From df1486af60e99aceb686f5ccc8894d6dae28df9f Mon Sep 17 00:00:00 2001
From: Wojciech Jaworski <wjaworski@mimuw.edu.pl>
Date: Sat, 25 Mar 2017 12:09:51 +0100
Subject: [PATCH] poprawki w interfejsie subsyntax
---
LCGlexicon/interface.ml | 205 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
LCGlexicon/makefile | 5 ++++-
NKJP2/ENIAM_NKJP.ml | 66 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++----
subsyntax/ENIAM_MWE.ml | 2 +-
subsyntax/ENIAMsubsyntax.ml | 3 ++-
subsyntax/ENIAMsubsyntaxTypes.ml | 1 +
subsyntax/resources/ne.tab | 34 ++++++++++++++++++++++++++++++++++
tokenizer/ENIAMacronyms.ml | 2 +-
tokenizer/ENIAMtokenizerTypes.ml | 3 ++-
tokenizer/makefile | 4 ++--
10 files changed, 314 insertions(+), 11 deletions(-)
create mode 100644 LCGlexicon/interface.ml
create mode 100644 subsyntax/resources/ne.tab
diff --git a/LCGlexicon/interface.ml b/LCGlexicon/interface.ml
new file mode 100644
index 0000000..63e8e6e
--- /dev/null
+++ b/LCGlexicon/interface.ml
@@ -0,0 +1,205 @@
+(*
+ * ENIAM_LCGlexicon is a library that provides LCG lexicon form Polish
+ * Copyright (C) 2016 Wojciech Jaworski <wjaworski atSPAMfree mimuw dot edu dot pl>
+ * Copyright (C) 2016 Institute of Computer Science Polish Academy of Sciences
+ *
+ * This library is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Lesser General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ *)
+
+open ENIAM_LCGlexiconTypes
+open ENIAM_LCGtypes
+open ENIAMsubsyntaxTypes
+
+let rules = ENIAM_LCGlexicon.make_rules ENIAM_LCGlexiconTypes.rules_filename
+
+let examples = [
+ (* "Szpak","Szpak śpiewa.";*)
+ (* "miał","Miałem miał."; *)
+(* "Ala","Ala ma kota.";
+ "Ale","Ale mają kota:"; *)
+ (* "zima","Szpak frunie zimą.";*)
+ (* "październik","Kot miauczy w październiku."; *)
+(* "Szpak-Kot","Szpak frunie. Kot miauczy.";
+ "powiedział","Szpak powiedział: „Frunę. Kiszę.”";*)
+ (* "teraz","Teraz frunie jakiś szpak.";
+ "chłopcy","Chłopcy mają ulicę kwiatami."; *)
+ (* "arabia","Arabia Saudyjska biegnie.";*)
+(* "Tom","Tom idzie."; *)
+ "liceum","W 1984-89 uczęszczał do VII Liceum Ogólnokształcącego im. K.K. Baczyńskiego w Szczecinie.";
+ "studia","Następnie studiował architekturę na Politechnice Szczecińskiej, dyplom uzyskał w 1994.";
+]
+
+let clarify_categories senses token =
+ match token.ENIAMtokenizerTypes.token with
+ ENIAMtokenizerTypes.Lemma(lemma,pos,interp) -> List.flatten (Xlist.map interp (fun interp -> ENIAMcategoriesPL.clarify_categories false senses (lemma,pos,interp)))
+ | ENIAMtokenizerTypes.Proper(lemma,pos,interp,_) -> List.flatten (Xlist.map interp (fun interp -> ENIAMcategoriesPL.clarify_categories true senses (lemma,pos,interp)))
+ | ENIAMtokenizerTypes.Interp lemma -> ENIAMcategoriesPL.clarify_categories false senses (lemma,"interp",[])
+ | _ -> []
+
+let create_chart tokens lex_sems paths last =
+ ENIAM_LCGrenderer.reset_variable_numbers ();
+ let chart = ENIAM_LCGchart.make last in
+ let chart = Xlist.fold paths chart (fun chart (id,lnode,rnode) ->
+ let t = ExtArray.get tokens id in
+ let s = ExtArray.get lex_sems id in
+ ENIAM_LCGrenderer.reset_variable_names ();
+ ENIAM_LCGrenderer.add_variable_numbers ();
