Refactor `ConstructedTransaction` to contain `Transaction` by jkczyz · Pull Request #4097 · lightningdevkit/rust-lightning

jkczyz · 2025-09-19T19:26:05Z

ConstructedTransaction tracks the result of the InteractiveTxConstruction and is used during an InteractiveTxSigningSession. This PR refactors it such that it holds an actual Transaction rather than the contributed parts, though some metadata in the form of NegotiatedTxInput and NegotiatedTxOutput need to be maintained.

This allows us to re-use the weight calculation provided by Transaction. It also facilitates reconstructing contributed inputs and outputs to produce a SpliceFailed event in #4077.

ldk-reviews-bot · 2025-09-19T19:26:08Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

codecov · 2025-09-19T19:37:15Z

Codecov Report

❌ Patch coverage is 92.20183% with 17 lines in your changes missing coverage. Please review.
✅ Project coverage is 88.71%. Comparing base (3564646) to head (86ed012).
⚠️ Report is 109 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/interactivetxs.rs	93.33%	11 Missing and 3 partials ⚠️
lightning/src/ln/channel.rs	25.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4097 +/- ##
==========================================
+ Coverage 87.81% 88.71% +0.89% 
==========================================
Files 176 177 +1 Lines 131770 133237 +1467 Branches 131770 133237 +1467 ==========================================
+ Hits 115719 118196 +2477 + Misses 13416 12326 -1090 - Partials 2635 2715 +80

Flag	Coverage Δ
fuzzing	`21.74% <0.54%> (+0.14%)`	⬆️
tests	`88.55% <92.20%> (+0.89%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2025-09-19T19:43:19Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


Wondering if we should have InteractiveTxSigningSession store both local and remote signatures until finalize_funding_tx is called. That way, ConstructedTransaction would always contain an unsigned transaction instead of one that may be partially signed.

SGTM, we already store the local ones separately

InteractiveTxConstructor::new converts inputs given in channel.rs to InputOwned. Instead of converting FundingTxInput in channel.rs to the type needed by that constructor, pass in FundingTxInput and have it converted directly to InputOwned there. This saves one conversion and Vec allocation. It also lets us use the satisfaction_weight in an upcoming refactor.

Instead of estimating the input weight, track the satisfaction_weight in InteractiveTxInput. This then can be used compute the transaction weight without storing individual weights in NegotiatedTxInput in an upcoming commit.

The weight is no longer needed once a ConstructedTransaction is created, so it doesn't need to be persisted.

Instead of defining a custom function for calculating a transaction's weight, have ConstructedTransaction hold a Transaction so that its weight method can be re-used. As a result NegotiatedTxInput no longer needs to store the transaction inputs.

jkczyz · 2025-09-19T20:06:11Z

Pushed rustfmt fixes.

ldk-reviews-bot · 2025-09-22T00:00:42Z

🔔 1st Reminder

Hey @wpaulino! This PR has been waiting for your review.
Please take a look when you have a chance. If you're unable to review, please let us know so we can find another reviewer.

wpaulino · 2025-09-22T18:38:59Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


SGTM, we already store the local ones separately

ldk-reviews-bot · 2025-09-22T18:44:16Z

👋 The first review has been submitted!

Do you think this PR is ready for a second reviewer? If so, click here to assign a second reviewer.

Now that NegotiatedTxInput no longer holds a TxIn, rename it accordingly.

Now that ConstructedTransaction holds the actual Transaction, there's no need to keep track of and persist the outputs separately. However, the serial IDs are still needed to later reconstruct which outputs were contributed.

The output values are never used so there's no need to persist them. If needed, they can be re-computed though the shared output's local and remote amounts would be lost.

The local and remote input values are used to determine which node sends tx_signatures first. Instead of persisting these values, compute them only when needed from the input metadata. The spec states that the entire shared input value is to be included for the node sending the corresponding tx_add_input, so it isn't necessary to know the local and remote balances which the metadata does not contain.

