Simon,
Do you have a link or could you send this scripts to me ? Maybe we could help each other.
Our migration script helped us in a very simple manner, by parsing the old GSI into the new one, using three points of conjugation between them. This gaves us one final GSI that needed to be translated, that contains only the new lines and can me merged again in the new GSI, by doing the same operation with our script.
Of course, that this has failures, we need something that shows where's the difference and in which line (we're gonna do that tomorow), if the three point conjugation are the same, but the translation's different.
> and maybe they are not as useful if you translate using PO or other tools
I'm simply using OOoCalc for translation :)
//VD
Simon Brouwer wrote on 01/17/2005 08:06 PM:
Hi Vitor, all,
At 16:03 17-1-2005, you wrote:
Hi,
The portuguese team have develop one tool for GSI migration of 1.x versions to 2.0 version. It's perl based and it's pretty simple.
Link: http://ooo.paradigma.co.pt/work/vd/l10n/2.0_Tools/trans_conv_gsi2sdf.perl
You only need to change the file directions. It produces three files: 1) The Final SDF Converted. 2) The CSV that needs translation - the new lines that doesnt exists in the 1.x version 3) The CSV that contains the already translated lines.
NOTE: Some locale information have changed, please use search and replace. Example: the 1.x GSI file has the "pt-PT" as locale information, but now the 2.0 SDF has only "pt" information.
I have made some tools for working with GSI/SDF files as well. They are written in ANSI C.
They assume the GSI/SDF files have line pairs of English and the target language, with English as the backup language (extracted with e.g. -l en-US,nl=en-US) They are:
* migrate-gsi-sdf , This reads an SDF file (e.g. OOo 1.1.x) and makes a database of all the strings and their translations (if the "translation" is different from the en-US string). It then reads a second SDF file (e.g. OOo 2.0). For each line pair in that file, where the strings following the language code for en-US and the target language are the same (so probably not translated), it looks in the database for translation(s). If found, these are written into an output SDF file. That SDF file can then be reviewed, edited and merged back.
I found that as strings in the OOo2.0 SDF appeart to have been moved around and duplicated a lot with respect to OOo 1.1 GSI, this tool is quite helpful with the migration.
* comparegsi. This reads two GSI or SDF files and compares them. I find this tool very useful for updating the UI translations between minor versions, because it outputs just the line pairs of which the en-US string has changed, and the ones that did not exist in the older file. You can also run it on the extracted SDF file and the SDF file that you have edited, so that you get an SDF file containing only the updated lines that need to be merged.
* mkgloss, a simple tool that reads an SDF file and outputs a CSV glossary containing the line number in SDF, en-US string, target language string.
These tools may need some more work, and maybe they are not as useful if you translate using PO or other tools, but if anyone is interested to have a look at them just let me know.
Vriendelijke groet, Simon Brouwer.
>>> nl.openoffice.org <<<
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
-- Vitor Domingos Paradigma.pt
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
