From: Tian Yuchen Date: Fri, 13 Mar 2026 17:15:23 GMT Subject: Re: [PATCH v1] builtin/mktree: remove USE_THE_REPOSITORY_VARIABLE Message-ID: <4fb9c915-7246-4c55-b7c6-b4ef7ca91230@malon.dev> In-Reply-To: On 3/14/26 00:03, Junio C Hamano wrote: > There is one corner case; upon "git foo -h", your cmd_foo() will get > repo==NULL when the command is run outside a repository. As long as > your cmd_foo() asks parse_options() to react to "-h" (which gives > the help message and then exits) before it uses repo assuming it > cannot be NULL, you are safe. I was completely blown away Σ( ° △ °) Thanks for pointing that out. Otherwise, I wouldn’t have been able to figure it out no matter how hard I try. I just took a quick look at the code: > const struct option option[] = { > OPT_BOOL('z', NULL, &nul_term_line, N_("input is NUL terminated")), > OPT_SET_INT( 0 , "missing", &allow_missing, N_("allow missing objects"), 1), > OPT_SET_INT( 0 , "batch", &is_batch_mode, N_("allow creation of more than one tree"), 1), > OPT_END() > }; > > ac = parse_options(ac, av, prefix, option, mktree_usage, 0); > getline_fn = nul_term_line ? strbuf_getline_nul : strbuf_getline_lf; > > while (!got_eof) { > while (1) { ... I think if there's a '-h' parameter, it gets intercepted in parse_options() and the process exits before repo is called. So there’s nothing to worry about, right? By the way, I find it a bit confusing that the'`-h' parameter — which is solely used for documentation query — is parsed and intercepted within a function that handles actual business logic. Wouldn’t it be more appropriate to intercept it at a higher level? Is this a technical debt? I don’t intend to write a patch to fix this, but this parameter has made me realize that this odd way of coding (at least in my understanding) is highly unpredictable for someone with limited experience like me. I don't know if there are any troubleshooting methods other than asking someone with experience. Regards, Yuchen