InteractiveTxSigningSession currently persists holder witnesses directly, but persists counterparty witnesses as part of its unsigned ConstructedTransaction. This makes the ConstructedTransaction actually partially signed even though it is held in a field named unsigned_tx. Instead, persists the counterparty witnesses alongside the holder witnesses directly in InteractiveTxSigningSession, leaving the transaction it holds unsigned.

TheBlueMatt

Code changes look fine to me. I'm not deep enough in the interactive negotiation world to say whether its all a good idea, but seems fine to me.

wpaulino · 2025-09-23T21:08:16Z

 .as_ref()
 .map(|shared_input| SharedInputSignature {
 holder_signature_first: shared_input.holder_sig_first,
- counterparty_signature: None,


Should probably rename SharedInputSignature since it doesn't contain a signature anymore

wpaulino · 2025-09-23T21:10:06Z

+ if let Some(shared_input_index) = self.shared_input_index {
+ let holder_shared_input_sig =
+ holder_tx_signatures.shared_input_signature.or_else(|| {
+ debug_assert!(false);


I think it'd be better to debug_assert_eq!(shared_input_index.is_some(), holder_tx_signatures.shared_input_signature.is_some()) to catch the other possible cases as well

jkczyz · 2025-09-19T19:26:05Z

ConstructedTransaction tracks the result of the InteractiveTxConstruction and is used during an InteractiveTxSigningSession. This PR refactors it such that it holds an actual Transaction rather than the contributed parts, though some metadata in the form of NegotiatedTxInput and NegotiatedTxOutput need to be maintained.

This allows us to re-use the weight calculation provided by Transaction. It also facilitates reconstructing contributed inputs and outputs to produce a SpliceFailed event in #4077.

ldk-reviews-bot · 2025-09-19T19:26:08Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

codecov · 2025-09-19T19:37:15Z

Codecov Report

❌ Patch coverage is 92.20183% with 17 lines in your changes missing coverage. Please review.
✅ Project coverage is 88.71%. Comparing base (3564646) to head (86ed012).
⚠️ Report is 109 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/interactivetxs.rs	93.33%	11 Missing and 3 partials ⚠️
lightning/src/ln/channel.rs	25.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4097 +/- ##
==========================================
+ Coverage 87.81% 88.71% +0.89% 
==========================================
Files 176 177 +1 Lines 131770 133237 +1467 Branches 131770 133237 +1467 ==========================================
+ Hits 115719 118196 +2477 + Misses 13416 12326 -1090 - Partials 2635 2715 +80

Flag	Coverage Δ
fuzzing	`21.74% <0.54%> (+0.14%)`	⬆️
tests	`88.55% <92.20%> (+0.89%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2025-09-19T19:43:19Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


Wondering if we should have InteractiveTxSigningSession store both local and remote signatures until finalize_funding_tx is called. That way, ConstructedTransaction would always contain an unsigned transaction instead of one that may be partially signed.

SGTM, we already store the local ones separately

InteractiveTxConstructor::new converts inputs given in channel.rs to InputOwned. Instead of converting FundingTxInput in channel.rs to the type needed by that constructor, pass in FundingTxInput and have it converted directly to InputOwned there. This saves one conversion and Vec allocation. It also lets us use the satisfaction_weight in an upcoming refactor.

Instead of estimating the input weight, track the satisfaction_weight in InteractiveTxInput. This then can be used compute the transaction weight without storing individual weights in NegotiatedTxInput in an upcoming commit.

The weight is no longer needed once a ConstructedTransaction is created, so it doesn't need to be persisted.

Instead of defining a custom function for calculating a transaction's weight, have ConstructedTransaction hold a Transaction so that its weight method can be re-used. As a result NegotiatedTxInput no longer needs to store the transaction inputs.

jkczyz · 2025-09-19T20:06:11Z

Pushed rustfmt fixes.

ldk-reviews-bot · 2025-09-22T00:00:42Z

🔔 1st Reminder

Hey @wpaulino! This PR has been waiting for your review.
Please take a look when you have a chance. If you're unable to review, please let us know so we can find another reviewer.

wpaulino · 2025-09-22T18:38:59Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


SGTM, we already store the local ones separately

ldk-reviews-bot · 2025-09-22T18:44:16Z

👋 The first review has been submitted!

