On 2020-09-10 at 17:01:52, Junio C Hamano wrote:
> Here is an update, after seeing Peff's review.
> 
> The overall structure of the series stays the same.
> 
>  * The second patch lost an incorrect use of 'flags' parameter down
>    to quote_c_style_counted() from quote_path().  In the function
>    declaration of quote_path(), the flags parameter is now named.
> 
>  * The third patch that moves the "optionally quote path with SP in
>    it" logic to quote_path() is essentialy unchanged.
> 
>  * But the fourth patch rewrites the implementation to avoid
>    shifting the output bytes with insertstr().
> 
>  * The fifth patch (formerly the fourth) that teachs "status --short"
>    to consistently quote hasn't changed, except that we test also
>    for the ignored paths.
> 
>  * The last two patches are unchanged.  They are optional clean-ups
>    that are not required.
> 
> We might want to add a bit more tests for:
> 
>  - how conflicted "funny" paths are shown.
> 
>  - how doubly funny paths (those that quote_c_style_counted() needs
>    to use backslash quoting *and* that contain SP) are shown.
> 
> but I'd say that is outside the scope of this round.  Seeing what is
> done in t3300, the latter test cannot be written portably as far as
> I can tell.

I took a look and this seems like a great improvement.  Thanks for
picking up where my patch left off.
-- 
brian m. carlson: Houston, Texas, US