| Pages in topic: [1 2] > | Creation of TM out of 2 TMs with different target languages Thread poster: avsie (X)
|
|---|
avsie (X) 
Local time: 11:52
English to French
+ ...
Hi everyone,
I'm sorry, my title doesn't really suit my actualy question!
We have 2 TMs: EN>SP and EN>IT. They pretty much have the same source segments (at least 75% of them, I'd say).
Now, we would like to have a IT>SP (or SP>IT) TM.
Is there any way to do this, if at all possible? We are using Trados (but Wordfast can also be an option)
Many thanks!
Marie-Claude ... See more Hi everyone,
I'm sorry, my title doesn't really suit my actualy question!
We have 2 TMs: EN>SP and EN>IT. They pretty much have the same source segments (at least 75% of them, I'd say).
Now, we would like to have a IT>SP (or SP>IT) TM.
Is there any way to do this, if at all possible? We are using Trados (but Wordfast can also be an option)
Many thanks!
Marie-Claude ▲ Collapse
| | | | | Anything is possible... | Sep 5, 2005 |
At first, I didn't read the post thoroughly enough, so I gave a wrong answer. See my new posting below...
 Jørgen
[Edited at 2005-09-05 10:50]
[Edited at 2005-09-05 10:51]
| | | | avsie (X)
Local time: 11:52
English to French
+ ...
TOPIC STARTER | You might not see it... | Sep 5, 2005 |
But for us it's actually very important at the moment
We usually translate only from English, but at the moment we have many big projects in Italian that we need to translate into Spanish. You can imagine that having a IT>SP TM would save time and money
I'll see if I can figure out a trick and I'll surely try out your suggestion ... See more
| | | | | Misunderstanding... | Sep 5, 2005 |
Sorry for not reading your post throughly...
I though you meant something else.
I'll think again and try to come up with a solution.
 Jørgen
| | |
|
|
|
avsie (X)
Local time: 11:52
English to French
+ ...
TOPIC STARTER
No worries, Jørgen!
I know, it's quite an unusual request, I've been puzzeling with it over the whole weekend and I can't come up with anything!
Maybe it's just not possible
| | | | Robert Tucker (X)
United Kingdom
Local time: 10:52
German to English
+ ...
| Align translated texts? | Sep 5, 2005 |
I don't know of any tools which currently exist to edit the tmx's directly as you require – but then I don't know much about Trados, DéjàVu, Wordfast or any of the other Windows only tools.
If none of these tools provide a solution then possibly one could consider aligning the Spanish and Italian texts into a tmx – I do know of tools that exist to do this. Whether you used the actual translated texts you created or extract the texts from the tmx's would depend on the complexit... See more I don't know of any tools which currently exist to edit the tmx's directly as you require – but then I don't know much about Trados, DéjàVu, Wordfast or any of the other Windows only tools.
If none of these tools provide a solution then possibly one could consider aligning the Spanish and Italian texts into a tmx – I do know of tools that exist to do this. Whether you used the actual translated texts you created or extract the texts from the tmx's would depend on the complexity of the formatting of the translated texts, I would guess. ▲ Collapse
| | | | avsie (X)
Local time: 11:52
English to French
+ ...
TOPIC STARTER | I thought of that... | Sep 5, 2005 |
I thought of that, but we are dealing with TMs that contain more than 75 000 units, coming from thousands of texts. So that wouldn't be possible
I thought of opening the TMs in Excel and sort of try to do an alignment with the units that are common to both TMs, but due to the size of the TMs, it would also be a fastidious job...
I appreciate your help ... See more I thought of that, but we are dealing with TMs that contain more than 75 000 units, coming from thousands of texts. So that wouldn't be possible
I thought of opening the TMs in Excel and sort of try to do an alignment with the units that are common to both TMs, but due to the size of the TMs, it would also be a fastidious job...
I appreciate your help ▲ Collapse
| | | | Hynek Palatin
Czech Republic
Local time: 11:52
Member (2003)
English to Czech
+ ...
| Database application | Sep 5, 2005 |
It's definitely not impossible.
There are probably special TM tools to do this, but I don't know any. Maybe somebody will suggest some of them later.
You can use a database application to create the TM. First export the two TMs, import them in a database, run a query to create the required language pairs, export the result of the query and import into a new TM. It should be quite easy for an experienced database user.
Also, any programmer should be able to ... See more It's definitely not impossible.
There are probably special TM tools to do this, but I don't know any. Maybe somebody will suggest some of them later.
You can use a database application to create the TM. First export the two TMs, import them in a database, run a query to create the required language pairs, export the result of the query and import into a new TM. It should be quite easy for an experienced database user.
Also, any programmer should be able to write a very short program or script to merge the two exported TMs into the new one. ▲ Collapse
| | |
|
|
|
Robert Tucker (X)
United Kingdom
Local time: 10:52
German to English
+ ...
| Well maybe not quite impossible | Sep 5, 2005 |
Marie-Claude Falardeau wrote: I thought of that, but we are dealing with TMs that contain more than 75 000 units, coming from thousands of texts. So that wouldn't be possible
I can believe that a single search and replace with a text of 75000 lines would take 10 minutes with a 2GHZ Celeron and that getting the result you want with the 75000 entries might be an hour's work, but not that it would be totally impossible. (I'm not talking about going through segment by segment.)
However, if you are competent with databases, exporting the tm to database as indicated above may well be the better, faster solution.
[Edited at 2005-09-05 12:58]
| | | | | | avsie (X)
Local time: 11:52
English to French
+ ...
