String manipulation - show certain characters based on condition

[Q1983]

data have;

infile datalines;

input ln_no $ Branch $50. ;

return;

datalines;

33 CHARLOTTE COUNTY - DEEP CREEK

22 CHASTAIN SQUARE PUBLIX (closed)

62 CHARLOTTESVILLE AREA - ALBEMARE

64 MARSHALL SQUARE PUBLIX (closed)

;run;

/*I want to accomplish the following:

1. if the Branch is separated by - only show the string after the dash (ie DEEP CREEK, ALBERMARLE)

2. if the Branch is not separated by - just show the string inside the "( )" only (ie closed)

FOOTNOTE

for #2 also how would I do this if I just want to show the characters before the  ( ) or in this case MARSHALL SQUARE PUBLIX 

[Tom]

SCAN() can probably do what you want.

data have;
  input ln_no $ Branch $50. ;
datalines;
33 CHARLOTTE COUNTY - DEEP CREEK
22 CHASTAIN SQUARE PUBLIX (closed)
62 CHARLOTTESVILLE AREA - ALBEMARE
64 MARSHALL SQUARE PUBLIX (closed)
;

data want;
  set have;
  length want1 want2 $50;
  want1 = strip(coalescec(scan(branch,2,'-'),scan(branch,-2,'()')));
  want2 = strip(coalescec(scan(branch,2,'-'),scan(branch,1,'()')));
run;

proc print;
run;

Obs    ln_no                Branch                 want1         want2

 1      33      CHARLOTTE COUNTY - DEEP CREEK      DEEP CREEK    DEEP CREEK
 2      22      CHASTAIN SQUARE PUBLIX (closed)    closed        CHASTAIN SQUARE PUBLIX
 3      62      CHARLOTTESVILLE AREA - ALBEMARE    ALBEMARE      ALBEMARE
 4      64      MARSHALL SQUARE PUBLIX (closed)    closed        MARSHALL SQUARE PUBLIX

[Patrick]

Below should do the job.

data have(drop=_:);
  infile datalines;
  input ln_no $ _str $50.;
  length branch  $20 branch2 $40;
  branch=scan(_str,2,'-');
  if missing(branch) then branch=scan(_str,-2,'()');
  branch2=scan(_str,1,'()-');
  datalines;
33 CHARLOTTE COUNTY - DEEP CREEK
22 CHASTAIN SQUARE PUBLIX (closed)
62 CHARLOTTESVILLE AREA - ALBEMARE
64 MARSHALL SQUARE PUBLIX (closed)
;
run;
proc print;
run;

[ed_sas_member]

Hi @Q1983 

You can also use pearl functions to achieve that:

data have;
	infile datalines;
	input ln_no $ Branch $50.;

	if prxmatch('/(.*-)(.*)/', Branch) then Branch=prxchange('s/(.*-)(.*)/$2/', -1, Branch);
	else Branch=prxchange('s/(.*)(\(.*\))/$2/', -1, Branch); /* put $1 instead of $2 if you want to display the characters before the ( ) */

	return;
	datalines;
33 CHARLOTTE COUNTY - DEEP CREEK
22 CHASTAIN SQUARE PUBLIX (closed)
62 CHARLOTTESVILLE AREA - ALBEMARE
64 MARSHALL SQUARE PUBLIX (closed)
;
run;

proc print;

Explanation :

- prxmatch('/(.*-)(.*)/',branch) looks for the pattern (.*-)(.*), meaning any character (.) zero, one or more times (*), an hyphen (-), followed by any character (.) zero, one or more times (*). If the condition is true, then the prxchange function retrieves the second argument in parenthesis ($2), so the string after the hyphen (prxchange('s/(.*-)(.*)/$2/', -1, Branch).

- It is quite the same logic for the second use case: the prxchange function looks for the pattern (.*)(\(.*\)), meaning any character (.) zero, one or more times (*), followed by any character (.) zero, one or more times (*) in parenthesis \( and \)

If you want to display the characters before the parenthesis, you can use $1, meaning the first group in parenthesis.

[Q1983]

Appreciate the explanation