topic extracting part of a string in SAS Data Science

extracting part of a string

ilikesas — Sat, 13 Feb 2016 19:23:49 GMT

Hi,

suppose I have a data set with datalines in the following manner:

the first site is www.abc.com

www.123.com is the second website.

I am trying to figure out how to extract the websites, that is, the part of the string which is between (and including) "www." and ".com"

Thank you!

Re: extracting part of a string

mohamed_zaki — Sun, 14 Feb 2016 00:37:50 GMT

one way

data have;
input x $37.;
cards;
the first site is www.abc.com
www.187878723.com is the second website.
www.computer.com is the second website.
;
run;

data want ;
set have;
www=index(x, "www.");
com=index(substr(x,www+4), ".com");
if www and com then website=substr(x,www,com+7);
drop www com;
run;

Re: extracting part of a string

mohamed_zaki — Sun, 14 Feb 2016 00:19:40 GMT

updated to handle more cases

Re: extracting part of a string

ilikesas — Sun, 14 Feb 2016 00:44:40 GMT

Hi Mohamed,

thank you for answering my question, everything works nicely!

Just on the sidenote, if I add a dataline " a pseudo site www. name .com", the code will still select "www. name .com" into want, but it isn't a real website becasue of the space after www. and before .com

So is there a way to avoide it by specifying that right after www. and right before .com there should be a character?

Thank you!

Re: extracting part of a string

mohamed_zaki — Sun, 14 Feb 2016 00:47:39 GMT

if www and com then website=compress(substr(x,www,com+7));

Re: extracting part of a string

mohamed_zaki — Sun, 14 Feb 2016 00:50:28 GMT

Do you still want to extreact it correctly? .... or to consider it wrong and neglect it?

Re: extracting part of a string

ilikesas — Sun, 14 Feb 2016 01:51:34 GMT

I ran the code with your new input and understand that it corrects it.

Could you please also show me the option to neglect such a case?

thnak you!

Re: extracting part of a string

mohamed_zaki — Sun, 14 Feb 2016 12:20:21 GMT

data want ;
set have;
www=index(x, "www.");
com=index(substr(x,www+4), ".com");
website=substr(x,www,com+7);
if www and com;
if index(trim(website),' ')> 0 then website="";
drop www com;
run;