Do you think this PR is ready for a second reviewer? If so, click here to assign a second reviewer.

Now that NegotiatedTxInput no longer holds a TxIn, rename it accordingly.

Now that ConstructedTransaction holds the actual Transaction, there's no need to keep track of and persist the outputs separately. However, the serial IDs are still needed to later reconstruct which outputs were contributed.

The output values are never used so there's no need to persist them. If needed, they can be re-computed though the shared output's local and remote amounts would be lost.

The local and remote input values are used to determine which node sends tx_signatures first. Instead of persisting these values, compute them only when needed from the input metadata. The spec states that the entire shared input value is to be included for the node sending the corresponding tx_add_input, so it isn't necessary to know the local and remote balances which the metadata does not contain.

InteractiveTxSigningSession currently persists holder witnesses directly, but persists counterparty witnesses as part of its unsigned ConstructedTransaction. This makes the ConstructedTransaction actually partially signed even though it is held in a field named unsigned_tx. Instead, persists the counterparty witnesses alongside the holder witnesses directly in InteractiveTxSigningSession, leaving the transaction it holds unsigned.

TheBlueMatt

Code changes look fine to me. I'm not deep enough in the interactive negotiation world to say whether its all a good idea, but seems fine to me.

wpaulino · 2025-09-23T21:08:16Z

 .as_ref()
 .map(|shared_input| SharedInputSignature {
 holder_signature_first: shared_input.holder_sig_first,
- counterparty_signature: None,


Should probably rename SharedInputSignature since it doesn't contain a signature anymore

wpaulino · 2025-09-23T21:10:06Z

+ if let Some(shared_input_index) = self.shared_input_index {
+ let holder_shared_input_sig =
+ holder_tx_signatures.shared_input_signature.or_else(|| {
+ debug_assert!(false);


I think it'd be better to debug_assert_eq!(shared_input_index.is_some(), holder_tx_signatures.shared_input_signature.is_some()) to catch the other possible cases as well

jkczyz · 2025-09-19T19:26:05Z

ConstructedTransaction tracks the result of the InteractiveTxConstruction and is used during an InteractiveTxSigningSession. This PR refactors it such that it holds an actual Transaction rather than the contributed parts, though some metadata in the form of NegotiatedTxInput and NegotiatedTxOutput need to be maintained.

This allows us to re-use the weight calculation provided by Transaction. It also facilitates reconstructing contributed inputs and outputs to produce a SpliceFailed event in #4077.

ldk-reviews-bot · 2025-09-19T19:26:08Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

codecov · 2025-09-19T19:37:15Z

Codecov Report

❌ Patch coverage is 92.20183% with 17 lines in your changes missing coverage. Please review.
✅ Project coverage is 88.71%. Comparing base (3564646) to head (86ed012).
⚠️ Report is 109 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/interactivetxs.rs	93.33%	11 Missing and 3 partials ⚠️
lightning/src/ln/channel.rs	25.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4097 +/- ##
==========================================
+ Coverage 87.81% 88.71% +0.89% 
==========================================
Files 176 177 +1 Lines 131770 133237 +1467 Branches 131770 133237 +1467 ==========================================
+ Hits 115719 118196 +2477 + Misses 13416 12326 -1090 - Partials 2635 2715 +80

Flag	Coverage Δ
fuzzing	`21.74% <0.54%> (+0.14%)`	⬆️
tests	`88.55% <92.20%> (+0.89%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2025-09-19T19:43:19Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


Wondering if we should have InteractiveTxSigningSession store both local and remote signatures until finalize_funding_tx is called. That way, ConstructedTransaction would always contain an unsigned transaction instead of one that may be partially signed.

