On Fri, Oct 25, 2019 at 04:42:24PM +0900, Amit Langote wrote:
> Hmm, if we're inventing a new API to replace the old one, why not use
> that opportunity to be consistent with our general style, which
> predominantly seems to be either words_separated_by_underscore() or
> UpperCamelCase(). Thoughts?
Not wrong. Using small-case characters separated with underscores
would be more consistent with the rest perhaps? We use that for the
initialization of custom variables and for all the relkind-related
interfaces.
> You might know already, but in short, the values in the passed-in
> relopt_parse_elts array (relopt_elems) must fit within
> relopt_struct_size. Writing an Assert turned out to be tricky given
> that alignment must be considered, but I have tried to add one. Pleas
> check, it very well might be wrong. ;)
Hmm. I didn't expect it to be this confusing with relopt_type_size[].
I'll try to think about something :(
+ * Parses reloptions provided by the caller in text array format and
+ * fills and returns a struct containing the parsed option values
The sentence structure is weird, perhaps:
This routine parses "reloptions" provided by the caller in text-array
format. The parsing is done with a table describing the allowed
options, defined by "relopt_elems" of length "num_relopt_elems". The
returned result is a structure containing all the parsed option
values.
> Attached updated patch. It would be nice to hear whether this patch
> is really what Nikolay intended to eventually do with this code.
Okay, let's check if Nikolay likes this idea.
--
Michael
