Re: [PATCH 2/3] builtins: utilize startup_info->help where possible

From: Nguyen Thai Ngoc Duy <pclouds@gmail.com>
Date: Oct 20, 2010, 01:13 UTC
Message-ID: <AANLkTinvSKAyXkzjUgxvrDWvuHWpa+46XVxTu=tWck53@mail.gmail.com>
In-Reply-To: <20101019172953.GC25139@burratino>

2010/10/20 Jonathan Nieder <jrnieder@gmail.com>:

Show 9 quoted lines

> Nguyễn Thái Ngọc Duy wrote:
>
>> It helps reduce false alarms while I'm looking for "git foo -h" code
>> path that accesses repository.
>
> Not sure I understand.  Is the idea that use of startup_info->help
> is a marker for "I've checked this code path"?
>
> If that were the only reason, I don't think I'd like the idea.

That. And simpler check. I mean "if (startup_info->help)" takes a tiny bit less energy for me to understand than "if (argc == 2 && argv[1] == "-h")". It's also good for grepping.

> As it is, I'm a bit conflicted: what if we decide to short-circuit
> "git foo --help-all" in the future just like we short-circuit
> "git foo -h" now?  Would that require a separate flag?

As long as it's help related, startup_info->help can be turned to a bit set. Although I'm not sure if --help-all or any other option will be implemented anytime soon. Remember it has to be implemented for _all_ builtin commands, or we need to introduce NO_DASH_H in run_builtin() to skip the shortcut for commands that do not support it.

Show 13 quoted lines

> In other words, I'm not sure startup_info->help is a good abstraction.
> Maybe (modulo names) it would be better to do
>
> struct startup_info {
>        ...
>        const char *short_circuit;      /* "-h", "--help-all", "--no-index", or NULL */
> };
>
> and use
>
>        if (!strcmp(startup_info->short_circuit, "-h"))
>
> to allow relaxing the argc == 2 check later?

Can it be relaxed later when someone comes up with "--help-all" or something?

-- 
Duy