+ let cats = clarify_categories ["X"] t in
+ let l = ENIAM_LCGlexicon.create_entries rules id t.ENIAMtokenizerTypes.orth cats s.ENIAMlexSemanticsTypes.schemata in
+ ENIAM_LCGchart.add_inc_list chart lnode rnode l 0) in
+ chart
+
+let test_example name tokens lex_sems paths last =
+ ENIAM_LCGreductions.reset_variant_label ();
+ let chart = create_chart tokens lex_sems paths last in
+ ENIAM_LCGlatexOf.print_chart "results/" (name^"1_chart") "a1" chart;
+ let chart,references = ENIAM_LCGchart.lazify chart in
+ ENIAM_LCGlatexOf.print_chart "results/" (name^"2_chart") "a4" chart;
+ ENIAM_LCGlatexOf.print_references "results/" (name^"2_references") "a4" references;
+ let chart = ENIAM_LCGchart.parse chart references 30. Sys.time in (* uwaga: niejawna zmiana imperatywna w references *)
+ ENIAM_LCGlatexOf.print_chart "results/" (name^"3_chart") "a4" chart;
+ ENIAM_LCGlatexOf.print_references "results/" (name^"3_references") "a4" references;
+ if ENIAM_LCGchart.is_parsed chart then (
+ let term = ENIAM_LCGchart.get_parsed_term chart in
+ Xlatex.latex_file_out "results/" (name^"4_term") "a4" false (fun file ->
+ Printf.fprintf file "\\[%s\\]\n" (ENIAM_LCGlatexOf.linear_term 0 term));
+ Xlatex.latex_compile_and_clean "results/" (name^"4_term");
+ let dependency_tree = ENIAM_LCGreductions.reduce term references in
+ ENIAM_LCGlatexOf.print_dependency_tree "results/" (name^"4_dependency_tree") "a0" dependency_tree;
+ if ENIAM_LCGreductions.is_reduced_dependency_tree dependency_tree then (
+ ENIAM_LCGreductions.assign_labels dependency_tree; (* uwaga: niejawna zmiana imperatywna w dependency_tree *)
+ ENIAM_LCGlatexOf.print_dependency_tree "results/" (name^"5_dependency_tree") "a4" dependency_tree;
+ ENIAM_LCGreductions.remove_cuts dependency_tree; (* uwaga: niejawna zmiana imperatywna w dependency_tree *)
+ ENIAM_LCGlatexOf.print_dependency_tree "results/" (name^"6_dependency_tree") "a4" dependency_tree;
+ ENIAM_LCGgraphOf.print_dependency_tree "results/" (name^"6_dependency_tree") dependency_tree;
+ ENIAM_LCGgraphOf.print_simplified_dependency_tree "results/" (name^"6_simple_dependency_tree") dependency_tree;
+ ())
+ else print_endline "not reduced")
+ else print_endline "not parsed"
+
+let rec parse_sentence name id tokens lex_sems = function
+ RawSentence s -> id
+ | StructSentence(paths,last) ->
+ test_example (name ^ string_of_int id ^ "_") tokens lex_sems paths last;
+ id + 1
+ | DepSentence(paths) -> id
+ | QuotedSentences sentences ->
+ Xlist.fold sentences id (fun id p ->
+ parse_sentence name id tokens lex_sems p.sentence)
+ | AltSentence l ->
+ Xlist.fold l id (fun id (mode,sentence) ->
+ parse_sentence name id tokens lex_sems sentence)
+
+let rec parse_paragraph name id tokens lex_sems = function
+ RawParagraph s -> id
+ | StructParagraph sentences ->
+ Xlist.fold sentences id (fun id p ->
+ parse_sentence name id tokens lex_sems p.sentence)
+ | AltParagraph l ->
+ Xlist.fold l id (fun id (mode,paragraph) ->
+ parse_paragraph name id tokens lex_sems paragraph)
+
+let rec parse_text name id tokens lex_sems = function
+ RawText s -> id
+ | StructText paragraphs ->
+ Xlist.fold paragraphs id (fun id paragraph ->
