Thank you to Daniel and Paul. So, am I to understand that there is no simple way to alter the segmentation of an xlf file in Trados Studio. Trados engineers, can you help with a simple fix? For example, MemoQ has a segmentation filter for this type of file.

Anne Fox

Is it a bilingual file - does it contain bilingual data (yet)? That is what would make it difficult.

(In general, if you can provide a sample file it's  a big help.)

Daniel

I thought I had attached the file in my original query. Apologies if not! Let me see how I add the file ...

Can you tell me how to attach or send the file?

Anne Fox 

Zip it, and then use the menu below to insert it:

I'm not sure if that file went. So here goes again

***File deleted***

Anne Fox 

It did come through.  I do have questions through.  I assume this is a sample you put together, and there seems to be three translatable segments in it.

The first has different source and target content... this seems incorrect and you should probably check with your client in case there are more like this.

The second only has source... this would be ok as you could deal with a file full of this even in 2019.

The third is translatable but the content is a name so probably won't get translated anyway.

Which is representative of the whole file?

Hi. This is a sample file, yes. But representative of the type of xlf I receive from this client. Usually French source to be translated into English. Usually one segment with the title, then a huge segment with the body of the text. When I import this type of file to Studio, all the text for translation (bar the title, which you see in the first segment here), is in one big clunky segment. That huge segment is my problem.

I attach another file as an example. Please delete after you look at it.

***File deleted***

Anne Fox 

I will delete the files from your posts.  But to show you how to do this I anonymised your file and then went through the process in the video below.  I just reinstalled afresh all my applications and seem to have some issue with our internal SSO so forgive me the errors at the end... but I used pseudo-translation to at least demonstrate how this would work in practice.

Thank you Paul. I appreciate your time and help. I'll try to follow all that in the morning when I'm fresh! But, is there something I could suggest to my client in terms of formating/creating their file that would avoid the need to do all this? Best, Anne

Thank you Paul. I appreciate your time and help. I'll try to follow all that in the morning when I'm fresh! But, is there something I could suggest to my client in terms of formating/creating their file that would avoid the need to do all this? Best, Anne

Anne Fox 

Anne Fox said:

But, is there something I could suggest to my client in terms of formating/creating their file that would avoid the need to do all this?

Yes... they could stop using CDATA dumps.  But this would involve work on their end they may not be prepared to do.  Dumping as CDATA is easy for them... tricky for the translator.

Bit they could also ensure that the source is always provided in the target element.  If they did this you could avoid the first part of the video where I put the content from the source into the target.  That way the files would be ready for you and you could simply create your projects with the custom XML and translate as normal.  This should be simple for them and would definitely make your life easier.