Kolusu, here are the first 3 column definitions: CHAR(10) N.N., CHAR(30) NOT NULL, CHAR(8) N.N.,
As you can see, each of the columns is separated by a comma, which is outside of the parens. The N.N. is an abbreviation for Not Null, and the abbreviation is used when the column width is less than the width of that descriptor. In terms of separation of the columns, I believe that all the fields have NN or Not Null, so you can only distinguish each column attribute by "NULL," or "N.N.," here. Another option is to convert ") N.N." to ") NOT NULL" or vice-versa, if that makes parsing easier. As to the input file, it is VB, and the max data record length is 200 (so, VB-204). There are actually several files, but 200 is the longest. I could write a Rexx program to handle this, but if Sort could do it, it would be much simpler. Thanks again! Don ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
