Uh oh!
There was an error while loading. Please reload this page.
Pre-scoring test clean-ups - #1120
Conversation
jkczyz
commented
Oct 13, 2021
I left a few TODO comments in the primary commit where I was unsure of the change. So please take a close look at those. Everything else should be fairly mechanical. |
Codecov Report
@@ Coverage Diff @@## main #1120 +/- ##
==========================================
+ Coverage 90.67% 90.73% +0.06%
==========================================
Files 66 66 Lines 34732 34808 +76 ==========================================
+ Hits 31494 31584 +90 + Misses 3238 3224 -14
Continue to review full report at Codecov.
|
TheBlueMatt
commented
Oct 13, 2021
The "Fix need for #[allow(unused_mut)] in tests" commit is largely duplicative with "Use Persister to return errors in tests not chain::Watch " in #1108, is it possible to drop that commit here and use the one from 1108? |
49e7016 to
7719633Comparejkczyz
commented
Oct 13, 2021
Dropped the commit.
Reverted the marked changes to use |
| check_closed_event!(nodes[1], 1, ClosureReason::CooperativeClosure); | ||
| } | ||
| // TODO: Determine why get_route fails when replacing get_payment_preimage_hash with |
There was a problem hiding this comment.
You have another todo here.
There was a problem hiding this comment.
Mainly to note why it may be that a simple replacement doesn't work here. Can drop it if you prefer.
|
| let (_, second_payment_hash, second_payment_secret) = get_payment_preimage_hash!(nodes[2]); | ||
| let sending_node = if forwarded_htlc { &nodes[0] } else { &nodes[1] }; | ||
| let (route, second_payment_hash, _, second_payment_secret) = get_route_and_payment_hash!(sending_node, nodes[2], 100000); | ||
| // TODO: Why use a different payment secret here? |
There was a problem hiding this comment.
Cause there's two different payments?
There was a problem hiding this comment.
Was unsure because they both used second_payment_hash but a different secret.
There was a problem hiding this comment.
Ah, I was comparing it to the first send event further up, its just a typo I believe, feel free to drop it if the test still passes and always use the second secret.
There was a problem hiding this comment.
Updated to use the second secret.
| assert_ne!(node_txn[0].input[0].previous_output, node_txn[2].input[0].previous_output); | ||
| assert_ne!(node_txn[1].input[0].previous_output, node_txn[2].input[0].previous_output); | ||
| // TODO: Determine why replacing get_route with get_route_and_payment_hash above causes |
There was a problem hiding this comment.
Because the tx ordering is based on the payment hash and we end up with a different hash, I assume? Or something like that, but maybe we should just drop the TODO and open a general "move the remaining explicit route calls to the routing macro" issue?
valentinewallace
left a comment
There was a problem hiding this comment.
ACK mod removing the todos in favor of an issue + squash
Uh oh!
There was an error while loading. Please reload this page.
7719633 to
b6a3f5cCompareb6a3f5c to
d700faaCompare| let _chan = create_chan_between_nodes(&nodes[0], &nodes[1], InitFeatures::known(), InitFeatures::known()); | ||
| let network_graph = &nodes[0].net_graph_msg_handler.network_graph; | ||
| let first_hops = nodes[0].node.list_usable_channels(); | ||
| let route = get_keysend_route(&payer_pubkey, &network_graph, &payee_pubkey, |
There was a problem hiding this comment.
This seems to be the only place we call get_keysend_route in tests through a private channel, so I think best to keep it.
There was a problem hiding this comment.
Reverted and updated the other keysend test to use get_keysend_route.
d700faa to
3e9678cCompareThe interface for get_route will change to take a scorer. Using get_route_and_payment_hash whenever possible allows for keeping the scorer inside get_route_and_payment_hash rather than at every call site. Replace get_route with get_route_and_payment_hash wherever possible. Additionally, update get_route_and_payment_hash to use the known invoice features and the sending node's logger.
3e9678c to
d4ec090Compare
Replaces
get_routewithget_route_and_payment_hashin tests wherever possible, since a scorer will need to be passed toget_route. Also includes some fairly trivial test clean ups.