On 29.10.24 09:47, Peter Eisentraut wrote:
I kind of wonder if we really want to do this. It adds no functionality, and it forecloses the possibility of changing the definition later. I understand and agree with your conclusion that it's pretty much impossible to do what the SQL standard suggests should happen --- but maybe we're both missing something that would make it feasible. (Have you asked your committee colleagues if anyone's actually implemented what they wrote about SIMILAR TO? If they've written something unimplementable, it seems like there is work for them to do in any case.)Good idea; I'll go ask there too.
So the result from that was that no one there knew what to do either. There was general interest in the various arguments and options, but there was no consensus about what the right solution should be.
For everyone's amusement, attached is the discussion paper I submitted, which contains some of my arguments from this thread as well as other information and examples.
I think a way forward would be to define more special purpose collations that are just "normal but case insensitive" or "normal but accent insensitive", like was discussed later in this thread, and what other implementations apparently also do (see BINARY_CI in the paper).
For now, I'm withdrawing this patch, but I (and I suspect others) will keep thinking about this.
w28015-Pattern matching versus collations.pdf
Description: Adobe PDF document
