topic Re: Compress or Strip in SAS Programming

Compress or Strip

Dsrountree — Thu, 03 Oct 2013 14:19:39 GMT

PROC SQL

CREATE TABLE WORK.PLAN AS

SELECT T1.BENEFIT,

(COMPRESS(T1.BENEFIT, '')) AS NBENEFIT,

FROM WORK.PLANSVC T1;

QUIT;

T1.BENEFIT = $0 copay B

NBENEFIT = $0/ copay.

I need a function to strip away the following:

<ul>

</ul>

<li>

</li>

When using COMPRESS it takes all the values including the B and removes it. I only want the HTML language removed from my text.

I simplified the proc sql for posting purposes, but when I extended the COMPRESS it removed "b, /, u, l, i, <, >" from all of my data.

My end results should have removed the and the without removing the B

I am new to SAS by the way.........:smileyconfused:

Re: Compress or Strip

LinusH — Thu, 03 Oct 2013 14:26:10 GMT

Maybe tranwrd()?

Re: Compress or Strip

ballardw — Thu, 03 Oct 2013 14:29:09 GMT

You may actually want TRANSTRN as it replaces complete specified substring.

TRANWRD will leave a single space when attempting to delete.

Re: Compress or Strip

Dsrountree — Thu, 03 Oct 2013 14:34:19 GMT

I've never heard of TRANSTRN.

Can you provide a sample of how it functions?

Ex: TRANWRD(source,target,replacement)

Re: Compress or Strip

ballardw — Thu, 03 Oct 2013 14:48:52 GMT

Syntax is the same: TRANSTRN(source,target,replacement)

With the same limitation one source, one target and one replacement string per call.

Your example would require a separate call for each of:

<ul>

</ul>

<li>

</li>

I don't work with trying to read HTML so I'm not sure of a better way. I suspect though you'd be better off doing this in a data step

data work.plan;

set work.plansvc (keep=benefit);

benefit = transtrn(benefit, '',''); /* that is 2 single quotes not a double quote*/

benefit = transtrn(benefit, '','');

continue with each replacement.

IF you have a lot of variables then use an array;

data work.plan;

set work.plansvc ;

array v <list variables to process here>;

do i = 1 to dim (v);

v = transtrn(v,'','');

v = transtrn(v,'','');

end;

drop i;

run;