[DATABASE] New Table for postagging

devops/postgres/schema.sql

@@ -39,6 +39,8 @@ CREATE TABLE public.nodes (
 );
 ALTER TABLE public.nodes OWNER TO gargantua;
 
+--------------------------------------------------------------
+-- | Ngrams
 CREATE TABLE public.ngrams (
     id SERIAL,
     terms CHARACTER varying(255),
@@ -47,6 +49,20 @@ CREATE TABLE public.ngrams (
 );
 ALTER TABLE public.ngrams OWNER TO gargantua;
 
+-- | Ngrams PosTag
+CREATE TABLE public.ngrams_postag (
+    id SERIAL,
+    lang_id INTEGER,
+    algo_id INTEGER,
+    postag CHARACTER varying(5),
+    ngrams_id INTEGER NOT NULL,
+    lemm_id   INTEGER NOT NULL,
+    score     INTEGER DEFAULT 1 ::integer NOT NULL,
+    FOREIGN KEY (ngrams_id) REFERENCES public.ngrams(id) ON DELETE CASCADE,
+    FOREIGN KEY (lemm_id) REFERENCES public.ngrams(id) ON DELETE CASCADE
+);
+ALTER TABLE public.ngrams_postag OWNER TO gargantua;
+
 --------------------------------------------------------------
 CREATE TABLE public.node_ngrams (
     id SERIAL,

src/Gargantext/Core/Text/Terms/Multi/PosTagging.hs

@@ -115,7 +115,7 @@ corenlp' lang txt = do
     let properties = case lang of
             EN -> "{\"annotators\": \"tokenize,ssplit,pos,ner\", \"outputFormat\": \"json\"}"
             -- FR -> "{\"annotators\": \"tokenize,ssplit,pos,ner\", \"outputFormat\": \"json\"}"
-            FR -> "{\"annotators\": \"tokenize,ssplit,pos,ner\", \"parse.model\":\"edu/stanford/nlp/models/lexparser/frenchFactored.ser.gz\", \"pos.model\":\"edu/stanford/nlp/models/pos-tagger/french/french.tagger\", \"tokenize.language\":\"fr\", \"outputFormat\": \"json\"}"
+            FR -> "{\"annotators\": \"tokenize,ssplit,pos,lemma,ner\", \"parse.model\":\"edu/stanford/nlp/models/lexparser/frenchFactored.ser.gz\", \"pos.model\":\"edu/stanford/nlp/models/pos-tagger/french/french.tagger\", \"tokenize.language\":\"fr\", \"outputFormat\": \"json\"}"
             _  -> panic $ pack "not implemented yet"
     url <- parseRequest $ "POST http://localhost:9000/?properties=" <> properties
     let request = setRequestBodyLBS (cs txt) url

src/Gargantext/Database/Schema/Ngrams.hs

 {-|
-Module      : Gargantext.Database.Schema.Ngrams
+Module      : Gargantext.Database.Schema.NgramsPostag
 Description : Ngram connection to the Database
 Copyright   : (c) CNRS, 2017-Present
 License     : AGPL + CECILL v3

src/Gargantext/Database/Schema/NgramsPostag.hs

+{-|
+Module      : Gargantext.Database.Schema.NgramsPostag
+Description : Ngram connection to the Database
+Copyright   : (c) CNRS, 2017-Present
+License     : AGPL + CECILL v3
+Maintainer  : team@gargantext.org
+Stability   : experimental
+Portability : POSIX
+
+Each Ngrams has a pos-tagging version to ease the default groups of
+ngrams in NgramsTerm Lists.
+
+-}
+
+{-# LANGUAGE Arrows                     #-}
+{-# LANGUAGE FunctionalDependencies     #-}
+{-# LANGUAGE QuasiQuotes            #-}
+{-# LANGUAGE TemplateHaskell        #-}
+
+module Gargantext.Database.Schema.NgramsPostag
+  where
+
+import Data.Text (Text)
+import Gargantext.Prelude
+import Gargantext.Database.Schema.Prelude
+
+data NgramsPosTagPoly id
+                      lang_id
+                      algo_id
+                      postag
+                      ngrams_id
+                      lemm_id
+                      score
+  = NgramsPosTagDB { _ngramsPosTag_id        :: !id
+                   , _ngramsPosTag_lang_id   :: !lang_id
+                   , _ngramsPosTag_algo_id   :: !algo_id
+                   , _ngramsPosTag_postag    :: !postag
+                   , _ngramsPosTag_ngrams_id :: !ngrams_id
+                   , _ngramsPosTag_lemm_id   :: !lemm_id
+                   , _ngramsPosTag_score     :: !score
+                   } deriving (Show)
+
+------------------------------------------------------------------------
+
+type NgramsPosTagWrite = NgramsPosTagPoly (Maybe (Column PGInt4))
+                                   (Column PGInt4)
+                                   (Column PGInt4)
+                                   (Maybe (Column PGText))
+                                   (Column PGInt4)
+                                   (Column PGInt4)
+                                   (Maybe (Column PGInt4))
+
+type NgramsPosTagRead  = NgramsPosTagPoly (Column PGInt4)
+                                   (Column PGInt4)
+                                   (Column PGInt4)
+                                   (Column PGText)
+                                   (Column PGInt4)
+                                   (Column PGInt4)
+                                   (Column PGInt4)
+
+type NgramsPosTagReadNull =  NgramsPosTagPoly (Column (Nullable PGInt4))
+                                   (Column (Nullable PGInt4))
+                                   (Column (Nullable PGInt4))
+                                   (Column (Nullable PGText))
+                                   (Column (Nullable PGInt4))
+                                   (Column (Nullable PGInt4))
+                                   (Column (Nullable PGInt4))
+
+type NgramsPosTagDB = NgramsPosTagPoly Int Int Int Text Int Int Int