*Well, I wouldn't have used the same variable name as the test data, and it's not clear whether the period in 9.12 is fortuitous. Your parse won't work against, e.g., myVar = 'word1 word2 666 word3.ext'*
I didn't use WORD3 in my test data. As for a fortuitous period in the test data, you work with what you are given. My less efficient solution is to use the REVERSE function. TRACE I MYVAR = 'WORD1 WORD2 9.12 XXXXZ.EXT' PARSE VALUE REVERSE(MYVAR) WITH EXT '.' WORD3 . SAY REVERSE(WORD3) PULL RESP Eliminates the fortuitous period from the mix. 5 *-* MYVAR = 'WORD1 WORD2 9.12 XXXXZ.EXT' >L> "WORD1 WORD2 9.12 XXXXZ.EXT" 7 *-* PARSE VALUE REVERSE(MYVAR) WITH EXT '.' WORD3 . >V> "WORD1 WORD2 9.12 XXXXZ.EXT" >F> "TXE.ZXXXX 21.9 2DROW 1DROW" >>> "TXE" >>> "ZXXXX" >.> "21.9 2DROW 1DROW" 9 *-* SAY REVERSE(WORD3) >V> "ZXXXX" >F> "XXXXZ" *On Wed, Feb 26, 2020 at 4:01 AM Seymour J Metz <sme...@gmu.edu <sme...@gmu.edu>> wrote:* > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > *The name is from CS, but the regexen in, e.g., Perl, Python, Ruby, can > describe grammars that are not RE grammars in the CS sense. -- Shmuel > (Seymour J.) Metz http://mason.gmu.edu/~smetz3 > <http://mason.gmu.edu/~smetz3> ________________________________________ > From: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU > <IBM-MAIN@LISTSERV.UA.EDU>> on behalf of Paul Gilmartin > <0000000433f07816-dmarc-requ...@listserv.ua.edu > <0000000433f07816-dmarc-requ...@listserv.ua.edu>> Sent: Tuesday, February > 25, 2020 11:51 AM To: IBM-MAIN@LISTSERV.UA.EDU <IBM-MAIN@LISTSERV.UA.EDU> > Subject: Re: Rexx parse using period as placeholder On Tue, 25 Feb 2020 > 08:14:33 -0500, Tony Thigpen wrote: >"regular expressions". What a misnamed > item. It should be called "geek >readable only expressions". :-) > The name > is adopted from formal language theory, clearly a geeky field. An > alternative name is Chomsky type-3. > https://secure-web.cisco.com/1yEhmTUBhK8QGgUQt09y0JsOgSqhJXPgIB7Wz4gZwgrTHdy8wAmwypFT0tCrS-JB2GheOTXaWsSeHJvgS2ofoCdp0m-1F0xEEQ1yXoQPhi6lbKSQQKlTBLZzS3TTM684q07ZikhoWrXEgICvTpo9FJcP95rTkyG3euEsTRYwDhlCXDQy_L5QL_LGkcmjt4G4xcTLA5yxPC3kMDpTFlLzZ31kiCv-pmGKv9GELh_RI2HeGU3cuxaX9kEFbeLzlK9eZvHT_rqfF1xTGp98og32FnHCCvu30HGdzhtB0IkRj8VaAbLArDRGTApSMFRK0kJHQ6mgZ9_m92aUcmsqyN6vlKLSyxEZtJKpHgap3ZEnOXvRq8yUe1zu3WkELf09COKLJTEn0ROC6L2B7BXUy1LsGE4Qj6Jwt3MuW_tao1O43pxGqN5xEtkLqIt86glUmt6Hu/https%3A%2F%2Fen.wikipedia.org%2Fwiki%2FChomsky_hierarchy#Type-3_grammars > <https://secure-web.cisco.com/1yEhmTUBhK8QGgUQt09y0JsOgSqhJXPgIB7Wz4gZwgrTHdy8wAmwypFT0tCrS-JB2GheOTXaWsSeHJvgS2ofoCdp0m-1F0xEEQ1yXoQPhi6lbKSQQKlTBLZzS3TTM684q07ZikhoWrXEgICvTpo9FJcP95rTkyG3euEsTRYwDhlCXDQy_L5QL_LGkcmjt4G4xcTLA5yxPC3kMDpTFlLzZ31kiCv-pmGKv9GELh_RI2HeGU3cuxaX9kEFbeLzlK9eZvHT_rqfF1xTGp98og32FnHCCvu30HGdzhtB0IkRj8VaAbLArDRGTApSMFRK0kJHQ6mgZ9_m92aUcmsqyN6vlKLSyxEZtJKpHgap3ZEnOXvRq8yUe1zu3WkELf09COKLJTEn0ROC6L2B7BXUy1LsGE4Qj6Jwt3MuW_tao1O43pxGqN5xEtkLqIt86glUmt6Hu/https%3A%2F%2Fen.wikipedia.org%2Fwiki%2FChomsky_hierarchy#Type-3_grammars> > It refers not to the template but to the target. I shudder to envision a > formal grammer for UNIX regexen. On Tue, 25 Feb 2020 12:11:24 +0000, > Seymour J Metz wrote: >What is happening is that everything in the template > before the string match '.' >is a sub-pattern applied to the fragment to > the left of the period. ... > Good description. Does IBM use the term > "sub-pattern" anywhere in its documentation? On Tue, 25 Feb 2020 13:24:09 > +0000, Seymour J Metz wrote: >While the Eunix RE syntax is grotesque, the > regexen are incredibly useful. I just wish that >they had adopted the > pattern syntax from SNOBOL or Wylbur. > There are two poles of UI design > style: o Ergonomic or keystroke economy. Frequently used commands are > single keystrokes with no lexical separator between the command and its > argument, as in the XEDIT "/target" for a string search. o Orderly > command grammar. I suspect that TSO begat CLIST and linemod EDIT, which > begat ISPF commands. Good in a script; terrible from the keyboard. A > very simple regex pattern, / ' " /, was formerly available in ISPF EDIT > only as F x'407d407f40'! Now that ISPF has regexen, it can be the > relatively transparent F R' ['] " '. But only the hex can be used as a > replacement. -- gil > ---------------------------------------------------------------------- For > IBM-MAIN subscribe / signoff / archive access instructions, send email to > lists...@listserv.ua.edu <lists...@listserv.ua.edu> with the message: INFO > IBM-MAIN > ---------------------------------------------------------------------- For > IBM-MAIN subscribe / signoff / archive access instructions, send email to > lists...@listserv.ua.edu <lists...@listserv.ua.edu> with the message: INFO > IBM-MAIN * *-- * *Wayne V. Bickerdike* ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