SGTM, we already store the local ones separately

InteractiveTxConstructor::new converts inputs given in channel.rs to InputOwned. Instead of converting FundingTxInput in channel.rs to the type needed by that constructor, pass in FundingTxInput and have it converted directly to InputOwned there. This saves one conversion and Vec allocation. It also lets us use the satisfaction_weight in an upcoming refactor.

Instead of estimating the input weight, track the satisfaction_weight in InteractiveTxInput. This then can be used compute the transaction weight without storing individual weights in NegotiatedTxInput in an upcoming commit.

The weight is no longer needed once a ConstructedTransaction is created, so it doesn't need to be persisted.

Instead of defining a custom function for calculating a transaction's weight, have ConstructedTransaction hold a Transaction so that its weight method can be re-used. As a result NegotiatedTxInput no longer needs to store the transaction inputs.

jkczyz · 2025-09-19T20:06:11Z

Pushed rustfmt fixes.

ldk-reviews-bot · 2025-09-22T00:00:42Z

🔔 1st Reminder

Hey @wpaulino! This PR has been waiting for your review.
Please take a look when you have a chance. If you're unable to review, please let us know so we can find another reviewer.

wpaulino · 2025-09-22T18:38:59Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


SGTM, we already store the local ones separately

ldk-reviews-bot · 2025-09-22T18:44:16Z

👋 The first review has been submitted!

Do you think this PR is ready for a second reviewer? If so, click here to assign a second reviewer.

Now that NegotiatedTxInput no longer holds a TxIn, rename it accordingly.

Now that ConstructedTransaction holds the actual Transaction, there's no need to keep track of and persist the outputs separately. However, the serial IDs are still needed to later reconstruct which outputs were contributed.

The output values are never used so there's no need to persist them. If needed, they can be re-computed though the shared output's local and remote amounts would be lost.

The local and remote input values are used to determine which node sends tx_signatures first. Instead of persisting these values, compute them only when needed from the input metadata. The spec states that the entire shared input value is to be included for the node sending the corresponding tx_add_input, so it isn't necessary to know the local and remote balances which the metadata does not contain.

InteractiveTxSigningSession currently persists holder witnesses directly, but persists counterparty witnesses as part of its unsigned ConstructedTransaction. This makes the ConstructedTransaction actually partially signed even though it is held in a field named unsigned_tx. Instead, persists the counterparty witnesses alongside the holder witnesses directly in InteractiveTxSigningSession, leaving the transaction it holds unsigned.

TheBlueMatt

Code changes look fine to me. I'm not deep enough in the interactive negotiation world to say whether its all a good idea, but seems fine to me.

wpaulino · 2025-09-23T21:08:16Z

 .as_ref()
 .map(|shared_input| SharedInputSignature {
 holder_signature_first: shared_input.holder_sig_first,
- counterparty_signature: None,


Should probably rename SharedInputSignature since it doesn't contain a signature anymore

wpaulino · 2025-09-23T21:10:06Z

+ if let Some(shared_input_index) = self.shared_input_index {
+ let holder_shared_input_sig =
+ holder_tx_signatures.shared_input_signature.or_else(|| {
+ debug_assert!(false);


I think it'd be better to debug_assert_eq!(shared_input_index.is_some(), holder_tx_signatures.shared_input_signature.is_some()) to catch the other possible cases as well

jkczyz · 2025-09-19T19:26:05Z

ConstructedTransaction tracks the result of the InteractiveTxConstruction and is used during an InteractiveTxSigningSession. This PR refactors it such that it holds an actual Transaction rather than the contributed parts, though some metadata in the form of NegotiatedTxInput and NegotiatedTxOutput need to be maintained.

