Richard, I think the confusion is between OVERRIDE and OVERLOAD, and in what cases defining a function is an complete override of the function (and all its pre-existing overloaded implementations), and in what cases it is merely an OVERLOAD of the function name.
And of course whether it is possible to override an overloaded implementation ... versus just adding a new overloaded implementation but leaving already declared implementations intact. > If you call sqlite3_create_function_v2() with a function name that is > the > name of a built-in function, then the built-in function goes away and is > replaced by your application-defined function. The original built-in > function is no longer accessible. *Any* built-in function can be > overloaded in this way. > > > On Wed, Sep 11, 2013 at 4:35 PM, Stephan Beal <sgb...@googlemail.com> > wrote: > > > Hi, all, > > > > i'm looking for a clarification on what is certainly a bit of pedantry > on > > my part: > > > > http://www.sqlite.org/c3ref/create_function.html > > > > specifies that we can overload built-in funcs with UDFs: > > > > "Built-in functions may be overloaded by new application-defined > > functions." > > > > Does "overload" imply "override" if the name/arg count/encoding/state > > match, or is it an error to override a function? The docs don't seem > to > > explicitly mention the (admittedly unusual) exact-match case, but the > > paragraph above that one seem to be intended that a name/arg-count > overload > > is an error: > > > > "It is permitted to register multiple implementations of the same > functions > > with the same name but with either differing numbers of arguments or > > differing preferred text encodings." > > > > > > i've been on this list long enough to know that someone out there is > going > > to ask, "why would you do that?" In brief: in porting the Fossil SCM > to a > > library API i need to use a different approach to how it overrides > > localtime() with its own variant (it uses a C macro to replace > localtime() > > with fossil_localtime(), which uses app-global state, whereas i have a > > library API and thus local state), and currently overriding it, as > opposed > > to overloading it, is what i'm aiming to do (when the time comes to > port > > those bits, which isn't tonight). But... i won't if you guys can tell > me in > > advance that it will fail. > > > > Happy Hacking! > > > > -- > > ----- stephan beal > > http://wanderinghorse.net/home/stephan/ > > http://gplus.to/sgbeal > > _______________________________________________ > > sqlite-users mailing list > > sqlite-users@sqlite.org > > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users > > > > > > -- > D. Richard Hipp > d...@sqlite.org > _______________________________________________ > sqlite-users mailing list > sqlite-users@sqlite.org > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
