Suggesting changes in transtions

[sprieto]

Hello everybody,

I'm a novice user of testlink, and I'm working in spanish translation for 1.7 version.
I'm using the current spanish strings in CVS repository against english_GB strings in CVS. The spanish strings is 16 months old and it hasn´t a lot of the new strings of 1.7 version, moreover changes in template names.
This make translation work very annoying and is too easy to make mistakes.

I'm thinking about it, and have one suggestion, that can do translation process more easy, but the first work will be a little bit boring. I suggest to break the strings file in many files, one for each tpl or php file to translate, and leave in strings.txt just call, with require_once function, all the files. I have little know about php and don't know if that is possible, but with this way, support the translation will be more easy. Every new template or file will have its own translation file, and only will require to add a require call to the file.

After the proccess /locale directiry would be:
strings.txt with the next content

Code: Select all

...
require_once("login.tpl.txt")
require_once("loginFirst.tpl.txt")
require_once("loginLost.tpl.txt")
require_once("mainPage.tpl.txt")
...

and in the directory, the next files:
login.tpl.txt
loginFirst.tpl.txt
loginLost.tpl.txt
mainPage.tpl.txt

Currently, there are 163 files (count from comment lines of strings.txt file).

If this would be interesting, I can do initial work for en_GB and es_ES translations.

BTW: this is amazing software. My congratulations to the team.

[brubu]

sprieto,
I have done the work you are doing few days ago for the Italian localization and I have found the same problems.
I have done the following.
Step 1)
Swapped all the lines in the Italian strings.txt file to match the exact order (line by line) of the GB file.
Step 2)
Added/Deleted all strings not matching.

I have done this using a file compare tool (I use Araxis Merge).
You can see the updated Italian file in CVS (and a new one submitted right now in Mantis).

Now whenever I discover a change in the GB file I simply run the file compare tool to detect the changed/Added/Moved/Removed line and update. This is very simple (now).

So I do not think that splitting in more file is needed.

Moreover I know that is possible to program Araxis Merge using regular expression to indicate only lines that change before the '='. When I will do this it will really be very simple?

Anyone know how to program this regular expression (I am not an expert)?

[fman]

The localization system, will not be changed.

[sprieto]

OK, it was just a suggestion.

I'm making finally checks to the Spanish translation file for testlink 1.7, following suggestions from brubu. I've made a simple console program to check string files against master file en_GB. If someone is interested, post a message.

[havlatm]

We could find some sophisticate solution for localization in future as the number of strings still increase. Anyway this checking string is handful. I will share it.