tech-kern archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

C hackery [was Re: Some changes to autoconfiguration APIs]

>> [...] statement expression [...]

> I think statement expressions can be a rather "dangerous"
> complication in C

I am moderately sure statement expressions don't exist in C, at least
as of C99; I believe them to be a gccism.  (I have never seen a spec
for any later C, though I have seen it said they've included things I
think have no business in C, such as closures, so maybe.)

Many gccisms can be dangerous, indeed.  Many of them - the same ones,
in some cases - can also be extremely useful and clarifying.

> I've only ever found them to be truly useful within a macro when I'm
> trying to avoid, or do something different than, the "usual
> promotions".

I find them, combined with nested functions, useful in that the result
can be close to lambda functions:

	n = map_sum_over(list, ({ int foo(struct thingy *x)
				   { ...
				  &foo; }));

However, some of the gcc variants I use produce either broken code or
broken debugging info when faced with that, so these days I am more
likely to make the function a nested function within a containing
block, creating a small block if necessary to keep point of definition
close to point of use.

> Kind of related to this, I have the following comment in my notes
> about C:

> - Positional parameters are evil (or at least error prone),
>    especially for variable numbers of parameters.

Personally, I think it's less "for variable numbers of parameters" and
more "for large numbers of parameters".

> Named parameters can be simulated in modern C with full structure
> passing: the price of defining a special-purpose struct for each such
function, or at least each such argument signature, and losing
unused-argument warnings (possibly among others, though none others
come to mind immediately).

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B

Home | Main Index | Thread Index | Old Index