Re: Cannot override `remote.origin.url` with `-c` option

[Junio C Hamano <gitster@pobox.com>]

Jeff King <peff@peff.net> writes:

> Of course that leaves two questions:
>
>   1. What are multiple URLs actually good for? I have no idea. I

It came from days back when there weren't a separate push URL IIRC.
You may pull from a central place like everybody else, but you may
have other publishing points of your work.

Place to fetch must be a single known stable place for obvious
reasons.  "I have N, so I'll pull from one chosen at random among
these N" would lead to madness ;-).

>   2. Is there a way to override the list?
>
>      Sadly, no. For some config keys, we allow a value-less boolean
>      entry to reset the list.

Hmph, I somehow thought that a more widely used "clear" was an empty
string, but if we can make the convention around the "I exist and I
mean true" entries, that would be great.  It would not make much
sense to have multi-valued Boolean variable to begin with.

> Of course none of that helps your immediate case. I did think of one
> workaround, though, which is to use the "insteadOf" config to rewrite
> the URL. So:
>
>   git -c url.right-url.insteadOf=wrong-url ...
>
> will rewrite all instances of "wrong-url" to use "right-url" instead (in
> origin and elsewhere).

Yup, that is an excellent idea.

> diff --git a/remote.c b/remote.c
> index dcb5492c85..69b0f28637 100644
> --- a/remote.c
> +++ b/remote.c
> @@ -63,6 +63,10 @@ static const char *alias_url(const char *url, struct rewrites *r)
>  
>  static void add_url(struct remote *remote, const char *url)
>  {
> +	if (!url) {
> +		remote->url_nr = 0;
> +		return;
> +	}
>  	ALLOC_GROW(remote->url, remote->url_nr + 1, remote->url_alloc);
>  	remote->url[remote->url_nr++] = url;
>  }
> @@ -430,10 +434,7 @@ static int handle_config(const char *key, const char *value,
>  	else if (!strcmp(subkey, "prunetags"))
>  		remote->prune_tags = git_config_bool(key, value);
>  	else if (!strcmp(subkey, "url")) {
> -		char *v;
> -		if (git_config_string(&v, key, value))
> -			return -1;
> -		add_url(remote, v);
> +		add_url(remote, xstrdup_or_null(value));
>  	} else if (!strcmp(subkey, "pushurl")) {
>  		char *v;
>  		if (git_config_string(&v, key, value))

I was expecting (with excitement) a mess, but the above is as clean
as we can make the idea, I would say.  Lack of documentation and
tests do count as incompleteness though of course.