| User | Thread poster: Hoshi-Trans TM from previously translated documents |
|---|
Hoshi-Trans
United Kingdom
English to Japanese
|
Hi,
is it possible to generate translation memory automatically from documents that had been previously translated in Word or Excel? (Source and target available)
Or is it necessary to create a new project, import source and paste the translation one by one?
Thanks in advance for any adivce
| | | |
Didier Briel 
France
Local time: 21:45
  Member (2007)
English to French
+ ...
| | What you need is an aligner | Mar 12, 2011 |
Hoshi-Trans wrote:
is it possible to generate translation memory automatically from documents that had been previously translated in Word or Excel? (Source and target available)
|
|
The process of creating a translation memory from existing translations is called an alignment, which is done with an aligner.
There are several open-source aligners.
See Other OmegaT resources and
Third-party software on OmegaT's website.
For Excel, it's generally quite easy to create a translation memory, since there is a one to one match between the source and target document.
Or is it necessary to create a new project, import source and paste the translation one by one? |
|
You could do it that way, but you should be faster with an aligner.
Didier
| | | |
Miloš Švantner
Slovakia
Local time: 21:45
German to Slovak
+ ...
|
I can recommend you LF Aligner, works on Windows, Mac and Linux as well.
It's really fast and accurate.
All you need to know you'll find in the LF_aligner_readme.
| | | |
Samuel Murray
Netherlands
Local time: 21:45
Member (2006)
English to Afrikaans
+ ...
| | Use an aligner | Mar 12, 2011 |
Hoshi-Trans wrote:
Is it possible to generate translation memory automatically from documents that had been previously translated in Word or Excel? |
|
Use an aligner, such as: http://bligner.org/
| | | |
Olaf Erikson
United States
| | Same situation but different | Jun 29, 2011 |
I am in the same boat except I've already got everything aligned perfectly. I've got thousands of segments sitting uselessly in Excel. I've tried what must be a dozen different software packages to convert the files but they inevitably turn my lovely Arabic-scripted Dari into complete gibberish despite supposedly being encoded in UTF-8. Any ideas?
| | | |
FarkasAndras
Hungary
Local time: 21:45
English to Hungarian
+ ...
|
Just generate a TMX from your data.
This is how I would do it:
Put the data in columns A and B of an xls. Put metadata (e.g. a reference to the source of the text) in column C. Each line of text can have different metadata. Copy-paste it all into a txt file. Save the txt file in UTF-8 encoding with Save as.
Use the TMX maker in the LF Aligner package to generate a TMX. This conserves the metadata as a text field.
Article:
http://www.proz.com/translation-articles/articles/3194/1/Converting-translation-memories-into-spreadsheets-and-vice-versa
Download: www.sourceforge.net/projects/aligner
As long as you save your txt as UTF-8, the Arabic characters should be fine in the TMX. The TMX will be encoded in UTF-8, and the text will essentially just be copied over to the TMX without modification. You can check this by opening the TMX in a text editor. Of course your problem might be in the program you're importing the TMX into.
| | | |
The translation workplace 