This allows us to re-use the weight calculation provided by Transaction. It also facilitates reconstructing contributed inputs and outputs to produce a SpliceFailed event in #4077.

ldk-reviews-bot · 2025-09-19T19:26:08Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

codecov · 2025-09-19T19:37:15Z

Codecov Report

❌ Patch coverage is 92.20183% with 17 lines in your changes missing coverage. Please review.
✅ Project coverage is 88.71%. Comparing base (3564646) to head (86ed012).
⚠️ Report is 109 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/interactivetxs.rs	93.33%	11 Missing and 3 partials ⚠️
lightning/src/ln/channel.rs	25.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4097 +/- ##
==========================================
+ Coverage 87.81% 88.71% +0.89% 
==========================================
Files 176 177 +1 Lines 131770 133237 +1467 Branches 131770 133237 +1467 ==========================================
+ Hits 115719 118196 +2477 + Misses 13416 12326 -1090 - Partials 2635 2715 +80

Flag	Coverage Δ
fuzzing	`21.74% <0.54%> (+0.14%)`	⬆️
tests	`88.55% <92.20%> (+0.89%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2025-09-19T19:43:19Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


Wondering if we should have InteractiveTxSigningSession store both local and remote signatures until finalize_funding_tx is called. That way, ConstructedTransaction would always contain an unsigned transaction instead of one that may be partially signed.

SGTM, we already store the local ones separately

InteractiveTxConstructor::new converts inputs given in channel.rs to InputOwned. Instead of converting FundingTxInput in channel.rs to the type needed by that constructor, pass in FundingTxInput and have it converted directly to InputOwned there. This saves one conversion and Vec allocation. It also lets us use the satisfaction_weight in an upcoming refactor.

Instead of estimating the input weight, track the satisfaction_weight in InteractiveTxInput. This then can be used compute the transaction weight without storing individual weights in NegotiatedTxInput in an upcoming commit.

The weight is no longer needed once a ConstructedTransaction is created, so it doesn't need to be persisted.

Instead of defining a custom function for calculating a transaction's weight, have ConstructedTransaction hold a Transaction so that its weight method can be re-used. As a result NegotiatedTxInput no longer needs to store the transaction inputs.

jkczyz · 2025-09-19T20:06:11Z

Pushed rustfmt fixes.

ldk-reviews-bot · 2025-09-22T00:00:42Z

🔔 1st Reminder

Hey @wpaulino! This PR has been waiting for your review.
Please take a look when you have a chance. If you're unable to review, please let us know so we can find another reviewer.

wpaulino · 2025-09-22T18:38:59Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


SGTM, we already store the local ones separately

ldk-reviews-bot · 2025-09-22T18:44:16Z

👋 The first review has been submitted!

Do you think this PR is ready for a second reviewer? If so, click here to assign a second reviewer.

Now that NegotiatedTxInput no longer holds a TxIn, rename it accordingly.

Now that ConstructedTransaction holds the actual Transaction, there's no need to keep track of and persist the outputs separately. However, the serial IDs are still needed to later reconstruct which outputs were contributed.

The output values are never used so there's no need to persist them. If needed, they can be re-computed though the shared output's local and remote amounts would be lost.

The local and remote input values are used to determine which node sends tx_signatures first. Instead of persisting these values, compute them only when needed from the input metadata. The spec states that the entire shared input value is to be included for the node sending the corresponding tx_add_input, so it isn't necessary to know the local and remote balances which the metadata does not contain.

InteractiveTxSigningSession currently persists holder witnesses directly, but persists counterparty witnesses as part of its unsigned ConstructedTransaction. This makes the ConstructedTransaction actually partially signed even though it is held in a field named unsigned_tx. Instead, persists the counterparty witnesses alongside the holder witnesses directly in InteractiveTxSigningSession, leaving the transaction it holds unsigned.

TheBlueMatt

Code changes look fine to me. I'm not deep enough in the interactive negotiation world to say whether its all a good idea, but seems fine to me.

wpaulino · 2025-09-23T21:08:16Z

 .as_ref()
 .map(|shared_input| SharedInputSignature {
 holder_signature_first: shared_input.holder_sig_first,
- counterparty_signature: None,


Should probably rename SharedInputSignature since it doesn't contain a signature anymore

wpaulino · 2025-09-23T21:10:06Z

+ if let Some(shared_input_index) = self.shared_input_index {
+ let holder_shared_input_sig =
+ holder_tx_signatures.shared_input_signature.or_else(|| {
+ debug_assert!(false);


I think it'd be better to debug_assert_eq!(shared_input_index.is_some(), holder_tx_signatures.shared_input_signature.is_some()) to catch the other possible cases as well

jkczyz · 2025-09-19T19:26:05Z

ConstructedTransaction tracks the result of the InteractiveTxConstruction and is used during an InteractiveTxSigningSession. This PR refactors it such that it holds an actual Transaction rather than the contributed parts, though some metadata in the form of NegotiatedTxInput and NegotiatedTxOutput need to be maintained.

