Function to extract substring between two delimiters that are the same

[mglogan]

Hello!

I need to extract a medication class that is between two delimiters that are the same. In this case, it is a semi-colon (;).

have (oldvar) = ATC|R|RESPIRATORY SYSTEM; ATC|R03|DRUGS FOR OBSTRUCTIVE AIRWAY DISEASES; ATC|R03A|ADRENERGICS, INHALANTS; ...

want (newvar) = ATC|R03|DRUGS FOR OBSTRUCTIVE AIRWAY DISEASES

can you suggest an effective way to do this?

Thank you!  

[novinosrin]

do you mean this?

data test;
have=' ATC|R|RESPIRATORY SYSTEM; ATC|R03|DRUGS FOR OBSTRUCTIVE AIRWAY DISEASES';
want=scan(have,-1,';','m');
run;

or just this?

data test;
have=' ATC|R|RESPIRATORY SYSTEM; ATC|R03|DRUGS FOR OBSTRUCTIVE AIRWAY DISEASES';
want=scan(have,-1,';');
run;

[mglogan]

Thanks for the quick response!

I should have included the entire length of the string which is as follows:

ATC|R|RESPIRATORY SYSTEM; ATC|R03|DRUGS FOR OBSTRUCTIVE AIRWAY DISEASES; ATC|R03A|ADRENERGICS, INHALANTS; ATC|R03AC|SELECTIVE BETA-2-ADRENORECEPTOR AGONISTS; PRODUCT|001395 01 001|SALBUTAMOL; PRODUCTSYNONYM|001395 01 002|VENTOLIN /00139501/;

[novinosrin · Accepted solution]

is it always in 2nd position?

if yes, extract the 2nd

data test;
have='ATC|R|RESPIRATORY SYSTEM; ATC|R03|DRUGS FOR OBSTRUCTIVE AIRWAY DISEASES; ATC|R03A|ADRENERGICS, INHALANTS; ATC|R03AC|SELECTIVE BETA-2-ADRENORECEPTOR AGONISTS; PRODUCT|001395 01 001|SALBUTAMOL; PRODUCTSYNONYM|001395 01 002|VENTOLIN /00139501/;';
want=strip(scan(have,2,';'));
run;

[mglogan]

It will always be between the first and second semi-colon.

[novinosrin]

ok, i updated in my previous. check that out plz

here again

data test;
have='ATC|R|RESPIRATORY SYSTEM; ATC|R03|DRUGS FOR OBSTRUCTIVE AIRWAY DISEASES; ATC|R03A|ADRENERGICS, INHALANTS; ATC|R03AC|SELECTIVE BETA-2-ADRENORECEPTOR AGONISTS; PRODUCT|001395 01 001|SALBUTAMOL; PRODUCTSYNONYM|001395 01 002|VENTOLIN /00139501/;';
want=strip(scan(have,2,';'));
run;

[mglogan]

It worked. thank you!