Detecting language please wait for.......
I am frequently in a situation where translated text contains a lot of very short segments, which I would classify as junk (for lack of a better word). Some examples of this would be acronyms, fragments of words, misspelled words or perhaps names of physical units. Generally I try to avoid cluttering my TMs with such junk entries. They cause a lot of problems later when you pre-translate and then have to correct a lot of segments which were not supposed to be pre-translated, but were replaced by these short junk entries. So here is my question.
Is there any way to control for each segment if it will be committed to TM but still have SDL Studio run its Autopropagation routine? What I am looking for is perhaps a key combination that would allow me to confirm a segment, run autopropagation but to avoid committing the segment to TM.
The way I do it now is go back one segment and delete TM entry manually. This involves too much effort and clicking. I would love to have a better solution.
If there is a solution, I would very much appreciate your help. If there isn't, let's make this a feature suggestion!
Use display filer for displaying just these segments, copy all source to target, select all segments and put them on "Sign-off rejected", then lock.
Even if that seems to be more work, it is usually faster and certainly more reliable as any auto propagation. This way all the segments will appear as translated, but will NEVER get into the TM.
You could also instead change this setting when dealing with such segments:
Thank you Jerzy, I am aware of all these options and features and also use RegEx filtering on large files, however nothing would make me happier than to be able to make the decision on saving an entry in the TM individually. I hope such feature (shortcut) will be available in the future.
Post this as idea.