On Mon, 1 Jun 2015 17:11:54 -0400, Shmuel Metz (Seymour J.) <shmuel+ibm-m...@patriot.net> wrote:
>In <0379604180364016.wa.bgodfrey.gzgmail....@listserv.ua.edu>, on >05/29/2015 > at 10:30 AM, Bill Godfrey <bgodfrey...@gmail.com> said: > >>I get identical results whether I use \n or $ in the OP's example. In >>OMVS. I'm not addressing your question but rather the OP's example. > >Which OMVS facilities match \n to end of line (record) and which to >LF? What do grep et al do about matching \n against legacy PS data >sets, where there is a logical end of record? > When commands like "cat" and "cp" read legacy PS data sets as text, the results reflect this description of "reading text files" in the C/C++ Programming Guide: http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/CBCPG1C0/2.9.4.2 Where it says: "For files opened in fixed text format, rightmost blanks are stripped off a record at input, and a new-line character is placed in the logical record." That new-line character is hex 15. So a 3-line data set of 80-byte records that look like this: a test testing will look like this in hex after being read by "cat" cat "//test.cntl" | od -tx1 -An 81 15 A3 85 A2 A3 15 A3 85 A2 A3 89 95 87 15 which is the same result as this command: printf %b "a\ntest\ntesting\n" | od -tx1 -An 81 15 A3 85 A2 A3 15 A3 85 A2 A3 89 95 87 15 The only facility with regular expressions that I have found that matches \n to end-of-line is in Perl. For example: printf %b "a\ntest\ntesting\n" | perl -ne 'print if /test\n/' test printf %b "a\ntest\ntesting\n" | perl -ne 'print if /test$/' test The "grep" and "awk" commands don't match \n to end-of-line on omvs, or on linux for that matter. The "grep" command can't read a legacy PS data set directly, but "awk" can. awk '/test$/' "//test.cntl" test Bill ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