+ parse_paragraph name id tokens lex_sems paragraph)
+ | AltText l ->
+ Xlist.fold l id (fun id (mode,text) ->
+ parse_text name id tokens lex_sems text)
+
+
+let _ =
+ Xlist.iter examples (fun (name,example) ->
+ let text,tokens = ENIAMsubsyntax.parse_text example in
+ let lex_sems = ENIAMlexSemantics.assign tokens text in
+ ignore(parse_text name 1 tokens lex_sems text))
+
+(*
+type output = Text | Xml | Html | Marsh | Graphviz
+
+let output = ref Text
+let comm_stdio = ref true
+let sentence_split = ref true
+let port = ref 0
+
+let spec_list = [
+ "-s", Arg.Unit (fun () -> sentence_split:=true), "Split input into sentences (default)";
+ "-n", Arg.Unit (fun () -> sentence_split:=false), "Do not split input into sentences";
+ "-i", Arg.Unit (fun () -> comm_stdio:=true), "Communication using stdio (default)";
+ "-p", Arg.Int (fun p -> comm_stdio:=false; port:=p), "<port> Communication using sockets on given port number";
+ "-t", Arg.Unit (fun () -> output:=Text), "Output as plain text (default)";
+ "-x", Arg.Unit (fun () -> output:=Xml), "Output as XML";
+ "-m", Arg.Unit (fun () -> output:=Marsh), "Output as marshalled Ocaml data structure";
+ "-h", Arg.Unit (fun () -> output:=Html), "Output as HTML";
+ "-g", Arg.Unit (fun () -> output:=Graphviz; sentence_split:=false), "Output as graphviz dot file; turns sentence split off";
+ (* "-r", Arg.String (fun p ->
+ ENIAMtokenizerTypes.set_resource_path p;
+ ENIAMmorphologyTypes.set_resource_path p;
+ ENIAMsubsyntaxTypes.set_resource_path p), "<path> Set resource path"; *)
+ ]
+
+let usage_msg =
+ "Usage: subsyntax <options>\nInput is a sequence of lines. Empty line ends the sequence and invoke parsing. Double empty line shutdown parser.\nOptions are:"
+
+let message = "ENIAMsubsyntax: MWE, abbreviation and sentence detecion for Polish\n\
+Copyright (C) 2016 Wojciech Jaworski <wjaworski atSPAMfree mimuw dot edu dot pl>\n\
+Copyright (C) 2016 Institute of Computer Science Polish Academy of Sciences"
+
+let anon_fun s = raise (Arg.Bad ("invalid argument: " ^ s))
+
+let input_text channel =
+ let s = ref (try input_line channel with End_of_file -> "") in
+ let lines = ref [] in
+ while !s <> "" do
+ lines := !s :: !lines;
+ s := try input_line channel with End_of_file -> ""
+ done;
+ String.concat "\n" (List.rev !lines)
+
+let rec main_loop in_chan out_chan =
+ let text = input_text in_chan in
+ if text = "" then () else (
+ (* print_endline "input text begin";
+ print_endline text;
+ print_endline "input text end"; *)
+ (if !sentence_split then
+ let text,tokens = ENIAMsubsyntax.parse_text text in
+ (match !output with
+ Text -> output_string out_chan (ENIAMsubsyntaxStringOf.text "" tokens text ^ "\n" ^ ENIAMsubsyntaxStringOf.token_extarray tokens ^ "\n\n")
+ | Xml -> output_string out_chan (Xml.to_string (ENIAMsubsyntaxXMLof.text_and_tokens text tokens) ^ "\n\n")
+ | Html -> output_string out_chan (ENIAMsubsyntaxHTMLof.text_and_tokens text tokens ^ "\n\n")
+ | Marsh -> Marshal.to_channel out_chan (text,tokens) []
+ | Graphviz -> failwith "main_loop: ni")
+ else
+ let tokens = ENIAMsubsyntax.parse text in
+ (match !output with
