To reproduce, please import the xlf files of the above mentioned zip file into a new task and assign the above mentioned tm to it.

source language: de

target language: pt-PT

Then pre-translate the task.

There still should be nine segments untranslated after pre-translation, even though there are 100% matches for them.

These are segments with the following source content:

• STOP
• Vorschwaden
• AUS °C

The latest one also contains some tags.

Result of check and solution

Result of the check is, that the segments have not been pre-translated, because they were repetitions of a 100% match. And thus they were recognized as internal fuzzy matches, and internal fuzzy matches are not pre-translated.

We discussed that and decided, that this behavior is correct for real internal fuzzy matches, but not for repeated 100% matches.

So the behavior for internal fuzzy matches should stay (even if fuzzy matches would be pre-translated, because someone sets the match rate for pre-translation below 100%).

But for repeated 100% matches (and higher than 100%) the segment should be pre-translated. And the pre-translation match-rate should be set to 102% in the segment, if otherwise it would be lower (so if it would be 100% or 101%). If the pre-translation match-rate would be higher, than 102%, the segment should be pre-translated as normal (so as a 103% or 104% match) and the repetition should not play any role.