Thanks Emmit,
With the variable in question I've tried declaring it with a REAL type
without setting it to a value (this should give it a null value),
initializing it to 0, and not decalaring it (it gives an error - variable not
found). The first cursor FETCHes values into the variable OK. The 2
other variables are populated by a variable form and receive valid
values.
Geoffrey
Geoffry,
Before you declare your cursors, declare and type all variables that will be
used within the cursor definitions, setting them to NULL.
Emmitt Dove
I have an Rbase 3.1 command file that I've updated to 7.6
I'm having difficulty with one cursor used in a nested WHILE.
I declare 2 cursors. The first cursor is opened before and fetched
before and in the outside while loop. The second cursor is OPEN
RESET before and fetched before and in the nested WHILE.
The first cursor is FETCHed OK every loop. The second
cursor never has any data (SQLCODE=100). I can never FETCH
values into my variables.
I checked to see what was happening with the cursor. I can look at
variables in the Watch Box but not cursors directly. And because I
couldn't FETCH with the cursor (SQLCODE=100) I couldn't examine
the 3 variables that it should have populated. So, I closed trace after I
had stepped through the loop a couple of times. The LIST CURSOR
command at the R>prompt then showed something very strange about
the second cursor.
The second cursor has 3 variables in the SELECT part of the
DECLARE statement. The LIST cursor command showed in the
cursor definition that one variable remained as a .vVariableName, one
had a constant text value and the other was set to -0- (Null). The one
set to NULL is the variable FETCHed by the first cursor in the
outside loop. I would expect the OPEN cursor2 RESET command
would re-evaluate the variables in the DECLARE statement. In this
case only one is re-evaluated because the other 2 have been set in
concrete by being evaluated in the DECLARE statement. What is
going on here? How can I get the variables in my second cursor to be
re-evaluated?
Thanks,
Geoffrey
