CafeTran is a great piece of software, I am really enjoying it.
Anyway, I have one question. In Trados I had an 'Update TM' option - when translation is over, you can read through your segments, change them as you want and then UPDATE your TM, just in order to 'insert' the changes you have just made to your target segments.
In other words, you 1) translate a file, 2) proofread it inside your CAT tool right after the translation process, and 3) update your TM afterwards.
It helps when you have identical files - you just open them, feed them to this updated TM, and in this case you don`t have to proofread all of these identical files one by one, since you updated your TM once, and then applied this updated TM to these new files.
is there any such 'Update TM' option in CafeTran? Thank you!
Thanks Jean for your comments.
I’m not sure I understand what you are missing.
What is your TM setup? Do you use TR?
If you use a Project TM for each project, just make sure you have “Keep newer duplicates” enabled in the TM options. Of course, if you use “Keep all duplicates” instead, all different translations of one source are kept.
“Keep newer duplicates” effectively overwrites and kills older duplicate source segments stored in a given TM.
In the end of each project, just export the final TM as a TMX or (that’s what I do=) use Memory > Import > Import segments from project. This way, you are sure that all project segments are stored in the project TM.
No need to fiddle around with TMX Editors to merge small files, just put them in a folder, and use CafeTran’s menu action Memory > Open Memory folder. Set the appropriate TM options, and run any relevant task from Task > TMX memory. You can then save the merged memory.
Doesn’t Task > TMX Memory > “Remove TMX units with duplicate source segments” remove the older duplicates as required?
A question to Igor though: with this task, in case of duplicates, which one is kept, the newest or the oldest?
Quite an old thread. Here I read about "updating" a TM, what I want is "updating by overwriting and killing" older segments. I'm fed with having two/three or more similar translations for one source. And this is not a problem of CT alone. I understand that there are occasions where you would want two different translations for one source, but - at least in my case - this is quite rare. How can this be solved? I could of course save my final translation as TMX Memory but then I would still have to fiddle around with TMX Editors to merge them small files into one big file. Can't we have an option that does just that, overwrite and kill segments with same source and different targets with from earlier dates?
One method would be to select the memory you wish to update and go to Memory > Import > Import segments from project. If your memory is set to Keep newer duplicates, only the most recent segments will be stored and your TM will effectively be updated.
Another method is to use Project > Export and exchange > To TMX memory. This allows you to save another up-to-date TM from the current project.
PS: This reference document offers an overview of what each menu action does: CafeTran Espresso - Menu and Interface, which is also explained in the Help tips that are displayed on the bottom CafeTran's window (status bar).
Your TM gets updated with the changed versions of project segments automatically as you proofread. Also, if you wish, you can create an updated TM (or overwrite the existing Project TM) with your latest changes via Project > Export and exchange > To TMX memory... .