Skip to content
Projects
Groups
Snippets
Help
Loading...
Help
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
P
purescript-gargantext
Project
Project
Details
Activity
Releases
Cycle Analytics
Repository
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Charts
Issues
0
Issues
0
List
Board
Labels
Milestones
Merge Requests
0
Merge Requests
0
CI / CD
CI / CD
Pipelines
Jobs
Schedules
Charts
Wiki
Wiki
Snippets
Snippets
Members
Members
Collapse sidebar
Close sidebar
Activity
Graph
Charts
Create a new issue
Jobs
Commits
Issue Boards
Open sidebar
Grégoire Locqueville
purescript-gargantext
Commits
c3987256
Commit
c3987256
authored
Nov 27, 2020
by
Nicolas Pouillard
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
highlightNgrams now supports nested/crossing ngrams
TODO UI & better testing
parent
dd85f69e
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
43 additions
and
47 deletions
+43
-47
Core.purs
src/Gargantext/Components/NgramsTable/Core.purs
+43
-47
No files found.
src/Gargantext/Components/NgramsTable/Core.purs
View file @
c3987256
...
...
@@ -93,7 +93,7 @@ import Data.Lens.Index (class Index, ix)
import Data.Lens.Iso.Newtype (_Newtype)
import Data.Lens.Record (prop)
import Data.List ((:), List(Nil))
import Data.List as L
ist
import Data.List as L
import Data.Map (Map)
import Data.Map as Map
import Data.Maybe (Maybe(..), fromMaybe, fromMaybe', isJust)
...
...
@@ -436,74 +436,70 @@ wordBoundaryReg2 = case R.regex ("(" <> wordBoundaryChars <> ")\\1") (R.global <
Left e -> unsafePartial $ crashWith e
Right r -> r
type HighlightAccumulator =
{ i0 :: Int -- where are we in input
, s :: String -- == drop i0 input
, l :: List (Tuple String (List (Tuple NgramsTerm TermList)))
}
type HighlightElement = Tuple String (List (Tuple NgramsTerm TermList))
type HighlightAccumulator = List HighlightElement
-- TODO: while this function works well with word boundaries,
-- it inserts too many spaces.
highlightNgrams :: CTabNgramType -> NgramsTable -> String -> Array
(Tuple String (List (Tuple NgramsTerm TermList)))
highlightNgrams :: CTabNgramType -> NgramsTable -> String -> Array
HighlightElement
highlightNgrams ntype table@(NgramsTable {ngrams_repo_elements: elts}) input0 =
-- trace {pats, input0, input, ixs} \_ ->
let sN = unsafePartial (foldl goFold {i0: 0, s: input, l: Nil} ixs) in
A.reverse (A.fromFoldable (consNonEmpty (undb (init sN.s)) sN.l))
A.fromFoldable ((\(s /\ ls)-> undb s /\ ls) <$> unsafePartial (foldl goFold ((input /\ Nil) : Nil) ixs))
where
spR x = " " <> R.replace wordBoundaryReg "$1$1" x <> " "
reR = R.replace wordBoundaryReg " "
db = S.replaceAll (S.Pattern " ") (S.Replacement " ")
sp x = " " <> db x <> " "
undb = R.replace wordBoundaryReg2 "$1"
init x = S.take (S.length x - 1) x
input = spR input0
pats = A.fromFoldable (Map.keys elts)
ixs = indicesOfAny (sp <<< ngramsTermText <$> pats) (normNgramInternal ntype input)
consOnJustTail s xs@(Tuple _ (_ : _) : _) = Tuple s Nil : xs
consOnJustTail _ xs = xs
consNonEmpty x xs
| S.null x = xs
| otherwise = Tuple x Nil : xs
goAcc :: Partial => Tuple NgramsTerm Int -> Int -> HighlightAccumulator -> HighlightAccumulator
goAcc (pat /\ lpat) i { i0, s, l } =
splitAcc :: Partial => Int -> HighlightAccumulator
-> Tuple HighlightAccumulator HighlightAccumulator
splitAcc i = go 0 Nil
where
go j pref acc =
case compare i j of
LT -> crashWith "highlightNgrams: splitAcc': i < j"
EQ -> L.reverse pref /\ acc
GT ->
case acc of
Nil -> crashWith "highlightNgrams: splitAcc': acc=Nil" -- pref /\ Nil
elt@(s /\ ls) : elts ->
let slen = S.length s in
case compare i (j + slen) of
LT -> let {before: s0, after: s1} = S.splitAt (i - j) s in
L.reverse ((s0 /\ ls) : pref) /\ ((s1 /\ ls) : elts)
EQ -> L.reverse (elt : pref) /\ elts
GT -> go (j + slen) (elt : pref) elts
extractInputTextMatch :: Int -> Int -> String -> String
extractInputTextMatch i len input = undb $ S.take len $ S.drop (i + 1) input
addNgramElt ng ne_list (elt /\ elt_lists) = (elt /\ ((ng /\ ne_list) : elt_lists))
goAcc :: Partial => Int -> HighlightAccumulator -> Tuple NgramsTerm Int -> HighlightAccumulator
goAcc i acc (pat /\ lpat) =
case lookupRootList pat table of
Nothing ->
crashWith "highlightNgrams: pattern missing from table"
Just ne_list ->
let
s1 = {-if i <= i0 then-} S.splitAt (i - i0) s {-else S.drop i input-}
s2 = S.splitAt lpat (S.drop 1 s1.after)
s3 = S.splitAt 1 s2.after
unspB = if i0 == 0 then S.drop 1 else identity
s3b = s3.before
text = undb s2.before
(acc0 /\ acc1_2) = splitAcc i acc
(acc1 /\ acc2) = splitAcc (lpat + 1) acc1_2
text = extractInputTextMatch i lpat input
ng = normNgram ntype text
in
-- trace {s, i, i0, s1, s2, s3, pat, lpat, s3b} \_ ->
-- `undb s2.before` and pat might differ by casing only!
{ i0: i + lpat + 2
, s: s3.after
, l: Tuple text ((normNgram ntype text /\ ne_list) : Nil) :
consOnJustTail s3b
(consNonEmpty (unspB (undb s1.before)) l)
}
-- NOTE that only the first matching pattern is used, the others are ignored!
acc0 <> (addNgramElt ng ne_list <$> acc1) <> acc2
goFold :: Partial => HighlightAccumulator -> Tuple Int (Array Int) -> HighlightAccumulator
goFold acc@{ i0, s, l } (Tuple i pis)
| i < i0 =
-- Skip this pattern which is overlapping with a previous one.
{ i0, s, l }
| otherwise =
let pats' = A.sortWith snd $
map (\pat -> pat /\ S.length (db (ngramsTermText pat))) $
fromMaybe' (\_ -> crashWith "highlightNgrams: out of bounds pattern") $
traverse (A.index pats) pis in
case List.fromFoldable pats' of
Nil -> { i0, s, l }
pat : _ -> goAcc pat i acc
goFold acc (Tuple i pis) = foldl (goAcc i) acc $
-- A.sortWith snd $
map (\pat -> pat /\ S.length (db (ngramsTermText pat))) $
fromMaybe' (\_ -> crashWith "highlightNgrams: out of bounds pattern") $
traverse (A.index pats) pis
-----------------------------------------------------------------------------------
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment