topic Re: SAS regular Expression in SAS Programming

SAS regular Expression

Satish_Parida — Fri, 02 Aug 2019 09:58:00 GMT

Hi Experts,

some of the praxmatch are not working for some regular expressions.

Could you please tell me whats wrong with the regular expressions where it is not working.

data test;
input StudyEventKey:$40.;
x = prxmatch(prxparse('/^VIST\[*\]*V*\[*\]/') ,StudyEventKey);	/*Does not work*/
x1 = prxmatch(prxparse('/^VIST\[*\]*V*\[1\]/') ,StudyEventKey);
x2 = prxmatch(prxparse('/^VIST\[*\]*V*\[2\]/') ,StudyEventKey);
y = prxmatch(prxparse('/^VIST\[*\]*V*/') ,StudyEventKey);
z = prxmatch(prxparse('/^VIST\[*\]\/*V*\[*\]/') ,StudyEventKey);  /*Does not work*/
cards;
VIST[1]/FV[1]
VIST[1]/SV[1]
VIST[1]/TV[1]
VIST[2]/FV[1]
VIST[2]/SV[1]
VIST[2]/TV[1]
;
run;

Re: SAS regular Expression

Criptic — Fri, 02 Aug 2019 10:46:38 GMT

Hey @Satish_Parida

The First Statement doesn't work because you are looking for the word VISIT then for 0 to n occurences of [ that is probabely not the intended effect.

Something like ^VIST\[\d\]*V*\[*\] would to the trick here.

I highly recommend using https://regexr.com/38ed7 to built your regex.

Kind regards

Criptic

Re: SAS regular Expression

Satish_Parida — Fri, 02 Aug 2019 10:51:33 GMT

Thank you, but x1 and x2 give correct output, that's why I am confused what is wrong with x?

Re: SAS regular Expression

Criptic — Fri, 02 Aug 2019 11:01:19 GMT

x1 findes the string VISIT[1] - why? because you match VISIT from the beginning, then 0 to n [, 0 to n ], 0 to n V, [, a 1 and finally a ] - so this regex finds the string VISIT[1] so this seems to work but it really doesn't do what you want.

x2 works in the same way, except that it matches VISIT[2]

Re: SAS regular Expression

Satish_Parida — Fri, 02 Aug 2019 11:05:58 GMT

Thank you.

Expression: /^VIST\[*[0-9|A-Z]\]\/*[0-9|A-Z]V\[*[0-9|A-Z]\]$/ works for all the scenarios I have.

Could you please help me with the expression, which will work for the following.

VIST[1A]/SSV[10A]

Re: SAS regular Expression

Criptic — Fri, 02 Aug 2019 11:11:09 GMT

^VIST\[[-zA-Z0-9]*\]\/SSV\[[a-zA-Z0-9]*\]$

Re: SAS regular Expression

Satish_Parida — Fri, 02 Aug 2019 11:13:34 GMT

This is a huge domain and I am new to REGEX,

/^VIST\[\w*\]\/\w*V\[\w*]$/

This is the best I could find.

Thank you for all your input.

The website https://regexr.com/38ed7 is very helpful.