Everything works well, except that the verification is done independently of the source. To give an example: we are using the word “auf” in German as a normal preposition. But “auf” is also one possible translation for “open” and in this context set as forbidden. Obviously in other contexts “auf” has to be used quite often. But if we run F8, every single “auf” is detected as an error even if in the source there is no sign of the word “open”. The termbase is really big and the texts very long, so this takes us very much extra time other than not seeming logical at all.
Does anyone know if there is a reason for it working like this? Or better: does anyone know how we can solve this problem?
Translate