Detecting language please wait for.......
I need to pretranslate documents to share the pretranslated files in their native format (mostly Word, Excel and PPT) with translators who do not have SDL Studio but still need to take advantage of all the previously translated content. In many cases, when Studio finds matches for segments with tags and inserts them in the target, the tags disappear in the target segments and then have to be added manually one by one with Quickplace. Is there any way to automatically reinsert in one go all the tags from the source to the target in those pretranslated segments where they have been lost (it happens often with hyperlinks but even though I have changed settings to extract and always process hyperlinks, it´s still a problem). Any suggestions on how to save post-editing time after the pretranslation in this scenario? Thank you.
Beatriz de Vega
There are no features to automatically add missing tags into existing translations. Nice idea though and I guess with the improvements in AI we could see this sort of thing soon.
I am interested in wheter these are 100% matches without the tags and ifthey are why the TM is returning them like this.
Hi Beatriz de Vega
As Paul mentioned, it sounds like your TM contains 100% matches that do not contain tags and this is why the tags disappearing. Unfortunately Studio ignores tags when matching (you can change this though) and there is no function to automatically copy over missing tags. Right now the only solution I have is to run the Tag verification to find the missing tags for you and then copying them over by hand. Having said that, it would be really easy to implement a plug-in that simply copies over all missing tags from source to target with a batch task. However it wouldn't be possible to place them in the correct positions without parsing the target segment, but you could implement so that the tags appear at the end of the segment.
That´s what I thought. It would be nice though. it´s quite frustrating when Studio offers you a suggestion of pretranslated text but unfortunately you lose the hyperlink, format (italics...) or other tags and have to replace them manually. These are no 100% matches exactly because the tags are missing. Otherwise they would be. For example, it could be the names of acts and regulations. In the original translation they did not include hyperlinks to the pages where they are posted but in the new text to translate they include them. Thanks for your quick reply.
Yes, that´s what happens. I apply the same solution, run the Verifier, apply a Display Filter to only show the segments with messages for missing tags and manually Quickplace them. The plug-in you suggest will be helpful. I understand tags would not be placed correctly but at least if would be easier to just move them around if they are already in the target segment. The plug-in should only copy the missing ones and there should be an option in the Display filters to only display those segments (with misplaced tags) so that you can manually place them in the correct positions without having to go over all the segments with tags. Thanks your for your help.
Beatriz de Vega said:The plug-in you suggest will be helpful.
I'm curious. Why would having the tags added to the end f the target segment be helpful? How would it be faster to place them than by taking them over from the source?
I just tried both ways with hyperlinks (by misplacing some at the end of segment). Current method: you need several clicks in order to: select the target text, right click and Quickplace (and if there are several tags, chose the one you want) or alternatively select the target text and Ctrl+, and then Enter correct tag. Possible automatic placement of tags at the end of segments: no need to select anything, just pick, drag and drop tag in correct spot (saves about half the time and it´s much less tiring).
Perhaps using the mouse would help? You can also hold down the Ctrl key and just select the tags with the mouse. So select the word(s) in the target, or the location for the tag, then double click the appropriate tag in the source with the Ctrl key held down.
any sample file(s) ? -I mean, not screenshots but real files
Putting it back exactly correct point within Taget could be "impossible" I guess. -is it possible ? Wow Amazing.. what a nice AI
So, possible workaround is simply put it (them) at the Target - at head or tail location, that should be not so difficult.
Following is not exactly your situation but could give you a kind of imagination
Thanks Paul. With the mouse is a bit better. Still, I think just picking, dragging and dropping the tags if they were already in the segment seems easier. For now, we´ll just use the tip you gave me.
Hello Kelly, Yes, that´s exactly what I tried. I manipulated a few segments and put the missing tags at the end of the target segments and then I compared how long it takes me to just pick them from there and drop them where they should be (and how tiring it is it I had to do that 100 times in a row several times a day) vs copying them from the target (via Quickplace, Ctrl , or using the mouse as Paul suggested). I find that it would be easier and quicker. I can prepare a real file if you still want it. How do I get it to you? Thanks
amazingly..people thinks, I have suggested above tip for "Manually"..
nO, NO, No.. I have done it with codes, and what I did (for use it, not write codes) to put that missing Tags for entire file is "One Single Click" only.
I am tired with peopleThey are just crying for help, doing nothing..
Good Luck To You
(If and only If you do really want to get helped.. Do what you have to do).. or not.
Of course it is possible. By AI I simply mean the ability to understand the language and identfy where the tags should go. Machine Translation is a form of AI. So definitley not impossible.
yap, I hope so too.
I use an AutoHotkey script to place tags sequentially in a segment by pressing a single key (I use the left Windows key, but it's customizable). It's a great timesaver in situations like the one you describe.
The script is here:
Nice Nora Díaz, I'd forgotten you did this.