Bungeni
Document Actions
Previous:
3.13.
Glossary of Parliamentary Terms
3.14. How to generate translation catalogs
How to setup i18n scripts
due to a minor issue in one of the recipes used in buildout, we need to modify the generated i18n scripts to include the path to our xapian library.
so after running buildout you'll need to add an additional
'$buildout_directory/parts/xapian/lib/python'
to the list at the top of the bin/i18nextract script, substituting the full path for $buildout_directory. after which you can use the i18nextract script to generate a message catalog (pot file) for translation.
replace the lines
import lovely.recipe.i18n.i18nextract
if __name__ == '__main__':
lovely.recipe.i18n.i18nextract.main(['i18nextract', '-d', 'bungeni', '-s', '/home/undesa/devel/svn/bungeni-portal/parts/i18n/configure.zcml', '-p', 'src/bungeni.core', '-o', 'locales'])
with
import lovely.recipe.i18n.i18nextract
import getopt
argv = sys.argv[1:]
try:
opts, args = getopt.getopt(argv, "p:d:", ["path=", "domain="])
except getopt.GetoptError:
usage()
sys.exit(2)
domain = 'bungeni'
path = 'src/bungeni.core'
for opt, arg in opts:
if opt in ("-p", "--path"):
path = arg
if opt in ("-d", "--domain"):
domain = arg
print 'generate locales for domain: ', domain, ' in: ', path
print
if __name__ == '__main__':
lovely.recipe.i18n.i18nextract.main(['i18nextract',
'-d', domain,
'-s', '/home/undesa/devel/svn/test-bungeni-portal/parts/i18n/configure.zcml',
'-p', path,
'-o', 'locales'])
Generate i18n catalogs
First, extract the msg strings of workflow states run (WHY!!!!! This needs to be fixed)
./bin/python src/bungeni.main/bungeni/core/workflows/exctractwfi18n.py
Then, to generate the POT files run the following commands
./bin/i18nextract -d bungeni.ui -p src/bungeni.main/bungeni/ui/
./bin/i18nextract -d bungeni.models -p src/bungeni.main/bungeni/models/
./bin/i18nextract -d bungeni.core -p src/bungeni.main/bungeni/core/
The POT files will be created in locales directory in the package eg. for bungeni.ui the POT file will be in src/bungeni.ui/bungeni/ui/locales
You may get errors like the ones below but the catalogs are still generated
/home/bungeni/cap_installs/bungeni_install/bungeni/releases/20091029070249/src/ore.alchemist/src/ore/alchemist/introspector.py:22: SADeprecationWarning: the information_schema module is deprecated.
from sqlalchemy.databases import information_schema
Exception in thread Thread-1 (most likely raised during interpreter shutdown):
Traceback (most recent call last):
File "/home/bungeni/cap_installs/python25/lib/python2.5/threading.py", line 486, in __bootstrap_inner
File "/home/bungeni/cap_installs/python25/lib/python2.5/threading.py", line 446, in run
File "/home/bungeni/cap_installs/bungeni_install/bungeni/releases/20091029070249/eggs/ore.xapian-0.5.0-py2.5.egg/ore/xapian/queue.py", line 64, in __call__
File "/home/bungeni/cap_installs/bungeni_install/bungeni/releases/20091029070249/eggs/ore.xapian-0.5.0-py2.5.egg/ore/xapian/queue.py", line 54, in operations
<type 'exceptions.AttributeError'>: 'NoneType' object has no attribute 'Empty'
Merge old catalogs with new ones
To merge old translation catalogs with new ones, one needs have gettext installed
sudo apt-get install gettext
use the msgmerge tool to merge the old translations with the new ones for each PO files eg. for the en translation catalog for bungeni.ui
cd src/bungeni.ui/bungeni/ui/locales/
msgmerge -U en/LC_MESSAGES/bungeni.ui.po bungeni.pot