So I coded a little Tcl script to automate this for me.
How does it work?
Well, first of all it's not fully automated, it just helps a little bit.
Step 1:
Generate the .tr file with the instructions at the U++ docs.
I got the attached PT-BR.UTF-8.tr file.
Step 2:
Go to http://translate.google.com and translate the file above.
Automatic translation should save some time, but it's not perfect.
Copy and paste it and save it as translated.txt. I got the attached translated.txt file.
Step3:
Run my attached Tcl script as follows:
$ tclsh parse.tcl PT-BR.UTF-8.tr translated.txt > test.txt
I got the attached test.txt file. Which is already in U++ translation format, but the translation is not perfect because it was translated by Google translate.
But now the job is half done. Most of the strings are already translated right. I just need to fix the broken ones.
Step 4:
The final step is to fix the broken translations (hopefully very few if Google does its job fine).
That's it. I hope you like my tool. Attached are the files including the Tcl script.
To run the script you need Tcl installed. On Ubuntu you can install it with "sudo apt-get install tcl8.5" if it's not already isntalled.
You can get a self-contained script interpreter here: http://equi4.com/tclkit/ for all platforms including Windows.
In this case run "tclkit parse.tcl PT-BR.UTF-8.tr translated.txt > test.txt".
Conclusion:
I just wanted to share this. I'll upload the fixed ptBR translation when I have some time to edit the file.