Problem

DeepL glossaries are 1:1 source to target only. However, it is possible to have multiple source terms point to the same target term. In many term collections there are synonyms in the source or forbidden terms. If a source text to be translated contains one of the synonyms or even the forbidden term (e.g. a wrong spelling is used) the glossary will not know which target term should be used although in the term collection this information is given and in the editor users will see the terms tagged.

Solution

Change the logic for creating the glossary:

• find all source terms given, admitted, preferred, regulated, even deprecated/forbidden
• pair them 1:1 with the best category target term

Example: