On Sun, Mar 13, 2022 at 07:34:11PM +0100, Omar Polo wrote:
> Hello,
> 
> the following diff adds a dry-run mode for got patch.  In this mode the
> patch is only faked and the changes are not really done to the work
> tree.
> 
> I chosen -n because it stands for "nop" which I like.  For comparison,
> Subversions seems to have --dry-run, git-apply(1) --check and patch(1)
> has -C, --dry-run and --check.  (I don't really like to type capital
> letters ;-)
> 
> It carries the same issues that patch(1) has: if the patchfile contains
> multiple diffs that affect the same file, got patch -n could report a
> failure even if 'got patch' would succeed.
> 
> (eventually we could sort the patches and try to apply them to the same
> temp file to handle also this case, but I'm not sure and this is a start
> anyway)

Good idea, I like it.

Some man page input below.

> @@ -1333,6 +1333,19 @@ Such cahnges can be viewed with
>  and can be reverted with
>  .Cm got revert
>  if needed.
> +.Pp
> +The options for
> +.Cm got patch
> +are as follows:
> +.Bl -tag -width Ds
> +.It Fl n
> +Check that the patch would apply cleanly but don't do any edits to the
> +worktree.

How about:

".It Fl n
Do not make any modifications to the work tree.
This can be used to check whether a patch would apply without issues."

> +Note that if
> +.Ar patchfile
> +contains diffs that affect the same file multiple times

Note that, in manual pages, "Note that" is discouraged ;)
https://marc.info/?l=openbsd-tech&m=163770139630230&w=2

"If the
.Ar patchfile
contains ...  " is better.

In general, man pages should never address the reader in first person.

> the results won't
> +likely be correct.

"... the results displayed may be incorrect." ?

> +.El