This could be an extra option in Translation memory settings > Language resources > Recognize group and should apply to any word excluding the first word of the segment (or that could be a suboption).
It is already implemented from time immemorial in other tools, i.e. Wordfast.
Top Comments