Posts: 10
Joined: Fri Feb 20, 2009 9:36 am
Thu Aug 13, 2009 3:21 pm
Strings Export
Hi again!
Okay, so I'm experimenting with some DLL files in Lingobit. The files are part of a translation updates project.
Anyway, so I've been looking at the best way to translate them, and I've come up with this process:
1) Load English files into Lingobit using the .NET Binary/ResX/Res parser.
2) Import translated versions of the files, this auto-translates everything translated so far.
3) Run "Translate Duplicates" (this usually translated some existing strings)
4) Then, since we're a SDL Trados translation house, I use the File > Export, as TMX feature to export all the strings from the project.
5) Using a custom XSLT transform I've put together I then strip the file of the already translated segments and make the TMX file into a clean XML file, ready for translation in Trados.
6) Translate the file in Trados, update the translation memory.
7) Import the translation memory back into Lingobit, and voila, fully translated files!
However, for a couple of my files it appears that step #4, exporting the strings from Lingobit, isn't working as perfectly as I had hoped. For one of the files, the word counts I get match exactly, the second one is fairly close, but the last one is really far off.
Is the Export function meant to extract all strings? Or only certain ones? Are there any settings that control what is extracted?
Right now for that one DLL file with the most differentiating word counts, here's what I'm getting:
From the Lingobit Report: 4994 Total Words, 1571 Imported translations, 3423 new translations.
Then I use Export, using a custom INI file I created for Trados to view only the xml:lang="en" segments. I then Analyze the whole file in Trados and get:
2129 Total Words. And once I filter out the already translated items, I only get 778 new translations out of it.
Any ideas?
Okay, so I'm experimenting with some DLL files in Lingobit. The files are part of a translation updates project.
Anyway, so I've been looking at the best way to translate them, and I've come up with this process:
1) Load English files into Lingobit using the .NET Binary/ResX/Res parser.
2) Import translated versions of the files, this auto-translates everything translated so far.
3) Run "Translate Duplicates" (this usually translated some existing strings)
4) Then, since we're a SDL Trados translation house, I use the File > Export, as TMX feature to export all the strings from the project.
5) Using a custom XSLT transform I've put together I then strip the file of the already translated segments and make the TMX file into a clean XML file, ready for translation in Trados.
6) Translate the file in Trados, update the translation memory.
7) Import the translation memory back into Lingobit, and voila, fully translated files!
However, for a couple of my files it appears that step #4, exporting the strings from Lingobit, isn't working as perfectly as I had hoped. For one of the files, the word counts I get match exactly, the second one is fairly close, but the last one is really far off.
Is the Export function meant to extract all strings? Or only certain ones? Are there any settings that control what is extracted?
Right now for that one DLL file with the most differentiating word counts, here's what I'm getting:
From the Lingobit Report: 4994 Total Words, 1571 Imported translations, 3423 new translations.
Then I use Export, using a custom INI file I created for Trados to view only the xml:lang="en" segments. I then Analyze the whole file in Trados and get:
2129 Total Words. And once I filter out the already translated items, I only get 778 new translations out of it.
Any ideas?