Trados Studio 2009: "Apply Translation" strips formatting of target text Thread poster: Harold Abilock
|
|---|
I am a Japanese-English translator. When translating, I sometimes apply formatting to the target English text that does not exist in the source Japanese text. For example, consider the following pair:
Japanese: 椅子１は、座席１０と、背もたれ１１と、足１２とを備えている。
English: A chair 1 comprises a seat 10, a backrest 11, and legs 12.
Note that the numeric symbols are boldfaced... See more I am a Japanese-English translator. When translating, I sometimes apply formatting to the target English text that does not exist in the source Japanese text. For example, consider the following pair:
Japanese: 椅子１は、座席１０と、背もたれ１１と、足１２とを備えている。
English: A chair 1 comprises a seat 10, a backrest 11, and legs 12.
Note that the numeric symbols are boldfaced in the target E text but not in the source J text. If I translate the above sentence so that it is stored in my translation memory and subsequently translate another patent document with the same source J text, both the source J text and the target E text (with the boldface tags properly included) appear in the Editor search window.
However, when I Apply Translation (Ctrl-T), the target E text is brought into the translation document as plain text without the boldface formatting. Is there anyway to set Trados 2009 so that it does not strip the formatting? Or do I have to go back to using Trados 2007 (which does not strip the formatting)?
Any advice would be much appreciated.
Best regards,
Harold ▲ Collapse
| | | | | Not sure why... | Apr 8, 2011 |
Hi Harold,
I tried to repro your problem but as you can see here the tags are recalled correctly:
Maybe you have an alternate translation that doesn't have the tags in your TM and then you are applying the tranlsation against this one?
If you want to share your TM and your file I'd be happy to take a look?
Regards
Paul... See more Hi Harold,
I tried to repro your problem but as you can see here the tags are recalled correctly:
Maybe you have an alternate translation that doesn't have the tags in your TM and then you are applying the tranlsation against this one?
If you want to share your TM and your file I'd be happy to take a look?
Regards
Paul
[email protected] ▲ Collapse
| | | | Harold Abilock 
Local time: 07:26
Member (2018)
Japanese to English
TOPIC STARTER | Looks like there's hope ... | Apr 9, 2011 |
Hi Paul,
Thank you so much for testing out my problem text. When I saw your response, I was much heartened to know that Trados 2009 does not strip formatting of target text as a rule. Now I just have to figure out why this is happening on my end.
First, I noticed that your file is .docx whereas test file is .doc. So I converted my test file to .docx, but the formatting was still stripped. I guess that the document type is not the problem.
Next, I deleted my... See more Hi Paul,
Thank you so much for testing out my problem text. When I saw your response, I was much heartened to know that Trados 2009 does not strip formatting of target text as a rule. Now I just have to figure out why this is happening on my end.
First, I noticed that your file is .docx whereas test file is .doc. So I converted my test file to .docx, but the formatting was still stripped. I guess that the document type is not the problem.
Next, I deleted my test project and created a fresh project based on the SDL template project. Still, the formatting was stripped. I guess the project settings are not the problem.
Next, I created a fresh translation memory (TM), whereupon "Apply Translation" worked properly and preserved the formatting. So, why did the new TM work properly but not the old TM? In both cases, I could see the proper tags clearly at the top part of the screen, but the formatting seems to get stripped when using the old TM.
The "old TM" was created by the "Upgrade Translation Memory" procedure, which created a Trados 2009 TM based on my Trados 2007 TM. When I use this converted TM, the formatting appears in the upper part but gets stripped when I "Apply Translation."
Any ideas? I will try re-running the "Upgrade Translation Memory" procedure using some variant approaches and get back to you. Meanwhile, let me know if you have any suggestions.
Best regards,
Harold ▲ Collapse
| | | | Harold Abilock
Local time: 07:26
Member (2018)
Japanese to English
TOPIC STARTER | But there is still a problem | Apr 9, 2011 |
Hi Paul,
I tried a variety of ways to upgrade my translation memory from 2007 to 2009, but it always results in a converted TM that ends up stripping the formatting (which is visible in the upper part of the screen) when "Apply Translation" is used to bring the target text into the translation document in the lower part of the screen.
