Import texts from Oracle database to Trados Studio and back after translation
Thread poster: Hana Netrefova
Hana Netrefova
Czech Republic
Jan 20, 2015

Hello colleagues,

I am a Trados Studio beginner. As a project manager I would like to ask you about some tips with importing existing data into Trados and back to the primary database. I am a user of Trados Studio 2014 Professional.

I have a large database (Oracle) with Czech/English/Slovak translations, with increasing number of Czech (source language) texts.

1. Firstly, I want to export the bilingual text pairs (CS-EN, CS-SK) into Trados Translation memory. Is the TMX the appropriate format for that?

2. Once a time I need to collect new arising Czech text in my database, import them into the Trados Studio, pretranslate them and send to my English/Slovak translators. With the source texts I need to keep also comments for translators (coming also from the primary database) plus some additional atributes necessary for the back-import into my database - eg. ID of the text record. How can I ensure that? Is there any format reccommended?

3. After I receive the translated texts from my translators, I'll update the Trados Translation Memory. Then, I'll export new Translation units via TMX format for updating my database. Still the question - how do I keep the record ID to know what record in my database I shall update? What if the database record is splitted in several Translation units according to segmentation rules?

My Oracle database remains to serve as a primary database. Is this suggestion of workflow with meeting my conditions reasonable? Would you ever faced issue like this one? I know the issue is very complex but for the first time I'am trying to describe it as a whole.

I very appreciate your opinion.

Hana Netrefova


Direct link Reply with quote
 

SDL Community  Identity Verified
United Kingdom
Local time: 16:02
English
The best solution... Jan 27, 2015

... might be to create a Translation Memory plugin provider directly to your Oracle database. This would probably be straightforward if you have a developer? You can do it using the API for Studio.

If not here's a quick answer to your questions:

Hana Netrefova wrote:
1. Firstly, I want to export the bilingual text pairs (CS-EN, CS-SK) into Trados Translation memory. Is the TMX the appropriate format for that?


Yes.

Hana Netrefova wrote:
2. Once a time I need to collect new arising Czech text in my database, import them into the Trados Studio, pretranslate them and send to my English/Slovak translators. With the source texts I need to keep also comments for translators (coming also from the primary database) plus some additional atributes necessary for the back-import into my database - eg. ID of the text record. How can I ensure that? Is there any format reccommended?


Use Fields and attributes held in the TMX against each translation unit. I'm not sure commenting on the TUs is supported out of the box so it would require development work to ensure they could be seen in the TM Results window. So using fields is probably the easiest way to go and then they could be seen in the TM results window too. I'd be interested to hear anyone elses opinion on this one?

Hana Netrefova wrote:
3. After I receive the translated texts from my translators, I'll update the Trados Translation Memory. Then, I'll export new Translation units via TMX format for updating my database. Still the question - how do I keep the record ID to know what record in my database I shall update? What if the database record is splitted in several Translation units according to segmentation rules?


Programmatically probably not a difficult task, but if your Oracle database is based on paragraph segmentation and you are bringing in work based on sentence segmentation then you will have to provide something more complex. More and more this looks like the creation of a TM Provider plugin for Studio would be a better idea.

Hana Netrefova wrote:
My Oracle database remains to serve as a primary database. Is this suggestion of workflow with meeting my conditions reasonable? Would you ever faced issue like this one?


I think it is reasonable if you really need to retain the Oracle database in this format. But I would definitely consider the creation of a plugin to at least make these things easier to manage:

- import/export of Translation Units between systems
- lookup to the Oracle database directly
- write access to the Oracle database directly
- concordance searching via the Oracle database directly

Regards

Paul


Direct link Reply with quote
 

Selcuk Akyuz  Identity Verified
Turkey
Local time: 18:02
Member (2006)
English to Turkish
+ ...
comments Jan 27, 2015

SDL Support wrote:

Use Fields and attributes held in the TMX against each translation unit. I'm not sure commenting on the TUs is supported out of the box so it would require development work to ensure they could be seen in the TM Results window. So using fields is probably the easiest way to go and then they could be seen in the TM results window too. I'd be interested to hear anyone elses opinion on this one?

Paul


Hi Paul,

I tested adding comments as additional fields years ago, with Studio 2009. It failed, but as you know I am not a power user of Studio.

Source and target side comments are very important in my workflow as well, they usually include definitions by the client or my own comments. Unfortunately they all go to a black hole when imported into TM. AFAIK, none of the CAT tools available on the market store comments/notes in the TM (or TMX file).

That is where xliff files and a tool like xBench can be useful.


Direct link Reply with quote
 

Meta Arkadia
Local time: 22:02
English to Indonesian
+ ...
Databases Jan 27, 2015

Selcuk Akyuz wrote:
That is where xliff files and a tool like xBench can be useful.

But that's a third party solution. I tried to import a database in CafeTran, and add comments. It's possible, but I'm still figuring out how to turn it in a decent workflow.



Importing the Oracle DB with comments shouldn't be a problem, adding comments on-the-fly neither. However, I have no idea how I can be alerted when there's a comment in the database, and that's elementary.

