On 2021-05-11 at 10:37:30, Eric Wong wrote:
> I'm not very familiar with the way some of this stuff works, but
> the patch below seems to fix clone for me.
> 
> I originally tried changing the cmd_main->set_option code path,
> but value is always set to "true" for object-format because
> it only sees "option object-format" with no arg
> 
> 		} else if (skip_prefix(buf.buf, "option ", &arg)) {
> 			char *value = strchr(arg, ' ');
> 			int result;
> 
> 			if (value)
> 				*value++ = '\0';
> 			else
> 				value = "true";
> 
> 			result = set_option(arg, value);
> 
> So when set_option gets called, hash_algo_by_name isn't:
> 
> 	} else if (!strcmp(name, "object-format")) {
> 		int algo;
> 		options.object_format = 1;
> 		if (strcmp(value, "true")) {
> 			/* XXX this branch is never taken: */
> 			algo = hash_algo_by_name(value);
> 			if (algo == GIT_HASH_UNKNOWN)
> 				die("unknown object format '%s'", value);
> 			options.hash_algo = &hash_algos[algo];
> 		}
> 		return 0;
> 
> So I'm not sure if the above is incomplete or dead code.
> Anyways, I arrived at the following and it works for me:
> 
> -----------8<---------
> Subject: [PATCH] remote-curl: fix clone on sha256 repos
> 
> The remote-https process needs to update it's own instance of
> `the_repository' when it sees an HTTP(S) remote is using sha256.
> Without this, parse_oid_hex() fails to handle sha256 OIDs when
> it's eventually called by parse_fetch().

This seems fine as a solution for now.  I tried to keep the transport
code mostly independent of the local repository settings, but in this
case because the HTTP walker mucks around with the internals of the
local pack files, I don't think we can avoid this without some major
restructuring, which I'm not interested in sitting down and writing this
evening.

I'll clean this up in a nicer way once I get interop working.  Thanks
for sending a patch for this.
-- 
brian m. carlson (he/him or they/them)
Houston, Texas, US