Fractions in regular expressions

[Stephen Porter]

I have some local government data with addresses that use the 1/2 character, e.g. "305 ½ SMITH RD " or "41½ JONES RD". I'm trying to extract the full numeric portion including the fraction, but in my search the only examples I've come across use fractions written like 1/2. Any suggestions?

[Andrew Musau]

Show us some examples using dataex.

[Hemanshu Kumar]

Consider this example and solution (this allows for fractions other than half as well):

clear
input str40 address
"305 ½ SMITH RD"
"41½ JONES RD"
"Platform 9¾, King's Cross Station"
end

gen wanted = ustrtrim(ustrregexs(1)) if ustrregexm(address, "([\d\s\x{00BC}-\x{00BE}\x{2150}-\x{215E}]+)")

which produces:

. list, noobs string(40)

  +--------------------------------------------+
  |                           address   wanted |
  |--------------------------------------------|
  |                    305 ½ SMITH RD    305 ½ |
  |                      41½ JONES RD      41½ |
  | Platform 9¾, King's Cross Station       9¾ |
  +--------------------------------------------+

[Stephen Porter]

That worked, thank you Hemanshu!