This allows us to re-use the weight calculation provided by Transaction. It also facilitates reconstructing contributed inputs and outputs to produce a SpliceFailed event in #4077.

ldk-reviews-bot · 2025-09-19T19:26:08Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

codecov · 2025-09-19T19:37:15Z

Codecov Report

❌ Patch coverage is 92.20183% with 17 lines in your changes missing coverage. Please review.
✅ Project coverage is 88.71%. Comparing base (3564646) to head (86ed012).
⚠️ Report is 109 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/interactivetxs.rs	93.33%	11 Missing and 3 partials ⚠️
lightning/src/ln/channel.rs	25.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4097 +/- ##
==========================================
+ Coverage 87.81% 88.71% +0.89% 
==========================================
Files 176 177 +1 Lines 131770 133237 +1467 Branches 131770 133237 +1467 ==========================================
+ Hits 115719 118196 +2477 + Misses 13416 12326 -1090 - Partials 2635 2715 +80

Flag	Coverage Δ
fuzzing	`21.74% <0.54%> (+0.14%)`	⬆️
tests	`88.55% <92.20%> (+0.89%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2025-09-19T19:43:19Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


Wondering if we should have InteractiveTxSigningSession store both local and remote signatures until finalize_funding_tx is called. That way, ConstructedTransaction would always contain an unsigned transaction instead of one that may be partially signed.

SGTM, we already store the local ones separately

InteractiveTxConstructor::new converts inputs given in channel.rs to InputOwned. Instead of converting FundingTxInput in channel.rs to the type needed by that constructor, pass in FundingTxInput and have it converted directly to InputOwned there. This saves one conversion and Vec allocation. It also lets us use the satisfaction_weight in an upcoming refactor.

Instead of estimating the input weight, track the satisfaction_weight in InteractiveTxInput. This then can be used compute the transaction weight without storing individual weights in NegotiatedTxInput in an upcoming commit.

The weight is no longer needed once a ConstructedTransaction is created, so it doesn't need to be persisted.

Instead of defining a custom function for calculating a transaction's weight, have ConstructedTransaction hold a Transaction so that its weight method can be re-used. As a result NegotiatedTxInput no longer needs to store the transaction inputs.

jkczyz · 2025-09-19T20:06:11Z

Pushed rustfmt fixes.

ldk-reviews-bot · 2025-09-22T00:00:42Z

🔔 1st Reminder

Hey @wpaulino! This PR has been waiting for your review.
Please take a look when you have a chance. If you're unable to review, please let us know so we can find another reviewer.

wpaulino · 2025-09-22T18:38:59Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


SGTM, we already store the local ones separately

ldk-reviews-bot · 2025-09-22T18:44:16Z

👋 The first review has been submitted!

Do you think this PR is ready for a second reviewer? If so, click here to assign a second reviewer.

Now that NegotiatedTxInput no longer holds a TxIn, rename it accordingly.

Now that ConstructedTransaction holds the actual Transaction, there's no need to keep track of and persist the outputs separately. However, the serial IDs are still needed to later reconstruct which outputs were contributed.

The output values are never used so there's no need to persist them. If needed, they can be re-computed though the shared output's local and remote amounts would be lost.

The local and remote input values are used to determine which node sends tx_signatures first. Instead of persisting these values, compute them only when needed from the input metadata. The spec states that the entire shared input value is to be included for the node sending the corresponding tx_add_input, so it isn't necessary to know the local and remote balances which the metadata does not contain.

