<i18n dev> localization of a java application

[Leo Jiang]

Hi Jean-Luc,

Any big company would manage the translation memory to help l10n work for their products. So the concrete workflow and 
tools have to be designed to match your own requirements.

In OpenJDK, we did it per several months to update the translation. About the translation quality, it's impossible for 
product team to verify the qualify, so you need to rely on your translation team and management.

Thanks,
Leo

On 03/22/2017 09:09 PM, THIRY, Jean-Luc wrote:
> Hi Leo,
>
> Thanks for your responses.
> Can you please tell me what your tools do ? If you manage exchanges with an agency in charge of the translation ? How often you send and receive translation requests ? How do you validate the quality of the translation and who is in charge of it in your team ?
>
> Best regards,
>
> Jean-Luc
>
> -----Message d'origine-----
> De : Leo Jiang [mailto:li.jiang at oracle.com]
> Envoyé : mercredi 22 mars 2017 09:40
> À : i18n-dev at openjdk.java.net; THIRY, Jean-Luc
> Objet : Re: <i18n dev> localization of a java application
>
> Hi, Jean-Luc,
>
> Yes, the problem we are facing is common, we have Oracle-specific tools and process to handle the data exchange with translation vendors. You need to build up yours.
>
> Thanks,
> Leo
>
>
> On 03/16/2017 10:07 PM, THIRY, Jean-Luc wrote:
>> Hi,
>>
>> I am a new subscriber to the i18n-dev list.
>> Currently I have to organize the localization of a big java application that has been developed for 15 years using 2 types of java resources to manage localization :
>>
>> -          .properties files used with resource bundles for common tags in forms, messages and java enums
>>
>> -          data collected in SQL tables for dynamic enums
>>
>> As the application is big (property files with more than a 1700 entries) and a big number of enums (over 500), I have to find tools to automate the localization process knowing that the application has numerous new features added each year, features that we need to keep localized as we deliver new versions of the application. Theses resources are located in different jar's in the project which makes the data retrieval of all the .properties tedious.
>>
>> The process I have in mind has to detect new entries in the property files or in the database, extract these new labels in files so that I can transmit the information to a localization provider which in return can send me back the files with the labels localized.
>>
>> I have been looking for information on standards and I found XLIFF 1.0 and 2.0 interesting to manage the exchange process with the localization provider and I found java APIs to translate a .properties in a XLIFF file.
>> The trouble is I have many things to code to have a full process operating so I have been looking also for other solutions.
>>
>> I found GNU getText well designed to manage localization of Linux projects.
>>
>> I would be interested in having more information on how to manage this knowing that you have exactly the same problem when you manage Localization of projects such as JDK or JRE. So how do you do this ?
>>
>> Thank's in advance,
>>
>> Cheers,
>> Jean-Luc
>>