[bug-gnulib] Re: [PATCH]: fix warning in the hash module

[Bruno Haible]

Jim Meyering wrote:
> What if we were to convert to the const-less signatures you prefer,
> and then an application wants to call hash_insert with a variable of
> type `char const *'? -- of course, the key-freer function is NULL
> in that case.  Then, it's not just the relatively obscure -Wcast-qual
> that evokes a warning, but instead it's gcc's more commonly used
> -W option:
>
>   warning: passing argument 2 of 'hash_insert' discards qualifiers
>     from pointer target type
>
> And since there are applications that do precisely that, I prefer
> to incur a little bit of technically unclean code under the covers
> than to require every such caller to cast-away their perfectly
> reasonable `const' attributes.

Agreed. It's more important for an interface (hash.h) to be usable without
warnings in the common cases, than for its implementation (hash.c) to be
without warnings. Here the common cases are:

   1) All keys are 'const something *'.
   2) All keys are 'something *'.

Because of 1, the argument type must be 'const void *'.
Because of 2, the return type of hash_lookup() must be 'void *'.

Same reasoning as for strchr(), memchr(), etc.

Bruno