Re: WILDCARD

>Date: Wed, 27 Aug 1997 18:50:29 +0400 >From: Goktug TEZUYSAL <goktug@prizma.net.tr> >To: informix-list@rmy.emory.edu >X-Informix-List-Id: <list.16204> > >Goktug TEZUYSAL wrote: >} DEAR SIR, >} WE ARE USING ONLINE V 7.22 AND 4GL V 6.03 ON SUN SOLARIS 2.5 >} SYSTEM.WE HAVE C SOURCES TO BE LINKED WITH fglgo WHICH WE HAVE >} CREATED NEW RUNNER SUCCESSFULLY.BEFORE CREATING NEW RUNNER >} WE WERE ABLE TO DO WILDCARD SEARCHES WITH '*' HOWEVER NOW WITH >} THE NEW fglgo, '*' DOES NOT WORK WITHIN THE CONSTRUCT STATEMENT. >} INSTEAD '?' DOES THE MATCHES COMMAND.THE NEW RUNNER searches >} the data with = 'T*' statement which returns no rows,the old runner >} does this with matches 'T*' which works fine. >} THIS SEEMED TO BE ABNORMAL IN OUR PROGRAMS, DO YOU HAVE ANY >} SUGGESTIONS ABOUT THIS PROBLEM? 1. THERE'S NO NEED TO SHOUT - please. 2. Why not version 6.05, rather than 6.03? 3. Does the basic fglgo runner handle a CONSTRUCT statement with a '*' correctly, or not? If it does handle it correctly, then somehow the code you added to the runner has (apparently) altered the behaviour of the CONSTRUCT code without breaking it (which is unusual in the extreme). If not, then it is bug in the basic product you are using. 4. I vaguely recall hearing of some such bug, but I've not checked in the PTS database. I've not encountered any problems with 6.05.UC1 I4GL-RDS on Solaris 2.5.1 -- I wrote the attached program to check that typing 'T*' on the form generates the correct condition "sometable.somecolumn matches 'T*'") and therefore I would recommend an upgrade if you have demonstrated a bug in 6.03. Yours, Jonathan Leffler (johnl@informix.com) #include <disclaimer.h> PS: Warning I do not reply to messages with anti-spam in the return path. -- kk.per -- DATABASE FORMONLY SCREEN { [f000 ] } END ATTRIBUTES f000 = FORMONLY.field01; END -- kk.4gl -- MAIN DEFINE s CHAR(50) OPEN FORM f FROM "kk" DISPLAY FORM f CONSTRUCT s ON sometable.somecolumn FROM FORMONLY.* CLOSE FORM f CLEAR SCREEN DISPLAY s END MAIN