+ Text -> output_string out_chan (ENIAMsubsyntaxStringOf.token_list tokens ^ "\n\n")
+ | Xml -> output_string out_chan (Xml.to_string (ENIAMsubsyntaxXMLof.token_list tokens) ^ "\n\n")
+ | Html -> output_string out_chan (ENIAMsubsyntaxHTMLof.token_list tokens ^ "\n\n")
+ | Marsh -> Marshal.to_channel out_chan tokens []
+ | Graphviz -> output_string out_chan (ENIAMsubsyntaxGraphOf.token_list tokens ^ "\n\n")));
+ flush out_chan;
+ main_loop in_chan out_chan)
+
+let _ =
+ prerr_endline message;
+ Arg.parse spec_list anon_fun usage_msg;
+ Gc.compact ();
+ prerr_endline "Ready!";
+ if !comm_stdio then main_loop stdin stdout
+ else
+ let sockaddr = Unix.ADDR_INET(Unix.inet_addr_any,!port) in
+ Unix.establish_server main_loop sockaddr
+*)
diff --git a/LCGlexicon/makefile b/LCGlexicon/makefile
index b4bb33a..37e64ec 100755
--- a/LCGlexicon/makefile
+++ b/LCGlexicon/makefile
@@ -42,6 +42,9 @@ test2: test2.ml
mkdir -p results
$(OCAMLOPT) -o test2 $(OCAMLOPTFLAGS2) test2.ml
+interface: interface.ml
+ $(OCAMLOPT) -o parser $(OCAMLOPTFLAGS2) interface.ml
+
print_lexicon: ENIAM_LCGlexiconLatexOf.ml
mkdir -p results
$(OCAMLOPT) -o print_lexicon $(OCAMLOPTFLAGS) ENIAM_LCGlexiconLatexOf.ml
@@ -67,4 +70,4 @@ print_lexicon: ENIAM_LCGlexiconLatexOf.ml
$(OCAMLOPT) $(OCAMLOPTFLAGS) -c $<
clean:
- rm -f *~ *.cm[aoix] *.o *.so *.cmxa *.a test test2 print_lexicon
+ rm -f *~ *.cm[aoix] *.o *.so *.cmxa *.a test test2 parser print_lexicon
diff --git a/NKJP2/ENIAM_NKJP.ml b/NKJP2/ENIAM_NKJP.ml
index 5390230..e58d53c 100644
--- a/NKJP2/ENIAM_NKJP.ml
+++ b/NKJP2/ENIAM_NKJP.ml
@@ -158,14 +158,72 @@ let load_morphosyntax path name =
List.rev (Xlist.rev_map entries load_morph_entry)
| _ -> failwith "load_morphosyntax"
-let rec merge_entries rev = function
+let parse_seg_corresp corresp =
+ if not (Xstring.check_prefix "text.xml#string-range(" corresp) then failwith "parse_seg_corresp" else
+ if not (Xstring.check_sufix ")" corresp) then failwith "parse_seg_corresp" else
+ let corresp = Xstring.cut_prefix "text.xml#string-range(" corresp in
+ let corresp = Xstring.cut_sufix ")" corresp in
+ let id,beg,len = match Xstring.split "," corresp with
+ [id;beg;len] -> parse_id id, int_of_string beg, int_of_string len
+ | _ -> failwith "parse_seg_corresp" in
+ let id_div,id_ab = match id with
+ {corref=""; prefix="txt"; numbers=[id_div;id_ab]; suffix="ab"} -> id_div,id_ab
+ | _ -> failwith "parse_seg_corresp" in
+ id_div,id_ab,beg,len
+
+let pos_set = StringSet.of_list
+ ["subst";"depr";"ppron12";"ppron3";"siebie";"prep";"adj";"adjc";"adjp";"adja";"num";
+ "adv";"ger";"pact";"ppas";"fin";"bedzie";"praet";"winien";"impt";
+ "imps";"pred";"aglt";"inf";"pcon";"pant";"qub";"comp";"conj";"interj";"burk";"interp";
+ "brev";"xxx";"numcol"]
+
+let parse_disamb disamb =
+ if disamb = "::interp" then ":","interp",[] else
+ if disamb = ":-):interp" then ":-)","interp",[] else
+ (* if Xstring.check_sufix ":interp" disamb then Xstring.cut_sufix ":interp" disamb, "interp", [] else *)
+ match Xstring.split_delim ":" disamb with
+ lemma1 :: lemma2 :: "subst" :: interp -> lemma1 ^ ":" ^ lemma2,"subst",interp
+ | lemma1 :: lemma2 :: lemma3 :: "subst" :: interp -> lemma1 ^ ":" ^ lemma2 ^ ":" ^ lemma3,"subst",interp
