Community
Participate
Working Groups
Created attachment 206338 [details] Test case. 1. CNTL-1 SQL statement for open rs from ds for row gets validation error when SQL statement is saved because "for row" not allowed with "with #sql". Recommend leaving the for row in as a comment in case developer needs to bebuild statement. open rs from ds /* for row */ with #sql ..... 2. CNTL-1 SQL statement for open rs behavior is different from the default open behavior. Default sql statement has no using, where, or order by. Behavior should be consistent. I would prefer CNTL-1 behavior (less thinking required to modify), but consistency is most important. See attached test. Download both test case and library load program since both are new. Test results: Start test for defect: CNTL-1 open validation error and inconsistent behavior Default open Read row 10 Read row 9 Read row 8 Read row 7 Read row 6 Read row 5 Read row 4 Read row 3 Read row 2 Read row 1 CNTL-1 open Read row 5
Created attachment 206339 [details] library for loading database for test program
Paul: For consistency, do you mean SQL statements generated by Ctl+1 and generators should be consistent? If so, Joe, which one do you consider is better for current phase, if need modification for Ctr+1 part, I can modify it. As for your first question, I have opened a bug for it: bug 362546 open rs from ds for basicRec using(basicRec.rid) with #sql{ select rid, rtrim(orderItem), price from Basic where rid = ? order by rid asc }; for the above statement, we cannot delete 'for basicRec' part because using phase is referring it; if we remove for and with clause, then we can only have OPEN statement operated on all table rows, not a subset of it.
fix the problem raised by Paul. But does not provide 'Remove SQL statement' for OPEN statement since the For Cluase is removed
verified and closed