InteractiveTxSigningSession currently persists holder witnesses directly, but persists counterparty witnesses as part of its unsigned ConstructedTransaction. This makes the ConstructedTransaction actually partially signed even though it is held in a field named unsigned_tx. Instead, persists the counterparty witnesses alongside the holder witnesses directly in InteractiveTxSigningSession, leaving the transaction it holds unsigned.

TheBlueMatt

Code changes look fine to me. I'm not deep enough in the interactive negotiation world to say whether its all a good idea, but seems fine to me.

wpaulino · 2025-09-23T21:08:16Z

 .as_ref()
 .map(|shared_input| SharedInputSignature {
 holder_signature_first: shared_input.holder_sig_first,
- counterparty_signature: None,


Should probably rename SharedInputSignature since it doesn't contain a signature anymore

wpaulino · 2025-09-23T21:10:06Z

+ if let Some(shared_input_index) = self.shared_input_index {
+ let holder_shared_input_sig =
+ holder_tx_signatures.shared_input_signature.or_else(|| {
+ debug_assert!(false);


I think it'd be better to debug_assert_eq!(shared_input_index.is_some(), holder_tx_signatures.shared_input_signature.is_some()) to catch the other possible cases as well

jkczyz · 2025-09-19T19:26:05Z

ConstructedTransaction tracks the result of the InteractiveTxConstruction and is used during an InteractiveTxSigningSession. This PR refactors it such that it holds an actual Transaction rather than the contributed parts, though some metadata in the form of NegotiatedTxInput and NegotiatedTxOutput need to be maintained.

This allows us to re-use the weight calculation provided by Transaction. It also facilitates reconstructing contributed inputs and outputs to produce a SpliceFailed event in #4077.

ldk-reviews-bot · 2025-09-19T19:26:08Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

codecov · 2025-09-19T19:37:15Z

Codecov Report

❌ Patch coverage is 92.20183% with 17 lines in your changes missing coverage. Please review.
✅ Project coverage is 88.71%. Comparing base (3564646) to head (86ed012).
⚠️ Report is 109 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/interactivetxs.rs	93.33%	11 Missing and 3 partials ⚠️
lightning/src/ln/channel.rs	25.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4097 +/- ##
==========================================
+ Coverage 87.81% 88.71% +0.89% 
==========================================
Files 176 177 +1 Lines 131770 133237 +1467 Branches 131770 133237 +1467 ==========================================
+ Hits 115719 118196 +2477 + Misses 13416 12326 -1090 - Partials 2635 2715 +80

Flag	Coverage Δ
fuzzing	`21.74% <0.54%> (+0.14%)`	⬆️
tests	`88.55% <92.20%> (+0.89%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2025-09-19T19:43:19Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


Wondering if we should have InteractiveTxSigningSession store both local and remote signatures until finalize_funding_tx is called. That way, ConstructedTransaction would always contain an unsigned transaction instead of one that may be partially signed.

SGTM, we already store the local ones separately

InteractiveTxConstructor::new converts inputs given in channel.rs to InputOwned. Instead of converting FundingTxInput in channel.rs to the type needed by that constructor, pass in FundingTxInput and have it converted directly to InputOwned there. This saves one conversion and Vec allocation. It also lets us use the satisfaction_weight in an upcoming refactor.

Instead of estimating the input weight, track the satisfaction_weight in InteractiveTxInput. This then can be used compute the transaction weight without storing individual weights in NegotiatedTxInput in an upcoming commit.

The weight is no longer needed once a ConstructedTransaction is created, so it doesn't need to be persisted.

Instead of defining a custom function for calculating a transaction's weight, have ConstructedTransaction hold a Transaction so that its weight method can be re-used. As a result NegotiatedTxInput no longer needs to store the transaction inputs.

jkczyz · 2025-09-19T20:06:11Z

Pushed rustfmt fixes.

ldk-reviews-bot · 2025-09-22T00:00:42Z

🔔 1st Reminder

Hey @wpaulino! This PR has been waiting for your review.
Please take a look when you have a chance. If you're unable to review, please let us know so we can find another reviewer.

wpaulino · 2025-09-22T18:38:59Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


SGTM, we already store the local ones separately

ldk-reviews-bot · 2025-09-22T18:44:16Z

👋 The first review has been submitted!