+ | lemma :: pos :: interp ->
+ if StringSet.mem pos_set pos then lemma,pos,interp
+ else failwith ("parse_disamb: " ^ disamb)
+ | _ -> failwith "parse_disamb"
+
+let rec merge_tokens name id_p rev = function
+ (corresp,nps,{corref=""; prefix="segm"; numbers=[id_segm_p;id_segm_s]; suffix="seg"}) :: segmentation,
+ ({corref="ann_segmentation.xml"; prefix="segm"; numbers=[c_segm_p;c_segm_s]; suffix="seg"},
+ {corref=""; prefix="morph"; numbers=[id_morph_p;id_morph_s]; suffix="seg"},orth,disamb) :: morphosyntax ->
+ (* if id_p <> id_segm_p then Printf.printf "merge_tokens inconsistent numbering: %s segm_%d-p segm_%d.%d-s\n" name id_p id_segm_p id_segm_s; *)
+ if id_segm_p <> c_segm_p || id_segm_p <> id_morph_p then failwith "merge_tokens 2" else
+ if id_segm_s <> c_segm_s || c_segm_s <> id_morph_s then failwith "merge_tokens 3" else
+ let id_div,id_ab,beg,len = parse_seg_corresp corresp in(
+ (* if id_div <> id_p then (*failwith*)print_endline (Printf.sprintf "merge_tokens 4: %s %d %s" name id_p corresp); (*else*) *)
+ let lemma,cat,interp = parse_disamb disamb in
+ merge_tokens name id_p ((id_div,id_ab,beg,nps,len,orth,lemma,cat,interp) :: rev) (segmentation,morphosyntax))
+ | [],[] -> List.rev rev
+ | _ -> failwith "merge_tokens 1"
+
+let rec merge_sentences name id_p rev = function
+ ({corref=""; prefix="segm"; numbers=[id_segm_p;id_segm_s]; suffix="s"},segm_tokens) :: segmentation,
+ ({corref="ann_segmentation.xml"; prefix="segm"; numbers=[c_segm_p;c_segm_s]; suffix="s"},
+ {corref=""; prefix="morph"; numbers=[id_morph_p;id_morph_s]; suffix="s"},morph_tokens) :: morphosyntax ->
+ (* if id_p <> id_segm_p then Printf.printf "merge_sentences inconsistent numbering: %s segm_%d-p segm_%d.%d-s\n" name id_p id_segm_p id_segm_s; *)
+ if id_segm_p <> c_segm_p || id_segm_p <> id_morph_p then failwith "merge_sentences 2" else
+ if id_segm_s <> c_segm_s || c_segm_s <> id_morph_s then failwith "merge_sentences 3" else
+ let tokens = merge_tokens name id_p [] (segm_tokens,morph_tokens) in
+ merge_sentences name id_p ((id_segm_p,id_segm_s,tokens) :: rev) (segmentation,morphosyntax)
+ | [],[] -> List.rev rev
+ | _ -> failwith "merge_sentences"
+
+let rec merge_entries name rev = function
({corref=""; prefix="txt"; numbers=[id_div]; suffix="div"},paragraphs) :: text,
({corref="text.xml"; prefix="txt"; numbers=[c_div]; suffix="div"},
{corref=""; prefix="segm"; numbers=[id_segm_p]; suffix="p"},segm_sentences) :: segmentation,
({corref="ann_segmentation.xml"; prefix="segm"; numbers=[c_segm_p]; suffix="p"},
{corref=""; prefix="morph"; numbers=[id_morph_p]; suffix="p"},morph_sentences) :: morphosyntax ->
if id_div <> c_div || c_div <> id_segm_p || id_segm_p <> c_segm_p || c_segm_p <> id_morph_p then failwith "merge_entries 2" else
- merge_entries ((id_div,paragraphs,segm_sentences,morph_sentences) :: rev) (text,segmentation,morphosyntax)
+ let sentences = merge_sentences name id_div [] (segm_sentences,morph_sentences) in
+ merge_entries name ((id_div,paragraphs,sentences) :: rev) (text,segmentation,morphosyntax)
| [],[],[] -> List.rev rev
| _ -> failwith "merge_entries"
