Question

Right now I’m building a small app that imports data from a spreadsheet and due to the nature of the original entry, there is a string being read in that has values such as 8½, 2½, etc.

My goal with a simple function is to convert 2½ into float 2.5, for example.

I’ve tried the .to_f method but that has left me with a weird value of 2.02½.

Any insight or suggestions here would be very much appreciated!

Answer 1

Unicode only supports a small number of vulgar fractions so a simple lookup table will do the trick:

# You might want to double check this mapping
vulgar_to_float = {
    "\u00BC" => 1.0 / 4.0,
    "\u00BD" => 1.0 / 2.0,
    "\u00BE" => 3.0 / 4.0,
    "\u2150" => 1.0 / 7.0,
    "\u2151" => 1.0 / 9.0,
    "\u2152" => 1.0 /10.0,
    "\u2153" => 1.0 / 3.0,
    "\u2154" => 2.0 / 3.0,
    "\u2155" => 1.0 / 5.0,
    "\u2156" => 2.0 / 5.0,
    "\u2157" => 3.0 / 5.0,
    "\u2158" => 4.0 / 5.0,
    "\u2159" => 1.0 / 6.0,
    "\u215A" => 5.0 / 6.0,
    "\u215B" => 1.0 / 8.0,
    "\u215C" => 3.0 / 8.0,
    "\u215D" => 5.0 / 8.0,
    "\u215E" => 7.0 / 8.0,
    "\u2189" => 0.0 / 3.0,
}

Then, a little bit of regex wrangling to pull your “number” apart:

s = "2½"
_, int_part, vulgar_part = *s.match(/(\d+)?(\D+)?/)

And finally, put them together taking care to properly deal with possible nils from the regex:

float_version = int_part.to_i + vulgar_to_float[vulgar_part].to_f

Remember that nil.to_i is 0 and nil.to_f is 0.0.