Wrong suggestions for German compounds

Development
1

Hello,

This is closely related to # Incorrect suggestion given by languagetool, but the solution of adding words to the dictionary is not viable for us on the long term.

LanguageTool flags some German compound words that are correct in our context, and provides surprising suggestions:

1.) Line 1, column 1, Rule ID: GERMAN_SPELLER_RULE premium: false prio=-3
Message: Möglicher Tippfehler gefunden.
Suggestion: Bodenverbbesserungsmitteln; Bodenverbessbergungsmitteln; Bodenverbessgrünessmitteln; Bodenverbesserdungsmitteln; Bodenverbessgeruchsmitteln
Bodenverbesserungsmitteln
^^^^^^^^^^^^^^^^^^^^^^^^^

1.) Line 3, column 1, Rule ID: GERMAN_SPELLER_RULE premium: false prio=-3
Message: Möglicher Tippfehler gefunden.
Suggestion: Bombenentenmaterial; Kolbenentenmaterial; Bonbonentenmaterial
Komponentenmaterial
^^^^^^^^^^^^^^^^^^^

Yet the compound splitter at German Compound Split yields:

Boden verbesserungs mitteln
Komponenten material

… so we would expect those compounds be accepted as they are. Is it possible? If not, is there a way to disable the suggestions for German compound words, and have LT simply flag a potential error?

Such compound words are very frequent in our texts, so adding them all manually is not an option.

We are using a local LT 6.3 server with custom resources and XML rules, so no Java.

Thank you,
Adrien

2

Not easily, there are several limitations to this, and it’s not documented, because it’s still being worked on. “Komponentenmaterial” and “Bodenverbesserungsmitteln” already accepted by the latest development version of LT, though.

No, there’s no option for that.

3

Thanks a lot for the quick reply.