[NgramsGroup] refact parameters

src/Gargantext/Core/Text/Group.hs

@@ -51,13 +51,16 @@ data StopSize = StopSize {unStopSize :: !Int}
 -- a first grouping option to user and get some
 -- enriched data to better learn and improve that algo
 
+data GroupParams = GroupParams { unGroupParams_lang  :: !Lang
+                               , unGroupParams_len   :: !Int
+                               , unGroupParams_limit :: !Int
+                               , unGroupParams_stopSize :: !StopSize
+                               }
 
-ngramsGroup :: Lang
-            -> Int
-            -> Int
+ngramsGroup :: GroupParams
             -> Text
             -> Text
-ngramsGroup l _m _n = Text.intercalate " "
+ngramsGroup (GroupParams l _m _n _) = Text.intercalate " "
                   . map (stem l)
                   -- . take n
                   . List.sort

src/Gargantext/Core/Text/List.hs

@@ -29,7 +29,6 @@ import qualified Data.Text as Text
 -- import Gargantext.API.Ngrams.Tools (getCoocByNgrams', Diagonal(..))
 import Gargantext.API.Ngrams.Types (NgramsElement, mkNgramsElement, NgramsTerm(..), RootParent(..), mSetFromList)
 import Gargantext.API.Ngrams.Types (RepoCmdM)
-import Gargantext.Core (Lang(..))
 import Gargantext.Core.Text (size)
 import Gargantext.Core.Text.List.Social (flowSocialList, invertForw)
 import Gargantext.Core.Text.Metrics (scored', Scored(..), normalizeGlobal, normalizeLocal)
@@ -53,15 +52,12 @@ buildNgramsLists :: ( RepoCmdM env err m
                     , HasNodeError err
                     )
                  => User
-                 -> Lang
-                 -> Int
-                 -> Int
-                 -> StopSize
+                 -> GroupParams
                  -> UserCorpusId
                  -> MasterCorpusId
                  -> m (Map NgramsType [NgramsElement])
-buildNgramsLists user l n m s uCid mCid = do
-  ngTerms     <- buildNgramsTermsList user l n m s uCid mCid
+buildNgramsLists user gp uCid mCid = do
+  ngTerms     <- buildNgramsTermsList user gp uCid mCid
   othersTerms <- mapM (buildNgramsOthersList user uCid identity)
                       [Authors, Sources, Institutes]
   pure $ Map.unions $ othersTerms <> [ngTerms]
@@ -105,14 +101,11 @@ buildNgramsTermsList :: ( HasNodeError err
                         , HasTreeError err
                         )
                         => User
-                        -> Lang
-                        -> Int
-                        -> Int
-                        -> StopSize
+                        -> GroupParams
                         -> UserCorpusId
                         -> MasterCorpusId
                         -> m (Map NgramsType [NgramsElement])
-buildNgramsTermsList user l n m _s uCid mCid = do
+buildNgramsTermsList user groupParams uCid mCid = do
 
 -- Computing global speGen score
   allTerms <- Map.toList <$> getTficf uCid mCid NgramsTerms
@@ -137,7 +130,7 @@ buildNgramsTermsList user l n m _s uCid mCid = do
 
   -- Grouping the ngrams and keeping the maximum score for label
   let grouped = groupStems'
-        $ map (\(t,d) -> let stem = ngramsGroup l n m t
+        $ map (\(t,d) -> let stem = ngramsGroup groupParams t
                           in ( stem
                              , GroupedText Nothing t d Set.empty (size t) stem Set.empty
                              )
@@ -186,7 +179,7 @@ buildNgramsTermsList user l n m _s uCid mCid = do
                 $ groupedMonoHead <> groupedMultHead
 
     -- grouping with Set NodeId
-    contextsAdded = foldl' (\mapGroups' k -> let k' = ngramsGroup l n m k
+    contextsAdded = foldl' (\mapGroups' k -> let k' = ngramsGroup groupParams k
                                     in case Map.lookup k' mapGroups'  of
                                       Nothing -> mapGroups'
                                       Just g  -> case Map.lookup k mapTextDocIds of

src/Gargantext/Database/Action/Flow.hs

@@ -66,7 +66,7 @@ import Gargantext.Core.Flow.Types
 import Gargantext.Core.Text
 import Gargantext.Core.Text.Corpus.Parsers (parseFile, FileFormat)
 import Gargantext.Core.Text.List (buildNgramsLists)
-import Gargantext.Core.Text.Group (StopSize(..))
+import Gargantext.Core.Text.Group (StopSize(..), GroupParams(..))
 import Gargantext.Core.Text.Terms
 import Gargantext.Core.Text.Terms.Mono.Stem.En (stemIt)
 import Gargantext.Core.Types (Terms(..))
@@ -221,7 +221,7 @@ flowCorpusUser l user corpusName ctype ids = do
 
   -- User List Flow
   (masterUserId, _masterRootId, masterCorpusId) <- getOrMk_RootWithCorpus (UserName userMaster) (Left "") ctype
-  ngs         <- buildNgramsLists user l 2 3 (StopSize 3) userCorpusId masterCorpusId
+  ngs         <- buildNgramsLists user (GroupParams l 2 3 (StopSize 3)) userCorpusId masterCorpusId
   _userListId <- flowList_DbRepo listId ngs
   _mastListId <- getOrMkList masterCorpusId masterUserId
   -- _ <- insertOccsUpdates userCorpusId mastListId

src/Gargantext/Database/Action/Metrics/NgramsByNode.hs

@@ -16,6 +16,7 @@ Ngrams by node enable contextual metrics.
 module Gargantext.Database.Action.Metrics.NgramsByNode
   where
 
+
 import Data.Map.Strict (Map, fromListWith, elems, toList)
 import Data.Map.Strict.Patch (PatchMap, Replace, diff)
 import Data.Set (Set)
@@ -24,18 +25,14 @@ import Data.Tuple.Extra (second, swap)
 import Database.PostgreSQL.Simple.SqlQQ (sql)
 import Database.PostgreSQL.Simple.Types (Values(..), QualifiedIdentifier(..))
 import Debug.Trace (trace)
-import qualified Data.List as List
-import qualified Data.Map.Strict as Map
-import qualified Data.Set as Set
-import qualified Data.Text as Text
-import qualified Database.PostgreSQL.Simple as DPS
-
-import Gargantext.Core (Lang(..))
 import Gargantext.Database.Admin.Config (nodeTypeId)
 import Gargantext.Database.Admin.Types.Node -- (ListId, CorpusId, NodeId)
 import Gargantext.Database.Prelude (Cmd, runPGSQuery)
 import Gargantext.Database.Schema.Ngrams (ngramsTypeId, NgramsType(..))
 import Gargantext.Prelude
+import qualified Data.Map.Strict as Map
+import qualified Data.Set as Set
+import qualified Database.PostgreSQL.Simple as DPS