17.06.2016 17:13, Dimitry Sibiryakov wrote: > > As most of you know, when routine protocol.cpp:alloc_cstring() receives zero > length, it > behaves differently, depending on previous allocations. In some cases it > returns pointer > to zero-length string, in some - NULL pointer. > Because of this, used in subj test query "EXECUTE STATEMENT ''" can throw > different > errors. If the routine has returned NULL, user will get code 335544608 from > Y-valve, > otherwise it will be set of 335544569, 335544436, 335544851 from DSQL. > In first case the test will fail, as it always expects the second case. > > Is it a problem in the test or in the server?
Wouldn't it be better to check for empty SQL string earlier, inside ExecStatementNode::execute(), and throw error directly from there, before passing it through the layers? Or perhaps do such a check in the remote client, or even better in the Y-valve? Dmitry ------------------------------------------------------------------------------ What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic patterns at an interface-level. Reveals which users, apps, and protocols are consuming the most bandwidth. Provides multi-vendor support for NetFlow, J-Flow, sFlow and other flows. Make informed decisions using capacity planning reports. http://sdm.link/zohomanageengine Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
