Re: [PATCH 7/7] t0000: verify that real_path() removes extra slashes

From: Junio C Hamano <gitster@pobox.com>
Date: Sep 6, 2012, 17:34 UTC
Message-ID: <7vharb3vib.fsf@alter.siamese.dyndns.org>
In-Reply-To: <20120906054407.GA25981@duynguyen-vnpc.dek-tpc.internal>

Nguyen Thai Ngoc Duy <pclouds@gmail.com> writes:

Show 21 quoted lines

> Just an idea. We could unify "[a-z]:" and "//host" into "dos root"
> concept. That would teach other code paths about UNC paths too.
> ...
> diff --git a/path.c b/path.c
> index 66acd24..0e4e2d7 100644
> --- a/path.c
> +++ b/path.c
> @@ -498,11 +498,12 @@ const char *relative_path(const char *abs, const char *base)
>  int normalize_path_copy(char *dst, const char *src)
>  {
>  	char *dst0;
> +	int i, len;
>  
> -	if (has_dos_drive_prefix(src)) {
> +	len = offset_1st_component(src, 1);
> +	for (i = 0; i < len; i++)
>  		*dst++ = *src++;
> -		*dst++ = *src++;
> -	}
> +
>  	dst0 = dst;

Modulo that I suspect you could get rid of offset_1st_component() altogether and has_dos_drive_prefix() return the length of the "d:" or "//d" part (which needs to be copied literally regardless of the "normalization"), what you suggest feels like the right approach. Why do you need the "keep_root" parameter and do things differently depending on the setting by the way? Wouldn't "skip the root level when computing the offset of the first path component" something the caller can easily decide to do or not to do, and wouldn't it make the semantics of the function cleaner and simpler by making it do only one thing and one thing well?