I'm working on a program that convert numbers written in kanji into their arabic numbers counterparts and one of the text I've been given to test my program contains numbers written in katakana. Most of these numbers are small decimals or version numbers (ジュリウスサンテンイチ=>ジュリウス3.1 or ニーテンニメガ=>2.2メガ for example) so they are easy to convert but there are also bigger numbers (like センサンビャクジュッテンイチナナ=>1310.17) which doesn't seem very natural to me, given how long it is to write.
I have implemented something that converts small numbers written in katakana (decimals up to ten) but implementing something that convert every numbers written in katakana would take a long time and I'm wondering if this is worth the hassle because it does seem unlilely to me that people would write big numbers in katakana. My deadline is close so I'd rather spend my time on something more useful.
What do you guys think?