Detecting language please wait for.......
I am sure that everybody knows this problem: you get a Word or Pdf original and the arrangement or layout of the text is messed up because your target sentences are longer than the text blocks or columns provided by the original document, so line breaks are inserted. Page headings are also misplaced some times.
Is there an App or trick or anything that can be done to solve this problem?
Of course not, you need to reformat the target on your own. You can however try not to be much longer as the source.
As for PDF, the only way to maintain the layout is to get the source document, from which the PDF has been created. Any conversion will alter the layout in more or less dramatical way, so there always will be a lot of work with PDF.
You could perhaps run a pseudotranslate before you translate so you at least have some idea of what will need to be changed. Then convert your PDF and edit the Word file so the necessary DTP work is complete prior to translation, or even have someone else do it while you're translating.
Then if needed pretranslate the "fixed" word file from your TM and hopefully have nothing else left to work on.
Thanks for your replies. That's unfortunate. My client decided that they will be taking care of that problem. I proposed that we could do it for a very reasonable amount but they didn't accept, which frees me from that ordeal, but I wish I could help them out.
Unfortunately there is no magic on reformatting translated documents. This can be a quite demanding work, for which you need to know a lot about how formatting work. That's fairly more than just changing font sizes...
I really wish someone made an app for this.
There are many: InDesign, Word, Powerpoint and so on. As any other app, they need to be used by a human being. Depends on how do you use these, you can have a lot or far less work with formatting. But formatting is a creative job, like painting or so - how should that be automated?
Well, the idea is to match the original format in the target file, I don't see much creativity there, just a downside of using digital documents. Creating the format yourself, that's a different thing.
Of course creating a new layout is much more demanding, but applying the correct formatting to an existing document without destroying the original style is quite demanding, too. As I said, it is not only changing font sizes...
And being annoyed by glitches, impossible to fix graphics, adding text boxes... That work is also cheaper than translation, which I also liked quite automated...
Yep. This is why it can be really demanding. However, if the document has been created with the possibility of creating versions of it in other languages, the reformatting task should just be recreating the TOC and the index plus maybe some polishing here and there. Not automated, though, as the automated process does not "see" the results. And a human eye does see layout problems quite fast.
Heh, "just" to match the original format... what could possibly be so hard on that... :) :)
I suppose, you have never tried that personally, right?That would explain these naive opinions ;-)
I suggest you try to do that with the very document you originally asked about. Just to see how much work would it be...
The issue is not formatting the original file, which I agree can be fun and is certainly creative, but to match your target file to the original after your CAT processed it and returned a bunch of problems that I have fixed many times. It's a software problem. This I don't find very creative.
No, it's NOT a software problem. That's why I suggested that you actually TRY to do that, to get some real-world idea what are we talking about. And indeed I meant to work on the TARGET file, not the source.
Or maybe if you can give some examples of problems which you think are a software problem?
Miguel Ríos said:It's a software problem.
If you mean it's a software problem because the conversion tools don't do a good enough job of representing the original PDF after changing this to Word despite not translating it yet, then yes I'd say you're right. It's an enormously complex problem because of the potential requirements to mirror something that might not have been created with Word in the first place, and may require features that are not even available in Word. I think you've trivialised the issue a little but essentially you are right... it's a software problem. Just not one we will ever even try to resolve in Trados Studio or with an app.
Perhaps you should look at other ways to handle a PDF. I wrote this a while ago and I think you need to handle them a different way rather than with an app since the developers have not improved or updated the solution they created for Studio... but the idea of extracting the text, translating and injecting it back in again might be a better approach for you:
You'd still need to do some DTP work if the text is longer, shorter or breaks in the wrong place once translated, but at least the layout would be good.
When you have the image of a car engine, for example, with lots of tiny labels in very specific places of the image and, after the conversion, said labels are all over the place. I can imagine a (not so complicated) software that could recognize the text around images and then automatically create flexible text boxes that could fit a longer target segment and you could easily move around.