Cheers,

Hans (still thinking)


Direct link Reply with quote
 

SDL Community  Identity Verified
United Kingdom
Local time: 16:02
English
I wonder if TM is the right place at all? Jan 27, 2015

Selcuk Akyuz wrote:

Source and target side comments are very important in my workflow as well, they usually include definitions by the client or my own comments. Unfortunately they all go to a black hole when imported into TM. AFAIK, none of the CAT tools available on the market store comments/notes in the TM (or TMX file).

That is where xliff files and a tool like xBench can be useful.


Hi Selcuk,

You're right that XLIFF is the way to go and if you keep the bilinguals then you always have them as a record. The problem I guess is being able to look up the XLIFF files and see the comment. Probably reasonably simple via the OpenExchange and I have many examples of companies who save all their translations as html on a central server so their translators can use them for better contextual matching. This would be easy to use and display the comments. Maybe it would be neat project for a developer wanting to make this a filebased solution too, so you build up your corpus and have ready access to the comments per segment in a new window as you work.

I think fields on the TU work well if the field is consistent, but if you have different comments for each segment, and different source and target comments, then the task of maintaining this level of detail starts to become counter productive very quickly.

How does Xbench provide for this?

Regards

Paul


Direct link Reply with quote
 

Selcuk Akyuz  Identity Verified
Turkey
Local time: 18:02
Member (2006)
English to Turkish
+ ...
xBench Jan 28, 2015

SDL Support wrote:

How does Xbench provide for this?


Unfortunately I can not test it now, my license for the new version expired and the free version does not search in the other fields.

Selcuk


Direct link Reply with quote
 

Meta Arkadia
Local time: 22:02
English to Indonesian
+ ...
Suggested Workflow Jan 28, 2015

1) Extract the relevant language pair from the Oracle DB
2) Filter on comments (and other relevant information)
3) Convert segments without comments to TMX file TM1
4) Convert segments with comments to TMX file TM2
5) Keep the file with the segments with comments in database format

The translator can leverage TM1 and TM2 as usual. If TM2 is docked in a way it’s always visible, it will indicate if any comments (etc.) have been added. He can then switch to the database, and read any comments. He can also add comments. In CafeTran, you wouldn’t need any other applications, in other CAT tools, you may want to use a (free) database viewer/editor, rather than a paid app that would require an extra conversion at that (I doubt if XLIFF is the right format for this kind of job anyway). He can then return the finished XLIFF file plus the updated database to Hana.

Cheers,

Hans


Direct link Reply with quote
 
Hana Netrefova
Czech Republic
TOPIC STARTER
Comments saving Jan 28, 2015

Hi Paul, Selcuk, Hans,
thank you for your answers. However, in the meantime between my post and this discussion, we decided, according to our workflow needed, to use the xliff format. It seems to fit the best our needs.

We can join the procedure of the primary creation of the TM from my translated texts (#1) and collecting the non-translated texts from the database (#2) in one unique export method.

There are still some questions I am not sure how to solve: how to keep the obligatory attributes coming from my database (ID, sysid, original comment) with the translation segment/unit. My idea is that the translators are working with the Studio Packages only (with all the necessary info).

What I mean with the “original comment”: e.g. the Czech word “Predmet” is translated differently in various contexts (“Subject” in the context of “e-mail subject”; “Course” in the context of “course taught at school”). This info with the original comment comes with the source text from the database. I’m thinking about “source text comments” within the xliff file (the translator can see them in the editor view) and about the “Custom Fields” in the TM, but I don’t know how to link them together. The other comments (e.g. target language comments made by the translator) are not crucial for me.

Thank you for your suggestions.
Hana


Direct link Reply with quote
 
Hana Netrefova
Czech Republic
TOPIC STARTER
Minimum set of elements/attributes Jan 29, 2015

I am still thinking about that, it seems to me, I would need to know the minimum set of elements and attributes (XLIFF or XML) that the Studio can recognize and interpret in the right way. This set would describe my obligatory fields I need to import to the Studio and export back to the database.

* ID, source text unit, target text unit, comment to the source text, date and user of the last change of both source and target texts

It can’t be so difficult. Is there any clue how to find it out? Maybe I have to go through http://docs.oasis-open.org/xliff/xliff-core/xliff-core.html
but there’s too much redundant information for me.

If you have any nice idea, thank you for sharing.

Hana


Direct link Reply with quote
 


To report site rules violations or get help, contact a site moderator:


You can also contact site staff by submitting a support request »

Import texts from Oracle database to Trados Studio and back after translation

Advanced search







WordFinder
The words you want Anywhere, Anytime

WordFinder is the market's fastest and easiest way of finding the right word, term, translation or synonym in one or more dictionaries. In our assortment you can choose among more than 120 dictionaries in 15 languages from leading publishers.

More info »
TM-Town
Manage your TMs and Terms ... and boost your translation business

Are you ready for something fresh in the industry? TM-Town is a unique new site for you -- the freelance translator -- to store, manage and share translation memories (TMs) and glossaries...and potentially meet new clients on the basis of your prior work.

More info »



Forums
  • All of ProZ.com
  • Term search
  • Jobs
  • Forums
  • Multiple search