I recently got the job of completing some of the leftover drupal localization files. Well 77% was done, I just had to worry about the left 23%. How hard can that be? Well, my judgment kinda failed me. While it wasn't so hard, it was pretty challenging. Fortunately, language sort happens to be one of my strong points but I was not familiar with the nepali unicode system. First of all, I needed to learn to type in this thing. While I had done similar thing for Java Micro Edition while making interface, I successfully circumvented the problem by making a VB program to do all the boring work while I typed.
Just typing wasn't enough, I had to then find how to use the poEdit without errors. Getting a little bored, I decided to explore in the drupal interface itself. Voila, I escaped from the ugly interface of poEdit. Ok I admit it, it wasn't so spectacular though :). After doing some modification to my working environment to make it more interesting, I went to work on the translation.
It was then I realized how poor Nepali language was. It wasn't the beautiful language Devkota once described as that spoken by the Gods(whatever!) but the cumbersome language that didn't fit anywhere. There were words that needed some imagination after the trusty ol' glossary.pdf failed to find those words. Everytime I needed to do something, I had to translate into English to see where I needed to go. Although this has improved over time, the looks really don't seem to improve.
The Nepali unicode is not really so beautiful for today's sleek "Web 2.0" sites, it's small and needs to be significantly increased in size (thankyou CSS) to be readable. While I finally did the work, there was a rather silly mistake left. Apparently some of the guys put the %(nepali-word) instead of the original variable (okay okay, %english-word) which showed sth like
लाइन %लाइनमा एरर छ।
That line number is everything for a debugger.... there were many other variables and did more harm than good. After all, who likes to look for errors and keep fixing in a slow working notepad?
At least most were removed.
There were some errors still caused while importing which were resolved easily. Mostly they were due to the plural thing for many fields.
Resolved using
msgstr[0] " "
msgstr [1] " "
(remember the two lines)
for the error like "missing msgstr"...
finally, the localization is ready for CAN...
Monday, January 8, 2007
Subscribe to:
Posts (Atom)