On Tue, 2016-05-10 at 22:09 +0200, Bernhard Reutner-Fischer wrote: > On Mon, May 09, 2016 at 08:14:47PM -0400, David Malcolm wrote: > > > - inform (loc, "valid arguments to %qs are: %s", option > > ->opt_text, s); > > + const char *hint = find_closest_string (arg, &candidates); > > + if (hint) > > + inform (loc, "valid arguments to %qs are: %s; did you mean > > %qs?", > > + option->opt_text, s, hint); > > + else > > + inform (loc, "valid arguments to %qs are: %s", option > > ->opt_text, s); > > + > > btw.. did you consider a format specifier for this ";did you mean > %qs" > common stanza? > > inform (loc, "valid arguments to %qs are: %s%qhs", x, y, hint); > (let's say) where the specifier would end up empty if hint was NULL.
It would save a conditional, but it seems a bit too much like "magic" to me; I prefer the explicitness of the existing approach. > I don't really like all these odd conditionals on the hints and it > also > adds some bloat to have all these duplicate error strings. > > Not sure but maybe this would even help translators? I'm not sure either, but it's generally a bad idea from an i18n perpective to be building strings at run-time, and the %qhs idea above feels a bit like that to me. > Is there a -Wno-spelling-suggestions to turn the suggestions off > globally, btw? There isn't one. Dave