@@ -174,7 +232,7 @@ let nkjp_path = "../../NLP resources/NKJP-PodkorpusMilionowy-1.2/"
let _ =
let names = get_folders nkjp_path in
Xlist.iter names (fun name ->
- print_endline name;
+ (* print_endline name; *)
let typ,channel = load_header nkjp_path name in
(* print_endline typ; *)
(* print_endline channel; *)
@@ -182,7 +240,7 @@ let _ =
let text = load_text nkjp_path name in
let segmentation = load_segmentation nkjp_path name in
let morphosyntax = load_morphosyntax nkjp_path name in
- let entries = merge_entries [] (text,segmentation,morphosyntax) in
+ let entries = merge_entries name [] (text,segmentation,morphosyntax) in
())
(*
diff --git a/subsyntax/ENIAM_MWE.ml b/subsyntax/ENIAM_MWE.ml
index a13eddf..e111b15 100644
--- a/subsyntax/ENIAM_MWE.ml
+++ b/subsyntax/ENIAM_MWE.ml
@@ -30,7 +30,7 @@ let load_dict dict filename =
let mwe_dict =
let dict = load_dict StringMap.empty brev_filename in
- let dict = load_dict dict fixed_filename in
+ let dict = try load_dict dict fixed_filename with _ -> (prerr_endline ("ENIAMsubsyntax file " ^ fixed_filename ^ " not found"); dict) in
(* let dict = load_dict dict complete_entries_filename in*)
let dict = load_dict dict mwe_filename in
dict
diff --git a/subsyntax/ENIAMsubsyntax.ml b/subsyntax/ENIAMsubsyntax.ml
index adb3e28..e0eef8c 100644
--- a/subsyntax/ENIAMsubsyntax.ml
+++ b/subsyntax/ENIAMsubsyntax.ml
@@ -200,7 +200,7 @@ let select_tokens paths =
(* | Dig(value,cat) -> t :: paths *)
| Other orth -> t :: paths
| Lemma(lemma,pos,interp) -> if pos = "brev" then paths else t :: paths
- | Proper(lemma,pos,interp,cat) -> t :: paths
+ | Proper(lemma,pos,interp,cat) -> if pos = "brev" then paths else t :: paths
(* | Compound _ -> t :: paths *)
| _ -> paths))
@@ -213,6 +213,7 @@ let load_proper_name proper = function
let proper_names =
let proper = File.fold_tab proper_names_filename StringMap.empty load_proper_name in
let proper = File.fold_tab proper_names_filename2 proper load_proper_name in
+ let proper = File.fold_tab proper_names_filename3 proper load_proper_name in
proper
let remove l s =
diff --git a/subsyntax/ENIAMsubsyntaxTypes.ml b/subsyntax/ENIAMsubsyntaxTypes.ml
index c1c4cb3..e92dace 100644
--- a/subsyntax/ENIAMsubsyntaxTypes.ml
+++ b/subsyntax/ENIAMsubsyntaxTypes.ml
@@ -55,6 +55,7 @@ let lemma_frequencies_filename = resource_path ^ "/subsyntax/NKJP1M-lemma-freq.t
let proper_names_filename2 = resource_path ^ "/subsyntax/proper_names.tab" *)
let proper_names_filename = resource_path ^ "/subsyntax/proper_names_sgjp_polimorf_20151020.tab"
let proper_names_filename2 = resource_path ^ "/subsyntax/proper_names_20160104.tab"
+let proper_names_filename3 = resource_path ^ "/subsyntax/ne.tab"
let int_of_mode = function
Raw -> 0
diff --git a/subsyntax/resources/ne.tab b/subsyntax/resources/ne.tab
new file mode 100644
index 0000000..1f7410b
--- /dev/null
+++ b/subsyntax/resources/ne.tab
@@ -0,0 +1,34 @@
+Akademia Sztuki ORGANIZACJA
+Atelier Bizio + Ligierko ORGANIZACJA
+Instytut Architektury i Planowania Przestrzennego ORGANIZACJA
+Katedra Architektury Współczesnej Teorii i Metodologii Projektowania ORGANIZACJA
+VII Liceum Ogólnokształcące im. K.K. Baczyńskiego ORGANIZACJA
