Maybe it's possible to load only locale define on common.properties
nicolas
Le 07/05/2015 00:08, Christian Geisert a écrit :
Hehe, this has already been discussed when switching to from .properties
to .xml
See https://issues.apache.org/jira/browse/OFBIZ-1442
And I was against this change
Hehe, this has already been discussed when switching to from .properties
to .xml
See https://issues.apache.org/jira/browse/OFBIZ-1442
And I was against this change (with size as one argument against it) ...
but meanwhile I like this format ;-)
Christian
Am 06.05.2015 22:57, schrieb Michael
- Posting in dev because I have no permission for commits -
Yes, thought about that recently when doing the german translations.
The drawback is that you have no overview about missing translations.
It's easy to go through the files as they are now and add missing
translations.
I quick
Except that byte site is not equal to in-memory site. Java doesn't use
utf-8 internally to store character data. I believe each char is 4
bytes in size. Then you have all the extra string objects with their
char array pointing into the large shared blob.
On 05/06/2015 03:57 PM, Michael
To be clear:
Yes, the entire file is parsed - which means it will reside in memory
during parsing - but only the requested language is kept in memory.
If file size becomes a problem, then we could convert to SAX parsing of
UI label files.
Adrian Crum
Sandglass Software