Identify Specific Letter/Numeric from a String Variable

[Robert Huang]

Dear all,
I have a list of countryname and countrycode, as follow:

countrycode	countryname
004	Afghanistan
008	Albania
012	Algeria
108	Burundi
191	Croatia
212	Dominica
233	Estonia
300	Greece
360	Indonesia
388	Jamaica
918	European Union
927	Euro Area
WLD	World
LMC	Lower Middle Income

I would like to add another variable regionalcat to identify if an observation is a country (0) or country group (1).
In this case, country groups have countrycode beginning with numeric 9 or letter (alphabet).

I tried using this code

gen regionalcat = .
gen regionalcat1 = substr(countrycode,1,1)
egen regionalcat2 = sieve(regionalcat1), keep(a)
capture assert regionalcat2 == regionalcat1
// if there's no alphabet
if _rc{
  // identify if the first character is "9"
    egen regionalcat9 = sieve(regionalcat1),omit(9)
    capture assert regionalcat9 == regionalcat1
    if _rc {
 // if the first character is 9
    replace regionalcat = 1
    }
    else {
// if the first character is numeric 012345678
    replace regionalcat = 0
    }
}
// if there's alphabet
else {
replace regionalcat = 1 
}

However, this code returns value 0 for all observations in regionalcat, can anybody help me? Thank you

[Jorrit Gosens]

* Example generated by -dataex-. To install: ssc install dataex
clear
input str3 countrycode str19 countryname
"004" "Afghanistan"        
"008" "Albania"            
"012" "Algeria"            
"108" "Burundi"            
"191" "Croatia"            
"212" "Dominica"           
"233" "Estonia"            
"300" "Greece"             
"360" "Indonesia"          
"388" "Jamaica"            
"918" "European Union"     
"927" "Euro Area"          
"WLD" "World"              
"LMC" "Lower Middle Income"
end

gen regionalcat = 1
replace regionalcat = 0 if real(substr(countrycode,1,1))<9