Commit Graph

7 Commits

Author SHA1 Message Date
Rusty Russell
b14cc0c9f7 lightningd/params: fix typesafe check.
typesafe_cb isn't suitable here, as it is simply a conditional cast,
and the result is passed through '...' and doesn't matter.

Reported-by: @wythe
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-07-07 23:49:25 +00:00
Mark Beckwith
7d9ad89010 params: removed the param_opt_tok macro
There doesn't seeem to be a need for this anymore (unless I'm missing something).
I added the sendpay_nulltok() unit test to confirm.

Signed-off-by: Mark Beckwith <wythe@intrig.com>
2018-07-05 00:19:16 +00:00
Mark Beckwith
1b50ea2abd params: removed tal context.
@rustyrussell showed we don't need temporary objects for params.

This means params no longer need a tal context.

Signed-off-by: Mark Beckwith <wythe@intrig.com>
2018-07-05 00:19:16 +00:00
Rusty Russell
6ff901d7b0 params: simplify lifetimes of params.
@wythe points out we don't need to keep the around now param_is_set()
is removed.  We can in fact go further and avoid marshalling them into
temporary objects at the caller altogether.

This means internally we have an array of struct param, rather than an
array of 'struct param *', which causes most of the noise in this
patch.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-07-05 00:19:16 +00:00
Rusty Russell
3f6f9e6fe0 param: make sure the name is a string literal.
We're using a macro anyway, so appending "" make it a compile-time check.

Complicates testing a bit, since we actually use generated names there.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-07-05 00:19:16 +00:00
Rusty Russell
9f83a9ae4d params: make optional args do allocation for you.
This is a bit more natural, IMHO.  The only issue is that json_tok_tok is
special, so we end up with param_opt_tok() if you really want an optional
generic token.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-07-05 00:19:16 +00:00
Mark Beckwith
4d1d0438e1 Typesafe callback system for parsing json
This is part of #1464 and incorporates Rusty's suggested updates from #1569.

See comment in param.h for description, here's the basics:

	unsigned cltv;
	const jsmntok_t *note;
	u64 msatoshi;
	struct param * mp;

	if (!param_parse(cmd, buffer, tokens,
			 param_req("cltv", json_tok_number, &cltv),
			 param_opt("note", json_tok_tok, &note),
			 mp = param_opt("msatoshi", json_tok_u64, &msatoshi),
			 NULL))
		return;

	if (param_is_set(mp))
		do_something()

There is a lot of developer mode code to make sure we don't make mistakes,
like trying to unmarshal into the same variable twice or adding a required param
after optional.

During testing, I found a bug (of sorts) in the current system.  It allows you
to provide two named parameters with the same name without error; e.g.:

	# cli/lightning-cli -k newaddr addresstype=p2sh-segwit addresstype=bech32
	{
		  "address": "2N3r6fT65PhfhE1mcMS6TtcdaEurud6M7pA"
	}

It just takes the first and ignores the second.  The new system reports this as an
error for now.  We can always change this later.
2018-07-05 00:19:16 +00:00