that I am trying to split into two different datasets, one is the top three rows - where roleId = 25, 26, 24, and the second dataset is the remainder of the list.
I was looking at the the example project but only found one where clause example like Select t_stamp, Ramp Where _r>(_from.rowCount-100) but it’s not exactly what I need.
I know I could currently do it with a LIMIT clause but a new role is something we could realistically add and I don’t doing that to mess things up so I’d rather select exacltly the rows I need.
How can I select exactly the rows I need and then how to select the remainder rows?
Only semi-related but I don’t suppose there’s more plans to expand Simulation Aids?
What would be great imo is a union keyword that combines two datasets (assuming same number of column/column types). I have situations where on the the same window I will have a dropdown list that needs to be selected by a user so the first option will be None and then the list of options, and on the same window a dropdown Filter based on those same options which needs an All as the first option before the actual options. Right now I do that with a named query with a string parameter that populates the first row with None or All or whatever is needed, but then the named query still needs to be called twice to populate both dropdowns.
Oh, there are always ideas percolating. Union is ugly. I wouldn’t make it part of view(), but a separate function. Join is the one I keep thinking about, but it is even more ugly.
What is most likely to happen first is an alternate version of view() that leverages Ignition’s expression machinery to execute off of an object tree instead of jython. I expect such a beast to be orders of magnitude faster than the current function. It could also implement some syntax closer to true SQL, like auto-grouping when aggregates are present in the Select list, or an efficient implementation of mixed ascending and descending ordering.
Yea I was thinking it would be an expression like union({Root container.datset1}, {Root Container.dataset2}, ...) would be the usage.
I expect such a beast to be orders of magnitude faster than the current function. It could also implement some syntax closer to true SQL, like auto-grouping when aggregates are present in the Select list, or an efficient implementation of mixed ascending and descending ordering.
