topic Re: find strings between (),[],{},' '," " and split them in new variables in SAS Programming

find strings between (),[],{},' '," " and split them in new variables

Alexxxxxxx — Mon, 18 Mar 2019 17:44:24 GMT

Dear all,

How can I find all strings between (),[],and {} (such as <BR>, [FONT],{BODY},'A',"JUICE") and split them in a new variable?

Especially, for the

'JUICE<BR>apple<footer>',I expect to add a blank between 'JUICE' and 'apple'

by using the following code,


data have ;
  infile datalines truncover;
  input name $100.;
  datalines;JUICE<BR>apple[footer] 
juice <BR> apple 
juice<BODY> 'apple' 
juice{BODY} apple 
[BR]juice apple
<figure> "juice" LTD
 ;
run;

data want1;
   set have;
   RegExID = prxparse('/<\w*>/');
   start=1;
   call prxnext(RegExID, start, length(name), name, pos, length);
      do while (pos > 0);
         html = substr(name, pos, length);
         newname=prxchange('s/<\w*>//', -1, name);
         output;
         call prxnext(RegExID, start, length(name), name, pos, length);
      end;
   keep name html newname;
run;

I get

name	html	newname
JUICE<BR>apple<footer>	<BR>	JUICEapple
JUICE<BR>apple<footer>	<footer>	JUICEapple

however, I expect to add a blank between 'JUICE' and 'apple'

name	html	newname
JUICE<BR>apple<footer>	BR	JUICE apple
JUICE<BR>apple<footer>	footer	JUICE apple

Could you please give me some suggestions about this?

thanks in advance.

Re: find strings between (),[],{},' '," " and split them in new variables

PeterClemmensen — Mon, 18 Mar 2019 17:03:14 GMT

Just made a very small change to your program in the PRXCHANGE Function. See if this does the trick

data have ;
  infile datalines truncover;
  input name $100.;
  datalines;JUICE<BR>apple[footer] 
juice <BR> apple 
juice<BODY> 'apple' 
juice{BODY} apple 
[BR]juice apple
<figure> "juice" LTD ;
run;
data want1;
   set have;
   RegExID = prxparse('/<\w*>/');
   start=1;
   call prxnext(RegExID, start, length(name), name, pos, length);
      do while (pos > 0);
         html = substr(name, pos, length);
         newname=prxchange('s/<\w*>/ /', -1, name);
         output;
         call prxnext(RegExID, start, length(name), name, pos, length);
      end;
   keep name html newname;
run;

Re: find strings between (),[],{},' '," " and split them in new variables

Alexxxxxxx — Mon, 18 Mar 2019 17:25:10 GMT

thanks draycut,

but I get

name	html	newname
JUICE<BR>apple[footer]	<BR>	JUICE apple[footer]
juice <BR> apple	<BR>	juice apple
juice<BODY> 'apple'	<BODY>	juice 'apple'
<figure> "juice" LTD	<figure>	"juice" LTD

besides,

I cannot get the expected result by following code,

data want1;
   set have;
   RegExID = prxparse('/<\w*>|[\w*]|{\w*}|'\w*'|"\w*"/');
   start=1;
   call prxnext(RegExID, start, length(name), name, pos, length);
      do while (pos > 0);
         html = substr(name, pos, length);
         newname=prxchange('s/<\w*>|[\w*]|{\w*}|'\w*'|"\w*"/ /', -1, name);
         output;
         call prxnext(RegExID, start, length(name), name, pos, length);
      end;
   keep name html newname;
run

Re: find strings between (),[],{},' '," " and split them in new variables

Alexxxxxxx — Mon, 18 Mar 2019 17:46:20 GMT

for example,

name

JUICE<BR>apple[footer]

juice <BR> apple

juice<BODY> 'apple'

<figure> "juice" LTD

I expect to get

name	html	newname
JUICE<BR>apple[footer]	BR	JUICE apple
JUICE<BR>apple[footer]	footer	JUICE apple
juice <BR> apple	BR	juice apple
juice<BODY> 'apple'	BODY	juice
juice<BODY> 'apple'	apple	juice
<figure> "juice" LTD	figure	LTD
<figure> "juice" LTD	juice	LTD

Re: find strings between (),[],{},' '," " and split them in new variables

PeterClemmensen — Mon, 18 Mar 2019 18:16:30 GMT

Ah, I see what the problem is. The old Quotes Within Quotes problem 🙂

This

data have ;
  infile datalines truncover;
  input name $100.;
  datalines;
JUICE<BR>apple[footer] 
juice <BR> apple 
juice<BODY> 'apple'
<figure> "juice" LTD
;

data want;
   format name html newname;
   set have;
   RegExID = prxparse('/<\w*>|\[\w*\]|\(\w*\)|"\w*"|''\w*''/');
   start=1;
   call prxnext(RegExID, start, length(name), name, pos, length);
   newname=prxchange('s/<\w*>|\[\w*\]|\(\w*\)|"\w*"|''\w*''/ /', -1, name);
      do while (pos > 0);
         html = substr(name, pos+1, length-2);
         output;
         call prxnext(RegExID, start, length(name), name, pos, length);
      end;
   keep name html newname;
run;

proc print data=want;
run;

gives you

Re: find strings between (),[],{},' '," " and split them in new variables

Alexxxxxxx — Thu, 28 Mar 2019 16:50:13 GMT

Hello @PeterClemmensen

I have a new question during the process.

the

HARDY(FRNS.)'A'

cannot be processed by the code

I expect to get

name	COMPANY_NAME_inB	COMPANY_NAME_noB
HARDY(FRNS.)'A'	FRNS.	HARDY
HARDY(FRNS.)'A'	A	HARDY

However, I only get

name	COMPANY_NAME_inB	COMPANY_NAME_noB
HARDY(FRNS.)'A'	A	HARDY(FRNS.)

Could you please give me some suggestions?