Patrick Karban 

You could use a segmentation rule alongside the use of DeepL.  But based on the minimal explanation of what you are doing I think you might be better off just creating a parser rule in your embedded content rules for Excel. If you're not sure what I mean can you provide a small sample file (zip it and attach using the Insert menu below your reply) and I can show you.

Hey!

Would I do the segmentation rule by adding the translation memory alongside DeepL as the translation resource? How would I set that up for the HTML brackets?

Also, what kind of information are you missing? The job is about employee training for new product lines. They supply me with their learning content in Excel, but every cell is associated to an ID their software uses. So I wouldn't be able to change the Excel file at all, or it would stop working for them. Can you explain what you mean if you are not talking about changing the Excel file? A parser rule for embedded content doesn't ring a bell.

Thanks!

 Patrick Karban 

Patrick Karban said:

Would I do the segmentation rule by adding the translation memory alongside DeepL as the translation resource?

No.

Patrick Karban said:

Also, what kind of information are you missing? The job is about employee training for new product lines. They supply me with their learning content in Excel, but every cell is associated to an ID their software uses. So I wouldn't be able to change the Excel file at all, or it would stop working for them. Can you explain what you mean if you are not talking about changing the Excel file? A parser rule for embedded content doesn't ring a bell.

I just wanted an example Excel file so I could have shown you exactly how to solve your specific problem.  But fair enough... you can do it yourself.  This article is 9-yrs old now but does explain the principle.  Today it's somewhat better:

https://multifarious.filkin.com/2013/08/21/embedded_content_excel

If you get stuck provide a SMALL sample file.  You can attach it to your reply by zipping it first.

Hey!

Thank you for the reply. I don't think the article helps me out here, since it is only about formatting HTMML tags to be Trados tags? My issue is that I am having up to 10 line breaks in a single Excel cell and when loading it into Trados, those 10 lines are formatted to only being a single line. I would like to have a new segment each line break. 

I attached a small sample of the content I am getting. Maybe that make the issue a bit more clear. I deleted everything besides the source column, so Trados only focuses on that part. I would export it after the translation and copy the translation into the target column of the initial file once it is finished. However, since Trados is reformatting the big cells to only being a single big line, I am not able to use Trados for the job at all.

Let me know if you need anything else!

PatrickKey Stories February 2023 translation sample.zip

Patrick Karban 

I created a short video explaining why the article will help you out... I hope this helps:

Hey!

Thank you for making that video. I followed the steps and entered all of the separate rules for the tags. However, when I let it translate using my usual project template, the overall quality is a lot worse due to all the tags in the segments, I guess. Translating the cells in Excel one by one is a lot more manual work, but the translations end up being better by a long shot. Too bad that I might not be able to use Trados for this client. Is there something else we could try to improve the overall translation quality as well? The translation volume of this client is pretty high and I have yet to find an optimal way to make it efficient.

Thanks!

Patrick Karban said:

However, when I let it translate using my usual project template, the overall quality is a lot worse due to all the tags in the segments, I guess.

Patrick Karban 

I don't really understand this, what do you mean by your usual project template?  You just changed all the rules so that you use a project template with the new rules for all your projects using these files.  The tags are in there because they are html tags and they are inline.  You cannot avoid them based on the type of content you have.  If you exclude them all then the sentences will also break up into small chunks and I doubt you want that either.  From what you have written in reply I suspect you simply don't know how to work with tagged files... but I can't be sure because the problem may be you just don't know how to work with project templates.  I really don't know.

Maybe you can share a screenshot of what you see when you have finished translating a few segments, and show us the Excel filetype settings for your new project?  It's really hard to truly understand what you are saying and match that to what you should be doing.  The example file you gave me should be trivial to handle in Trados so I really don't know why you think you cannot use Trados for this.

Patrick Karban said:

Is there something else we could try to improve the overall translation quality as well?

I think we need to make sure you are handling these files properly first!  The translation quality is something that is 100% down to you and you surely know this.  So let's just make sure we're both talking and thinking about exactly the same thing!