| 32 | + IntMap.add synmap (int_of_string syn_id) (pos,lu_ids) | |
| 33 | + | l -> failwith ("load_syn: " ^ String.concat "\t" l)) | |
| 34 | + | |
| 35 | +let load_ex_hipo filename = | |
| 36 | + File.fold_tab filename IntMap.empty (fun ex_hipo -> function | |
| 37 | + [parent; child; cost] -> | |
| 38 | + let parent = int_of_string parent in | |
| 39 | + let child = int_of_string child in | |
| 40 | + let cost = int_of_string cost in | |
| 41 | + let children = try IntMap.find ex_hipo parent with Not_found -> IntMap.empty in | |
| 42 | + let children = IntMap.add_inc children child cost (fun _ -> failwith "load_ex_hipo") in | |
| 43 | + IntMap.add ex_hipo parent children | |
| 44 | + | l -> failwith ("load_ex_hipo: " ^ String.concat "\t" l)) | |
| 45 | + | |
| 46 | +let lumap = load_lu lu_filename | |
| 47 | +let synmap = load_syn syn_filename | |
| 48 | +let ex_hipo = load_ex_hipo ex_hipo_filename | |
| ... | ... |
lexSemantics/makefile
| ... | ... | @@ -3,20 +3,20 @@ OCAMLOPT=ocamlopt |
| 3 | 3 | OCAMLDEP=ocamldep |
| 4 | 4 | INCLUDES=-I +xml-light -I +xlib -I +zip -I +bz2 -I +eniam |
| 5 | 5 | OCAMLFLAGS=$(INCLUDES) -g |
| 6 | -#OCAMLOPTFLAGS=$(INCLUDES) unix.cmxa xml-light.cmxa str.cmxa nums.cmxa zip.cmxa bz2.cmxa xlib.cmxa eniam-tokenizer.cmxa eniam-morphology.cmxa eniam-subsyntax.cmxa eniam-lcg-parser.cmxa eniam-lcg-lexicon.cmxa #eniam-plWordnet.cmxa #eniam-lexSemantics.cmxa | |
| 7 | -OCAMLOPTFLAGS=$(INCLUDES) unix.cmxa xml-light.cmxa str.cmxa nums.cmxa zip.cmxa bz2.cmxa xlib.cmxa eniam-tokenizer.cmxa eniam-morphology.cmxa eniam-subsyntax.cmxa eniam-lcg-parser.cmxa eniam-lcg-lexicon.cmxa eniam-lexSemantics.cmxa | |
| 6 | +OCAMLOPTFLAGS=$(INCLUDES) unix.cmxa xml-light.cmxa str.cmxa nums.cmxa zip.cmxa bz2.cmxa xlib.cmxa eniam-tokenizer.cmxa eniam-morphology.cmxa eniam-subsyntax.cmxa eniam-lcg-parser.cmxa eniam-lcg-lexicon.cmxa #eniam-plWordnet.cmxa #eniam-lexSemantics.cmxa | |
| 7 | +#OCAMLOPTFLAGS=$(INCLUDES) unix.cmxa xml-light.cmxa str.cmxa nums.cmxa zip.cmxa bz2.cmxa xlib.cmxa eniam-tokenizer.cmxa eniam-morphology.cmxa eniam-subsyntax.cmxa eniam-lcg-parser.cmxa eniam-lcg-lexicon.cmxa eniam-lexSemantics.cmxa | |
| 8 | 8 | INSTALLDIR=`ocamlc -where`/eniam |
| 9 | 9 | |
| 10 | 10 | SOURCES= entries.ml ENIAMwalTypes.ml ENIAMwalStringOf.ml ENIAMwalParser.ml ENIAMwalReduce.ml ENIAMvalence.ml ENIAMwalRenderer.ml ENIAMadjuncts.ml \ |
| 11 | - ENIAMlexSemanticsTypes.ml ENIAMlexSemanticsStringOf.ml ENIAMlexSemanticsHTMLof.ml ENIAMlexSemanticsXMLof.ml ENIAMlexSemantics.ml #ENIAMlexSemanticsData.ml | |
| 11 | + ENIAMlexSemanticsTypes.ml ENIAMlexSemanticsStringOf.ml ENIAMlexSemanticsHTMLof.ml ENIAMlexSemanticsXMLof.ml ENIAMplWordnet.ml ENIAMlexSemantics.ml #ENIAMlexSemanticsData.ml | |
| 12 | 12 | |
| 13 | 13 | all: eniam-lexSemantics.cma eniam-lexSemantics.cmxa |
| 14 | 14 | |
| 15 | 15 | install: all |
| 16 | 16 | mkdir -p $(INSTALLDIR) |
| 17 | 17 | cp eniam-lexSemantics.cmxa eniam-lexSemantics.a eniam-lexSemantics.cma $(INSTALLDIR) |
