Re: [PATCH v3 4/4] progress.c: silence cgcc suggestion about internal linkage

From: Johannes Schindelin <johannes.schindelin@gmx.de>
Date: May 1, 2020, 20:09 UTC
Message-ID: <nycvar.QRO.7.76.6.2005012206300.18039@tvgsbejvaqbjf.bet>
In-Reply-To: <d66d9aa677944205cdd5af77192c35ec929692ea.1587996766.git.congdanhqx@gmail.com>

Hi,

On Mon, 27 Apr 2020, Đoàn Trần Công Danh wrote:

Show 19 quoted lines

> diff --git a/progress.h b/progress.h
> index 847338911f..f1913acf73 100644
> --- a/progress.h
> +++ b/progress.h
> @@ -3,6 +3,14 @@
>
>  struct progress;
>
> +#ifdef GIT_TEST_PROGRESS_ONLY
> +
> +extern int progress_testing;
> +extern uint64_t progress_test_ns;
> +void progress_test_force_update(void);
> +
> +#endif
> +
>  void display_throughput(struct progress *progress, uint64_t total);
>  void display_progress(struct progress *progress, uint64_t n);
>  struct progress *start_progress(const char *title, uint64_t total);

A slightly cleaner strategy would be to add a separate header file, say, `progress-internal.h` or `internal/progress.h` to declare the variables and the function.

I guess at some stage we really need to define a proper API of libgit.a. Just because we refuse to declare such an API doesn't mean that people don't use libgit.a in that way (e.g. cgit, cinnabar). We might just as well face reality and split up `cache.h` and put the API part into header files in a subdirectory `include/`.

Ciao, Dscho