[Progress Communities] [Progress OpenEdge ABL] Forum Post: RE: Double jeopardy when a NO-UNDO temp table violates a uniqueness constraint

Status
Not open for further replies.
D

dbeavon

Guest
Thanks jamesmc. Neither of the workarounds seems that appealing. Switching to an UNDO'able temptable will incur unnecessary overhead. We've gotten to the point where all our TT's are NO-UNDO by convention (similar to our memory variables). Deleting the default record buffer in a catch block is also a very questionable workaround... especially if you are doing it in a generic catch block ... in an outer scope that is not supposed to know about the internal implementation details of any methods that are called. Perhaps the implementation of ABL'S "EMPTY-TEMP-TABLE" operation should be responsible for deleting the default buffer as a first step before doing the rest of the work the normal way. That seems fairly straightforward. And it would unburden us from micro-managing things at that level. Or maybe there should be a way to "unbind" the temp-table, essentially flagging it as an unused and uninteresting object for the current session & program (similar to the way a "REFERENCE-ONLY" TT is uninteresting before it is assigned a reference to the instance of another TT). If a temp-table was "unbinded" , then ABL wouldn't need to validate it since it was abandoned anyway.

Continue reading...
 
Status
Not open for further replies.
Top