| 18 | - cp entries.cmi ENIAMwalTypes.cmi ENIAMwalStringOf.cmi ENIAMwalParser.cmi ENIAMwalReduce.cmi ENIAMvalence.cmi ENIAMwalRenderer.cmi ENIAMadjuncts.cmi ENIAMlexSemanticsTypes.cmi ENIAMlexSemanticsStringOf.cmi ENIAMlexSemanticsHTMLof.cmi ENIAMlexSemanticsXMLof.cmi ENIAMlexSemantics.cmi $(INSTALLDIR) | |
| 19 | - cp entries.cmx ENIAMwalTypes.cmx ENIAMwalStringOf.cmx ENIAMwalParser.cmx ENIAMwalReduce.cmx ENIAMvalence.cmx ENIAMwalRenderer.cmx ENIAMadjuncts.cmx ENIAMlexSemanticsTypes.cmx ENIAMlexSemanticsStringOf.cmx ENIAMlexSemanticsHTMLof.cmx ENIAMlexSemanticsXMLof.cmx ENIAMlexSemantics.cmx $(INSTALLDIR) | |
| 18 | + cp entries.cmi ENIAMwalTypes.cmi ENIAMwalStringOf.cmi ENIAMwalParser.cmi ENIAMwalReduce.cmi ENIAMvalence.cmi ENIAMwalRenderer.cmi ENIAMadjuncts.cmi ENIAMlexSemanticsTypes.cmi ENIAMlexSemanticsStringOf.cmi ENIAMlexSemanticsHTMLof.cmi ENIAMlexSemanticsXMLof.cmi ENIAMplWordnet.cmi ENIAMlexSemantics.cmi $(INSTALLDIR) | |
| 19 | + cp entries.cmx ENIAMwalTypes.cmx ENIAMwalStringOf.cmx ENIAMwalParser.cmx ENIAMwalReduce.cmx ENIAMvalence.cmx ENIAMwalRenderer.cmx ENIAMadjuncts.cmx ENIAMlexSemanticsTypes.cmx ENIAMlexSemanticsStringOf.cmx ENIAMlexSemanticsHTMLof.cmx ENIAMlexSemanticsXMLof.cmx ENIAMplWordnet.cmx ENIAMlexSemantics.cmx $(INSTALLDIR) | |
| 20 | 20 | mkdir -p /usr/share/eniam/lexSemantics |
| 21 | 21 | # cp resources/* /usr/share/eniam/lexSemantics |
| 22 | 22 | ln -s /usr/share/eniam/lexSemantics/proper_names_20160104.tab /usr/share/eniam/lexSemantics/proper_names.tab |
| ... | ... | @@ -25,8 +25,8 @@ install: all |
| 25 | 25 | install-local: all |
| 26 | 26 | mkdir -p $(INSTALLDIR) |
| 27 | 27 | cp eniam-lexSemantics.cmxa eniam-lexSemantics.a eniam-lexSemantics.cma $(INSTALLDIR) |
| 28 | - cp entries.cmi ENIAMwalTypes.cmi ENIAMwalStringOf.cmi ENIAMwalParser.cmi ENIAMwalReduce.cmi ENIAMvalence.cmi ENIAMwalRenderer.cmi ENIAMadjuncts.cmi ENIAMlexSemanticsTypes.cmi ENIAMlexSemantics.cmi $(INSTALLDIR) | |
| 29 | - cp entries.cmx ENIAMwalTypes.cmx ENIAMwalStringOf.cmx ENIAMwalParser.cmx ENIAMwalReduce.cmx ENIAMvalence.cmx ENIAMwalRenderer.cmx ENIAMadjuncts.cmx ENIAMlexSemanticsTypes.cmx ENIAMlexSemantics.cmx $(INSTALLDIR) | |
| 28 | + cp entries.cmi ENIAMwalTypes.cmi ENIAMwalStringOf.cmi ENIAMwalParser.cmi ENIAMwalReduce.cmi ENIAMvalence.cmi ENIAMwalRenderer.cmi ENIAMadjuncts.cmi ENIAMlexSemanticsTypes.cmi ENIAMlexSemanticsStringOf.cmi ENIAMlexSemanticsHTMLof.cmi ENIAMlexSemanticsXMLof.cmi ENIAMplWordnet.cmi ENIAMlexSemantics.cmi $(INSTALLDIR) | |
| 29 | + cp entries.cmx ENIAMwalTypes.cmx ENIAMwalStringOf.cmx ENIAMwalParser.cmx ENIAMwalReduce.cmx ENIAMvalence.cmx ENIAMwalRenderer.cmx ENIAMadjuncts.cmx ENIAMlexSemanticsTypes.cmx ENIAMlexSemanticsStringOf.cmx ENIAMlexSemanticsHTMLof.cmx ENIAMlexSemanticsXMLof.cmx ENIAMplWordnet.cmx ENIAMlexSemantics.cmx $(INSTALLDIR) | |
| 30 | 30 | mkdir -p /usr/local/share/eniam/lexSemantics |
| 31 | 31 | # cp resources/* /usr/local/share/eniam/lexSemantics |
| 32 | 32 | ln -s /usr/local/share/eniam/lexSemantics/proper_names_20160104.tab /usr/local/share/eniam/lexSemantics/proper_names.tab |
| ... | ... |