Kata

I understand your suggestion. How would you clarify it?

1.Any single 'i' followed...

2.Any single letter 'i' followed...

3.Specially for letter 'i', when followed...

4.Given the special behavior of letter 'i', when followed...

Please, let me know how you think it best clarified to improve the description.

Thank you for your suggestion! :)

I would word it along the lines of:

From the input, any '1' that is preceded by a space and followed by a non-alphanumeric
character should translate to its upper-case letter assignment.
Any other numbers do not follow this special case.

*edited for a more precise definition

Great! I'm glad you liked it! :)

Actually I didn't consider that scenario because I considered that writing a character just after a dot would be an invalid input.

However, now that you mention it, I see the possibility of having numbers in the string (not just words as it is right now), in which case this kind of input could be valid. I think any word with a dot in the middle should be considered a number, as a space must be left between a period and a new word. Therefore, the decyphered version of your string, as far as I understand the problem, should be: 234.234, not Rea.Rea nor Rea.zea. What was your result in that case?

I'm thinking of doing another Kata cyphering and decyphering strings of words and numbers. What do you think?

Thank you very much for your comment! :)

Honestly? I think it's broken by design as a cipher algorithm. :P

I see two directions to take it forward: more language parsing (and before you know it, you are implementing English just to obfuscate it as if you were a l33t h4x0r, but one who can write numbers :), or a stronger (different) encryption algorithm (and before you know it, you are the next Alan Turing and GCHQ comes ringing your doorbell with a cozy little white jacket and a cozy little padded cubicle for you to work in).

Actually, seeing how your first impulse was to start doing more input parsing, the language angle might be the more obvious choice. Not a bad one either. :]

But I have a feeling there might be more interesting things to do with parsed English than translate it into 1337sp34k. How's weighing text elements for a web search grab you? Filter out the fluff, extract the key words, match them allowing for synonyms (a bear being entirely different from to bear)?

OK, back to reality now ..