On 19 juin 07, at 20:52, Rafaella Braconi wrote:
Hi Jean-Christophe,
thank you for verifying the file format provided.
Based on all discussions, I think that now we have 2 possibilities:
1) to provide you with 2 tmx files 1 for UI and 1 for OLH
(unfortunately splitting the tms OLH file into single components
it's really time consuming and also, there are some parts which are
shared among other modules so I think that having 1 OLH tmx is much
better). These tmx files will be based on all translated strings
which will be used for the 2.3 and which do not need any update
(what we flag as translated and reviewed)
2) to provide you 2 tmx files: 1 for UI and 1 for OLH with exactly
the same strings you received for translation as sdf files. This
tmx file will contain what we call new and changed strings, exactly
the same you have in the sdf file. This way you would be able to
see - if I understand it correctly - the previous translations
directly in the OmegaT window.
Just let me know what works for you best.
I am not sure I understand the second option. Would that be a "pseudo
translated" file that you turn into TMX, like I did myself ?
Or would that be the set of modified segments _before_ they were
modified along with their correct translation ?
The minimum I need is this set of modified segments _before_ they
were modified, ie, the state of the (limited) corpus at the time of
2.2.1.
If that is possible I'll take that. If not, I'll take option 1 above.
Cheers,
JC
Rafaella
Jean-Christophe Helary wrote:
On 19 juin 07, at 18:28, Rafaella Braconi wrote:
Hi Jean-Christophe,
attched you can find a TEST tmx file created out of the Calc
Help project. This test file should be used to verify if you can
read it, use it, etc. Once you verify the file and everything is
fine (let's keep the finger crossed) I will proceed doing a
full extraction of tmx files out of the database.
Just let me know the results.
It loaded properly. I see that the whole package eats up a lot of
memory:
PID COMMAND %CPU TIME #TH #PRTS #MREGS RPRVT RSHRD
RSIZE VSIZE
7123 java 0.0% 9:30.89 18 523 333 108M 54.6M
126M 2.56G
And I noticed matches (even low level ones) early in the file. So
it seems to work.
The French file has not been split by application but if it were
possible to get a split TMX that would be make the set more
useable in the future.
Regarding the escape sequences:
I suppose you created the TMX directly from the material that
gives you the .sdf files ? So basically the TMX seems to contain
the exact same level of escaping. Let's say it keeps the file at
an arbitrary "level 1".
When I use the "level 1" .sdf to create a po file with oo2po, the
process escapes the escape sequence character "\" so I get to a
"level 2" escaping. "Level 2" contains one more "\" character for
each escape sequence group.
So basically, my .po file and your .tmx file do not match at the
number of "\" for escape sequences.
I had the same issue with the oo2po + po2tmx process (see the
other thread on the list).
the .sdf is at "level 1", the oo2po process puts the resulting .po
at "level 2", adding an extra "\" to each escape sequence and the
po2tmx process removes that extra character to lower the escaping
to "level 1", which makes it incompatible with the .po source file.
In the case of your internal conversion I'd say it is bad luck
for me, but in the case of the translate-toolkit processes I
think that should not be the expected behavior: the .tmx should
contain the same contents as the .po file. I think there is a bug
in the po2tmx process somehow.
Regards,
Jean-Christophe
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
Jean-Christophe Helary
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]