I tried the following approaches:
* Import Trados 2007 WinAlign Files
First, the TXT files were converted to TMW files... See more Hi Paul,
I tried a variety of ways to upgrade my translation memory from 2007 to 2009, but it always results in a converted TM that ends up stripping the formatting (which is visible in the upper part of the screen) when "Apply Translation" is used to bring the target text into the translation document in the lower part of the screen.
I tried the following approaches:
* Import Trados 2007 WinAlign Files
First, the TXT files were converted to TMW files, then imported into a 2009 glossary. When this TM is used, the formatting of the target text (visible in the upper window) is stripped when brought into the lower part of the screen using "Apply Translation."
* Import Trados 2007 TM File
Same result.
I'm at a loss as to why this problem is ocurring. What is most peculiar is that the proper formatting of the target text (e.g., chair 1) appears in the upper Search window both when a newly created TM is used and when the converted TM is used. The problem is that the formatting is not transferred to the lower window (i.e., the translation doc) when the converted TM is used.
My converted TM has 70,000+ TUs, and most of them have formatting in the target text. Therefore, this problem cannot be solved manually.
Any advice would be much appreciated.
Best regards,
Harold ▲ Collapse
| | |
|
|
|
| An explanation.. | Apr 9, 2011 |
Hello Harold,
You didn't mention you were using work translated in Workbench, and then matching against the upgraded TMW, and this piece of information is important in this scenario. I took some advice from a developer this afternoon and he has explained what is happening as follows, and I would stress that this is only because you have added tags into the target that are not in the source and is only an issue when migrating from one tool to another.
What we have here ... See more Hello Harold,
You didn't mention you were using work translated in Workbench, and then matching against the upgraded TMW, and this piece of information is important in this scenario. I took some advice from a developer this afternoon and he has explained what is happening as follows, and I would stress that this is only because you have added tags into the target that are not in the source and is only an issue when migrating from one tool to another.
What we have here are “out of parity tags” – the target segment contains tags which don’t occur in the source. Usually, the “actual formatting” is taken from the document and injected into the translation proposal, by using tag alignment information between the involved segments. If the “tag alignment chain” (document to TU source, TU source to TU target) cannot be established, penalties will incur unless the tags can be resolved against the current filter’s quick tag list.
Now, native Studio TMs store a “tag formatting class” with target-only tags, such as “Bold”. Studio can, upon leverage, resolve such target-only tags against the quick tag list of the current filter. If there is some formatting only in the target segment, and there’s no alignment to actual document formatting, Studio will attempt to resolve the “dangling tags” against the quick tag list.
In a “native” Studio scenario, this use case works fine, as you have seen yourself and as I illustrated earlier, even across file types (provided that both filters define an actual formatting for “Bold” in their quick tags list).
The migration process however cannot currently “reinterpret” the original (TMW) tag contents to map them to a quick tag ID. That is, the quick tag information cannot be restored. If the quick tag IDs are unknown, or are not defined for the current filter, Studio can’t resolve them, and “anonymous tags” remain in the target segment, leading to penalties. This is what appears to happen in the migration case for segments like these.
I would stress again that if you had these tags in the source and the target when you first translated in Workbench then you would not have this issue, so this is something that probably only affects a minority of users who are at the inbetween stage of migrating between Workbench and Studio.
Out of interest I tested this scenario in another couple of tools and had a very similar result in one of them with only text being brought in despite the tags being present in the Translation Memory. It was interesting to note that this still got a 100% match though, so at least you have a benefit in Studio of there being a penalty so you can correct these as you build up your Studio TM to achieve what you want. The other tool provided no match at all. This was an interesting exercise, particularly when looking at the sort of issues you can experience when you CAT hop.
I hope this is a clear explanation anyway? It may be that we can do something to address this specific migration case in a future version, but for the time being I think you should either correct these in Studio as you work by editing the TM, or create a new TM that is updated as you work whilst the old upgraded TM is used for lookup only.
Regards
Paul ▲ Collapse
| | | | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Trados Studio 2009: "Apply Translation" strips formatting of target text | Anycount & Translation Office 3000 | Translation Office 3000
Translation Office 3000 is an advanced accounting tool for freelance translators and small agencies. TO3000 easily and seamlessly integrates with the business life of professional freelance translators.
More info » |
| | Trados Business Manager Lite | Create customer quotes and invoices from within Trados Studio
Trados Business Manager Lite helps to simplify and speed up some of the daily tasks, such as invoicing and reporting, associated with running your freelance translation business.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
Translation news related to SDL Trados
