Thanks a lot! It works! ... View more

They have to be enclosed in ## ... View more

Do all of your examples have # characters?  blablablabla blablalba Not all examples have # characters. Do they have 2 and only 2 # characters? Could there be just one? some other text#blablablabla blablalba No, some lines have more than 2 # characters, but not less than 2 Could there be three? some other text#ABC#CDE#more text It could be text #abc# text #abs# text Could there be four or more? some other text#ABC#more text#CDE#and even more text Yes Is the # character ever the first character? #AB#some other text # character can be first, in the middle of the text or last in the sentence: #abc# text text #abc# text Text #abc# text #abc# text #abc# text   ... View more

Could you please explain metecharacters you've used: prxchange('s/.*#(.+)#.*/\1/', -1, string) ... View more

Draycut, it works! Thank you!!! ... View more

I don't use cards, because there are thousands of row with different data data have; input text&$100.; news=scan(text,2,'#'); cards;blablablabla #ABC# blablabla; run; I use ode:   PROC SQL; CREATE TABLE WANT AS SELECT DISTINCT t1.ID, t1.CREATED, t1.COMMENT_TXT, (scan(t1.COMMENT_TXT, 2, "#")) AS ABC FROM data.COMMENT t1 ;QUIT; ... View more

Thanks, Jag! scan(t1.COMMENT_TXT, 2, "#") works, but the result is ABC blablabla, all characters after last # stay too 😞 ... View more

Thanks ... View more

Jagadishkatam, thank you so much, it works! ... View more