TOPIC STARTER
Robert Tucker wrote:
I can believe that a single search and replace with a text of 75000 lines would take 10 minutes with a 2GHZ Celeron and that getting the result you want with the 75000 entries might be an hour's work, but not that it would be totally impossible. (I'm not talking about going through segment by segment.)
I'm currently giving a try to the database solution, I think I have an idea. If it works, I'll let you know
| | | | Gerard de Noord
France
Local time: 11:52
Member (2003)
English to Dutch
+ ...
| It's possible | Sep 5, 2005 |
You could try to use Excel in combination with Wordfast. Open your Wordfast EN>SP TM in Excel. In the second step of the Excel Text Import Wizard you should select {none} as Text qualifier and not ". Also, you shouldn't open the Wordfast TM in Excel by right clicking and selecting Open with. This can bypass the wizard.
In Excel delete the TM header and all columns but EN and SP. Now your column A is EN and your column B is SP. Copy column A to column C so it is EN too. Hide column B... See more You could try to use Excel in combination with Wordfast. Open your Wordfast EN>SP TM in Excel. In the second step of the Excel Text Import Wizard you should select {none} as Text qualifier and not ". Also, you shouldn't open the Wordfast TM in Excel by right clicking and selecting Open with. This can bypass the wizard.
In Excel delete the TM header and all columns but EN and SP. Now your column A is EN and your column B is SP. Copy column A to column C so it is EN too. Hide column B and C. Click once on the first cell to be translated without selecting the text. Open Word too and Reorganise your EN>IT Wordfast TM after making the right settings. Set your Fuzzy threshold to 100 and use Tools/Translate to translate column A. In theory your Excel document will be trilingual afterwards (great for Reference search). Make a copy of the Excel document, sort it to get rid of the 25 % of no matches, delete all columns but the SP and the IT column and export the Excel file to a tab-delimited TXT file.
If you're still with me:
Open the tab-delimited file in Word.
Search for "^p"
Replace with "<0}^p{0>"
Then search for "^t"
And replace with "<}0{>"
Repair the first and the last segment manually. (The segment delimiters don’t need the purple style.)
Create a new SP>IT Wordfast TM and make all settings you want. Clean up the document and you’ll have the TM you want.
It looks like a lot of work but this is how you can create a 50,000 TU TM in 30 minutes. This is how I juggled with the Microsoft Glossaries. (I now have DE>NL and FR>NL Microsoft Glossaries too).
Regards,
Gerard
[Edited at 2005-09-05 22:57] ▲ Collapse
| | |
|
|
|
Robert Tucker (X)
United Kingdom
Local time: 10:52
German to English
+ ...
| Further suggestion | Sep 6, 2005 |
I think what I would try taking the database route would be to:
Save the tm's to database
Delete all columns except the Spanish column in the English/Spanish tm (and, of course, all the columns except the Italian column in the English/Italian tm)
Export the database tables in text format, so that you now have two text files; one with an Italian translation of a segment on each line and the other with a Spanish translation of a segment on each line
... See more I think what I would try taking the database route would be to:
Save the tm's to database
Delete all columns except the Spanish column in the English/Spanish tm (and, of course, all the columns except the Italian column in the English/Italian tm)
Export the database tables in text format, so that you now have two text files; one with an Italian translation of a segment on each line and the other with a Spanish translation of a segment on each line
Then either:
Open the two text files in a text editor with numbers down the side for each line (e.g. Bluefish) so that the files can be checked line by line against each other and when satisfied that you have equal numbers of lines in each file and that each line in one file corresponds correctly with the line of the same number in the other file combine the files into a tmx using:
Frankenstein (Java based) www.wilandra.com/~hans/
aligner.py (Python based) www.omegat.org/omegat/resources.html (Aligner utility)
Heartsome TMX Editor (Java based)
Or:
Combine the files first (making up any difference in numbers of segments by inserting dummy segments if necessary) with one of the above combiners
Edit the resulting tmx with you favourite tmx editor
See also:
www.proz.com/topic/33804 ▲ Collapse
| | | | avsie (X)
Local time: 11:52
English to French
+ ...
TOPIC STARTER | Many thanks to all!! | Sep 6, 2005 |
Thanks Gerard and Robert, I'll look into it today. What I tried yesterday didn't work, but these suggestions are very helpful.
Thanks for taking the time to help, I appreciate it
| | | | | Heartsome TMX Editor | Sep 10, 2005 |
Marie-Claude Falardeau wrote:
We have 2 TMs: EN>SP and EN>IT. They pretty much have the same source segments (at least 75% of them, I'd say).
Now, we would like to have a IT>SP (or SP>IT) TM.
Heartsome TMX Editor ( http://www.heartsome.net/EN/tmxedit.html ) does what you want.
The process is this:
1) Combine the TMX files using the "Merge TMX Files" from Tasks menu
2) Use the "Reorganize File" option to place segments with the same EN entry in one row. See details at http://wiki.heartsome.org/hswiki/Tips_and_tricks#reorganizing where you have an example.
3) Create a TM database using the editor and import your file.
4) Export the TM database and select to export only ES and IT.
Following these steps tou will have the desired ES->IT in less than 5 minutes.
Hope this helps,
Rodolfo
[Edited at 2005-09-10 19:40]
| | | | | Pages in topic: [1 2] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Creation of TM out of 2 TMs with different target languages | 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 » |
| | Protemos translation business management system | Create your account in minutes, and start working! 3-month trial for agencies, and free for freelancers!
The system lets you keep client/vendor database, with contacts and rates, manage projects and assign jobs to vendors, issue invoices, track payments, store and manage project files, generate business reports on turnover profit per client/manager etc.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
Translation news related to SDL Trados