+IV Liceum Ogólnokształcące im. L. Szenwalda ORGANIZACJA
+Muzeum Narodowe ORGANIZACJA
+Nagroda Artystyczna m. Szczecina WYRÓŻNIENIE
+Zachodniopomorski Nobel WYRÓŻNIENIE
+Politechnika Krakowska ORGANIZACJA
+Politechnika Szczecińska ORGANIZACJA
+Pracownia Podstaw Projektowania ORGANIZACJA
+Przegląd Teatrów Małych Form „Kontrapunkt” ORGANIZACJA
+Mistrzowska Szkoła Reżyserii Filmowej Andrzeja Wajdy ORGANIZACJA
+Uniwersytet im. M. Kopernika ORGANIZACJA
+Zachodniopomorski Uniwersytet Technologiczny ORGANIZACJA
+Wydział Budownictwa i Architektury ORGANIZACJA
+Wydział Stuk Wizualnych ORGANIZACJA
+Zakład Teorii Architektury, Historii i Konserwacji Zabytków ORGANIZACJA
+Festiwal Polskich Sztuk Współczesnych R@Port WYDARZENIE
+Sosnowiec MIASTO
+Stefan IMIĘ
+Józefa IMIĘ
+Szczecin MIASTO
+Waldemar IMIĘ
+Marzęcki NAZWISKO
+Austria KRAJ
+Czechy KRAJ
+Niemcy KRAJ
+Francja KRAJ
+Litwa KRAJ
+USA KRAJ
+Rosja KRAJ
+
diff --git a/tokenizer/ENIAMacronyms.ml b/tokenizer/ENIAMacronyms.ml
index 36b7646..1c113fe 100644
--- a/tokenizer/ENIAMacronyms.ml
+++ b/tokenizer/ENIAMacronyms.ml
@@ -21,7 +21,7 @@ open ENIAMtokenizerTypes
let mte_patterns =
let lines = try File.load_lines mte_filename
- with _ -> (print_endline ("ENIAMtokenizer mte file " ^ mte_filename ^ " not found"); []) in
+ with _ -> (prerr_endline ("ENIAMtokenizer mte file " ^ mte_filename ^ " not found"); []) in
let l = List.rev (Xlist.rev_map lines (fun line ->
match Str.split (Str.regexp "\t") line with
[orths; lemma; interp] -> Str.split (Str.regexp " ") orths, lemma, interp
diff --git a/tokenizer/ENIAMtokenizerTypes.ml b/tokenizer/ENIAMtokenizerTypes.ml
index 4c4d1cf..72c45cc 100644
--- a/tokenizer/ENIAMtokenizerTypes.ml
+++ b/tokenizer/ENIAMtokenizerTypes.ml
@@ -72,4 +72,5 @@ let resource_path =
if Sys.file_exists "/usr/local/share/eniam" then "/usr/local/share/eniam" else
failwith "resource directory does not exists"
-let mte_filename = resource_path ^ "/tokenizer/mte.tab"
+(* let mte_filename = resource_path ^ "/tokenizer/mte.tab" *)
+let mte_filename = resource_path ^ "/tokenizer/mte_20151215.tab"
diff --git a/tokenizer/makefile b/tokenizer/makefile
index 2d23090..cabf196 100755
--- a/tokenizer/makefile
+++ b/tokenizer/makefile
@@ -18,7 +18,7 @@ install: all
mkdir -p /usr/share/eniam/tokenizer
cp resources/mte_20151215.tab /usr/share/eniam/tokenizer/mte_20151215.tab
cp resources/README /usr/share/eniam/tokenizer/README
- ln -s /usr/share/eniam/tokenizer/mte_20151215.tab /usr/share/eniam/tokenizer/mte.tab
+# ln -s /usr/share/eniam/tokenizer/mte_20151215.tab /usr/share/eniam/tokenizer/mte.tab
install-local: all
mkdir -p $(INSTALLDIR)
@@ -28,7 +28,7 @@ install-local: all
mkdir -p /usr/local/share/eniam/tokenizer
cp resources/mte_20151215.tab /usr/local/share/eniam/tokenizer/mte_20151215.tab
cp resources/README /usr/local/share/eniam/tokenizer/README
- ln -s /usr/local/share/eniam/tokenizer/mte_20151215.tab /usr/local/share/eniam/tokenizer/mte.tab
+# ln -s /usr/local/share/eniam/tokenizer/mte_20151215.tab /usr/local/share/eniam/tokenizer/mte.tab
eniam-tokenizer.cma: $(SOURCES)
ocamlc -linkall -a -o eniam-tokenizer.cma $(OCAMLFLAGS) $^
--
libgit2 0.22.2