On 28 nov. 07, at 19:21, Reiko Saito wrote:
Hi JC,
I understood your point.
To increase the leverage, we may be able to lower the lowest match-
rate,
or will use this tmx just for reference and search a certain string
as a file... what do you think ?
It is possible to use the sdf directly and not the PO file as source.
For that it is necessary to do a few things to the sdf file as I
described sometimes in the summer if I remember well. Simply put, it
amounts to that:
Basically the sdf comes as pairs of lines:
"string in English+meta data"
"string in target language+meta data"
The format for each line is very close to CSV so the idea is to:
1) convert each pairs to:
"string in English+meta data" "string in target language+meta data"
2) import that into OOo and to select the columns that correspond to
"string in English".
3) put that into a text file for use as source in OmegaT or anything
else with the TMX provided by Rafaella as reference.
All this is done with a text editor and a few regex search/replace.
Once the translation is completed, it is pasted into the "string in
target" part of the CSV file, the file is converted back to a 2 lines
format and the result is delivered.
Can anyone generate csv file from these tmx ?
It is very easy but now that I am seeing the contents, I wonder it is
is a good idea. Look at one <tuv>:
<tu tuid="47197">
<tuv xml:lang="en-US">
<seg>The query already exists. Do you want to delete
it?</seg>
</tuv>
<tuv xml:lang="ja-JP">
<seg>このクエリはすでに存在します。削除します
か。</seg>
</tuv>
</tu>
Obviously this is a full sentence and not a menu item. So maybe
Rafaella's idea of using Sun Gloss for glossary reference is better
after all ?
Anyway, just in case we need the conversion it is not a very complex
task to do that by hand with a few regex. I'll do it for the Japanese
group if you want.
Jean-Christophe
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
