This is an automated email from the ASF dual-hosted git repository.
sergeykamov pushed a commit to branch NLPCRAFT-70_NEW
in repository https://gitbox.apache.org/repos/asf/incubator-nlpcraft.git
The following commit(s) were added to refs/heads/NLPCRAFT-70_NEW by this push:
new 389e733 WIP.
389e733 is described below
commit 389e73342c55d7ed45ee58037d8ec28c87589103
Author: Sergey Kamov <[email protected]>
AuthorDate: Tue Jul 6 17:20:20 2021 +0300
WIP.
---
.../nlpcraft/server/mdo/NCProbeModelMdo.scala | 1 -
.../ctxword/NCContextWordCategoriesEnricher.scala | 38 +++++-----------------
.../server/sugsyn/NCSuggestSynonymManager.scala | 17 ++--------
3 files changed, 11 insertions(+), 45 deletions(-)
diff --git
a/nlpcraft/src/main/scala/org/apache/nlpcraft/server/mdo/NCProbeModelMdo.scala
b/nlpcraft/src/main/scala/org/apache/nlpcraft/server/mdo/NCProbeModelMdo.scala
index d330ec7..cb85a83 100644
---
a/nlpcraft/src/main/scala/org/apache/nlpcraft/server/mdo/NCProbeModelMdo.scala
+++
b/nlpcraft/src/main/scala/org/apache/nlpcraft/server/mdo/NCProbeModelMdo.scala
@@ -19,7 +19,6 @@ package org.apache.nlpcraft.server.mdo
import org.apache.nlpcraft.server.mdo.impl._
-
@NCMdoEntity(sql = false)
case class NCCtxWordCategoriesConfigMdo(
@NCMdoField probeId: String,
diff --git
a/nlpcraft/src/main/scala/org/apache/nlpcraft/server/nlp/enrichers/ctxword/NCContextWordCategoriesEnricher.scala
b/nlpcraft/src/main/scala/org/apache/nlpcraft/server/nlp/enrichers/ctxword/NCContextWordCategoriesEnricher.scala
index 503e504..8307863 100644
---
a/nlpcraft/src/main/scala/org/apache/nlpcraft/server/nlp/enrichers/ctxword/NCContextWordCategoriesEnricher.scala
+++
b/nlpcraft/src/main/scala/org/apache/nlpcraft/server/nlp/enrichers/ctxword/NCContextWordCategoriesEnricher.scala
@@ -259,20 +259,13 @@ object NCContextWordCategoriesEnricher extends
NCServerEnricher {
* @return
*/
private def getCorpusData(cfg: NCCtxWordCategoriesConfigMdo, key:
ModelProbeKey, parent: Span = null):
- Map[
-
- /** Element ID */
- String, ElementData] =
- elemsCorpuses.synchronized {
- elemsCorpuses.get(key)
- } match {
+ Map[/** Element ID */String, ElementData] =
+ elemsCorpuses.synchronized { elemsCorpuses.get(key) } match {
case Some(cache) => cache
case None =>
val res = askSamples(cfg, parent)
- elemsCorpuses.synchronized {
- elemsCorpuses += key -> res
- }
+ elemsCorpuses.synchronized { elemsCorpuses += key -> res }
res
}
@@ -284,9 +277,7 @@ object NCContextWordCategoriesEnricher extends
NCServerEnricher {
* @return
*/
private def getValuesData(cfg: NCCtxWordCategoriesConfigMdo, key:
ModelProbeKey): ValuesHolder =
- valuesStems.synchronized {
- valuesStems.get(key)
- } match {
+ valuesStems.synchronized { valuesStems.get(key) } match {
case Some(cache) => cache
case None =>
def mkMap(convert: String => String): Map[String, Set[String]]
=
@@ -301,9 +292,7 @@ object NCContextWordCategoriesEnricher extends
NCServerEnricher {
val h = ValuesHolder(normal = normsMap, stems =
stemsMap.filter(p => !normsMap.keySet.contains(p._1)))
- valuesStems.synchronized {
- valuesStems += key -> h
- }
+ valuesStems.synchronized { valuesStems += key -> h }
h
}
@@ -338,10 +327,8 @@ object NCContextWordCategoriesEnricher extends
NCServerEnricher {
* @return
*/
@throws[NCE]
- private def askSamples(cfg: NCCtxWordCategoriesConfigMdo, parent: Span =
null): Map[
-
- /** Element ID */
- String, ElementData] = {
+ private def askSamples(cfg: NCCtxWordCategoriesConfigMdo, parent: Span =
null):
+ Map[/** Element ID */String, ElementData] = {
val corpusSeq = cfg.corpus.toSeq
val corpusWords = corpusSeq.map(parser.parse(_).map(_.word))
val nlpWords = corpusSeq.map(s => parser.parse(s))
@@ -392,16 +379,7 @@ object NCContextWordCategoriesEnricher extends
NCServerEnricher {
val respsSeq: Seq[(NCSuggestionRequest, Seq[NCWordSuggestion])] =
resps.toSeq
def mkMap(convert: (NCSuggestionRequest, NCWordSuggestion) =>
String):
- Map[
-
- /** Element ID */
- String,
-
- /** Word key */
- Map[String,
-
- /** Confidences */
- Seq[Double]]] = {
+ Map[/** Element ID */ String, /** Word key */ Map[String, /**
Confidences */ Seq[Double]]] = {
val seq: Seq[(String, Map[String, Double])] =
respsSeq.
map { case (req, suggs) =>
diff --git
a/nlpcraft/src/main/scala/org/apache/nlpcraft/server/sugsyn/NCSuggestSynonymManager.scala
b/nlpcraft/src/main/scala/org/apache/nlpcraft/server/sugsyn/NCSuggestSynonymManager.scala
index c2c8d23..3d83030 100644
---
a/nlpcraft/src/main/scala/org/apache/nlpcraft/server/sugsyn/NCSuggestSynonymManager.scala
+++
b/nlpcraft/src/main/scala/org/apache/nlpcraft/server/sugsyn/NCSuggestSynonymManager.scala
@@ -502,23 +502,12 @@ object NCSuggestSynonymManager extends NCService {
if (cnt.incrementAndGet() == batches.size) {
val min = minScoreOpt.getOrElse(DFLT_MIN_SCORE)
- val map: Map[NCSuggestionRequest,
Seq[NCWordSuggestion]] =
+ promise.success(
data.asScala.groupBy(_.request).map {
case (req, ress) =>
req ->
ress.flatMap(_.suggestions.filter(_.score >= min).toSeq).sortBy(-_.score)
- }
-
- // TODO ? logic?
-// val map: Map[NCSuggestionRequest,
Seq[NCWordSuggestion]] = data.asScala.groupBy(_.request).map(p =>
-// p._1 ->
-// p._2.
-// map(_.suggestions.map(p =>
(toStem(p.word), p.score))).
-// map(_.groupBy(_._1)).
-// flatMap(p => p.map(p => p._1 -> p._1 ->
p._2.map(_._2).sum / p._2.size).
-// filter(_._2 >= min).
-// map(p => NCWordSuggestion(p._1._2,
p._2)).toSeq).toSeq)
-
- promise.success(map)
+ }
+ )
}
()
},
