On Tue, Nov 3, 2026 at 1:16 AM David G. Johnston <[email protected]> wrote: > > On Mon, Nov 2, 2026 at 8:47 PM Matemática A3K <[email protected]> > wrote: >> >> To inspect if each function parameter is documented correctly for each >> function of each supported release seems like a gargantuan task to be >> done. > > > We're pretty good about sgml markup in this area, it doesn't seem that hard > to build a deterministic detection system if someone wants to try. I'd make > it require an installed instance and use the catalog as the authoritative > dataset.
I can'f figure that code in my mind, only manual checking, it would be nice to see the code that automates it. > >> >> Perhaps a note in somewhere that says "named parameters in functions >> are supported but may not be up to date, using positional arguments is >> the recommended way for function calling"? >> > > I don't think we actually recommend that. Telling people how to find the > names of function parameters should they wish to use them and the docs aren't > correct would be helpful though. For procedure-like functions/procedures I'd > say named arguments are recommended. The feature works (calling a function with named arguments), it can't be used if the names documented doesn't match those defined in the code or the catalog. Stating this as caveat isn't neccesary if the correspondence is ensured. The patches attached go in this sense, so they should be accepted. If someone can provide the code you mentioned, then there wouldn't be anything more to be fixed here (besides the mismatches that code may find). > > David J. >
