Updating stanford corenlp to 4.3.2; adding more languages

There are models for Spanish, that have been added now. Also the
Hungarian language has been added to the list of supported
languages (for tesseract mainly, no nlp models)
This commit is contained in:
eikek
2021-11-20 14:31:39 +01:00
parent 20fc9955ba
commit 501c6f2988
18 changed files with 162 additions and 40 deletions

View File

@ -30,7 +30,7 @@ object Language {
override val allowsNLP = true
}
object NLPLanguage {
val all: NonEmptyList[NLPLanguage] = NonEmptyList.of(German, English, French)
val all: NonEmptyList[NLPLanguage] = NonEmptyList.of(German, English, French, Spanish)
}
case object German extends NLPLanguage {
@ -53,11 +53,16 @@ object Language {
val iso3 = "ita"
}
case object Spanish extends Language {
case object Spanish extends NLPLanguage {
val iso2 = "es"
val iso3 = "spa"
}
case object Hungarian extends Language {
val iso2 = "hu"
val iso3 = "hun"
}
case object Portuguese extends Language {
val iso2 = "pt"
val iso3 = "por"
@ -125,6 +130,7 @@ object Language {
French,
Italian,
Spanish,
Hungarian,
Dutch,
Portuguese,
Czech,