Walter Blaser 

I’ve heard Daniel Brockmann saying that not saving leading/trailing spaces in the TM is by design. And now I see this post in the Ideas section:  Respect spaces at the end of segments

And there is another idea:  Keeping trailing whitespaces when updating TM

I don’t know if the workaround mentioned of the hard space works with the Transit files…

I wished Auto-propagation kept the trailing spaces. Just noticed that a couple of days ago. Maybe it’s related to the fact that trailing spaces are not saved to the TM, but there must be a reasion for adding leading or trailing spaces, so they must be preserved at least in the Editor. I may add an idea about this tomorrow morning…

Hi Jesús , hi Daniel 

I understand Daniel's position. Trailing spaces should not be saved because they are not really a part of the sentence. The sentence ends at the punctuation mark.

Unfortunately, STAR Transit handles this differently, their segments include the trailing space. We do not have any problem with that in the Editor because segments imported using the Transit Package Handler plugin do contain the trailing space. The contents of the SDLXLIFF are in line with the way segments show up in Transit.

The problem only occurs with autopropagation: the autopropagated segments lack the trailing space, which then needs to be added manually. 
Unfortunately, QA Checker is not able to detect this, it only detects punctuation differences or added spaces, but does not check if the target segment has the same number of spaces at the end as the source segment. This could possibly be done using a Regex rule in the QA Checker. Does anyone have a suggestion for such a Regex?

Another option would be to enhance the options for the Autopropagate function to issue a warning if an autopropagated segment lost its trailing space. I may post another IDEA!

Walter 

Walter Blaser 

Walter Blaser said:

Does anyone have a suggestion for such a Regex?

I think this regex will make it:

Regex source:

\S( +)$

Regex source:

\S$1$

Thanks a lot Jesús 

The regex works fine. I included it in a QA profile that concerned translators can import. 
This will at least allow us to spot the faulty segments and add the missing space