Hi Bruno, Paul and Reuben,
> Le 31 oct. 2020 à 20:41, Bruno Haible a écrit :\
>
> You are right that there is a problem here: On 2012-03-07 this patch [1]
> from Akim did a good thing — it made the quote.h functions' effect more
> customizable — and a not so good thing: it combined the implement
Akim Demaille wrote:
> I see gnulib already has both -private.h and -internal.h files. I don't know
> if this is a case where we'd want to create such a file.
I like the idea. Yes, moving the struct to a "quotearg-private.h" would
not accidentally reveal the details to the user who reads quotearg
On 11/1/20 4:07 AM, Bruno Haible wrote:
I like the idea. Yes, moving the struct to a "quotearg-private.h" would
not accidentally reveal the details to the user who reads quotearg.h.
Sounds good to me too. This way, we wouldn't need the QUOTING_OPTIONS_INIT
macro. And renaming quotearg_n_option
I wrote:
> If you see "options" as a generalization of "style" and "custom", then
> the systematic name should be 'quotearg_n_options_mem'. But then the
> functions which return their result in a buffer or freshly allocated
> should better be named
>
> quotearg_options_buffer instead of qu
Paul Eggert wrote:
> On 11/1/20 4:07 AM, Bruno Haible wrote:
> > I like the idea. Yes, moving the struct to a "quotearg-private.h" would
> > not accidentally reveal the details to the user who reads quotearg.h.
>
> Sounds good to me too. This way, we wouldn't need the QUOTING_OPTIONS_INIT
> macro
I am simultaneously amused, horrified, and slightly apologetic at what I
appear to have unleashed here :)
The patch introduced a compilation error on MSVC, because the declaration and
the definition of 'test_assume_noreturn' are not consistent. This patch should
fix it.
2020-11-01 Bruno Haible
verify tests: Fix compilation error with MSVC (regression 2020-10-30).
* tests/test-ver
Two more portability fixes are needed:
- On Linux/SPARC, UCHAR_MAX is not defined unless is included.
- On Linux/PowerPC, getpagetsize() returns 65536, which is unusually large.
2020-11-01 Bruno Haible
ssfmalloc tests: Portability to Linux/PowerPC and Linux/SPARC.
* tests
On 10/31/20 12:46 PM, Bruno Haible wrote:
> Bernhard Voelker wrote:
>> I noticed a new build failure in coreutils when configuring with
>> --enable-gcc-warnings.
>> GCC is 10.2.1.
>> Fixed with the attached.
>
> While Gnulib generally doesn't support compiling with '-Werror' — this is the
> optio
On 10/31/20 7:09 PM, Bruno Haible wrote:
> In a gnulib testdir, I now see this warning:
>
> In file included from ../../gltests/test-verify.c:22:0:
> ../../gltests/test-verify.c: In function ‘main’:
> ../../gltests/../gllib/verify.h:316:25: warning: ‘((unsigned char*)&s)[0]’ is
> used uninitializ
10 matches
Mail list logo
