topic Re: How to find a substring with variable numbers in SAS Programming

How to find a substring with variable numbers

kz_ — Fri, 27 Jan 2023 20:25:30 GMT

I would like to find the location of a substring of the form #.#Z, where the # sign represents any numeric digit.

For example, the substring might be 7.6Z or 1.3Z.

It is possible that there may be other digits in the string overall, though this particular pattern will only be found once in the string.

Re: How to find a substring with variable numbers

PaigeMiller — Fri, 27 Jan 2023 21:11:58 GMT

Find the location of the Z, and then you know the location of the start of the substring as well.

where = find(string,'Z','i');

Of course, that fails if a Z could appear prior to the digits.

Re: How to find a substring with variable numbers

Patrick — Sat, 28 Jan 2023 11:24:56 GMT

Assuming you then also want to extract this sub-string below one way to go.

data sample;
  input have_str :$40.;
  _prxid=prxparse('/\d+\.\d+Z/');
  want_startpos=prxmatch(_prxid,trim(have_str));
  call prxsubstr(_prxid, trim(have_str), _pos , _len);
  length want_str $10;
  want_str=substr(have_str,_pos,_len);

  /* or if you just want the number and Z could be upper or lowercase */
  _prxid2=prxparse('/\d+\.\d+(?=Z)/i');
  call prxsubstr(_prxid2, trim(have_str), _pos , _len);
  want_num=input(substr(have_str,_pos,_len), ?? best32.);
  datalines;
asdfasd78adsfad7.6Zdafad
adfadZ90Zadsf1.3Zsafd7890p0
adfadZ90Zadsf13Zsafd7890p0
;

proc print data=sample;
run;

Re: How to find a substring with variable numbers

andreas_lds — Sun, 29 Jan 2023 20:18:24 GMT

Can you post a working data step containing example data? That would make it easier to suggest working code. Showing the expected results would be useful, too.

Re: How to find a substring with variable numbers

johnagalloway0 — Mon, 30 Jan 2023 09:36:50 GMT

With apologies to Patrick for stealing his test data, perl expressions will do the job but can be difficult to understand - why not just use translate and the method your question suggests:

If the character '#' may also appear in the string then just substitute another printable character - or even a non-printable character such as '00'x

Re: How to find a substring with variable numbers

johnagalloway0 — Mon, 30 Jan 2023 09:41:42 GMT

apologies I misunderstood - the +3 finds the position of 'Z' - remove it (or change to +0) to get position of string