Do you think this PR is ready for a second reviewer? If so, click here to assign a second reviewer.

Now that NegotiatedTxInput no longer holds a TxIn, rename it accordingly.

Now that ConstructedTransaction holds the actual Transaction, there's no need to keep track of and persist the outputs separately. However, the serial IDs are still needed to later reconstruct which outputs were contributed.

The output values are never used so there's no need to persist them. If needed, they can be re-computed though the shared output's local and remote amounts would be lost.

The local and remote input values are used to determine which node sends tx_signatures first. Instead of persisting these values, compute them only when needed from the input metadata. The spec states that the entire shared input value is to be included for the node sending the corresponding tx_add_input, so it isn't necessary to know the local and remote balances which the metadata does not contain.

InteractiveTxSigningSession currently persists holder witnesses directly, but persists counterparty witnesses as part of its unsigned ConstructedTransaction. This makes the ConstructedTransaction actually partially signed even though it is held in a field named unsigned_tx. Instead, persists the counterparty witnesses alongside the holder witnesses directly in InteractiveTxSigningSession, leaving the transaction it holds unsigned.

TheBlueMatt

Code changes look fine to me. I'm not deep enough in the interactive negotiation world to say whether its all a good idea, but seems fine to me.

wpaulino · 2025-09-23T21:08:16Z

 .as_ref()
 .map(|shared_input| SharedInputSignature {
 holder_signature_first: shared_input.holder_sig_first,
- counterparty_signature: None,


Should probably rename SharedInputSignature since it doesn't contain a signature anymore

wpaulino · 2025-09-23T21:10:06Z

+ if let Some(shared_input_index) = self.shared_input_index {
+ let holder_shared_input_sig =
+ holder_tx_signatures.shared_input_signature.or_else(|| {
+ debug_assert!(false);


I think it'd be better to debug_assert_eq!(shared_input_index.is_some(), holder_tx_signatures.shared_input_signature.is_some()) to catch the other possible cases as well

jkczyz · 2025-09-19T19:26:05Z

ConstructedTransaction tracks the result of the InteractiveTxConstruction and is used during an InteractiveTxSigningSession. This PR refactors it such that it holds an actual Transaction rather than the contributed parts, though some metadata in the form of NegotiatedTxInput and NegotiatedTxOutput need to be maintained.

This allows us to re-use the weight calculation provided by Transaction. It also facilitates reconstructing contributed inputs and outputs to produce a SpliceFailed event in #4077.

ldk-reviews-bot · 2025-09-19T19:26:08Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

codecov · 2025-09-19T19:37:15Z

Codecov Report

❌ Patch coverage is 92.20183% with 17 lines in your changes missing coverage. Please review.
✅ Project coverage is 88.71%. Comparing base (3564646) to head (86ed012).
⚠️ Report is 109 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/interactivetxs.rs	93.33%	11 Missing and 3 partials ⚠️
lightning/src/ln/channel.rs	25.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4097 +/- ##
==========================================
+ Coverage 87.81% 88.71% +0.89% 
==========================================
Files 176 177 +1 Lines 131770 133237 +1467 Branches 131770 133237 +1467 ==========================================
+ Hits 115719 118196 +2477 + Misses 13416 12326 -1090 - Partials 2635 2715 +80

Flag	Coverage Δ
fuzzing	`21.74% <0.54%> (+0.14%)`	⬆️
tests	`88.55% <92.20%> (+0.89%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2025-09-19T19:43:19Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


Wondering if we should have InteractiveTxSigningSession store both local and remote signatures until finalize_funding_tx is called. That way, ConstructedTransaction would always contain an unsigned transaction instead of one that may be partially signed.

SGTM, we already store the local ones separately

InteractiveTxConstructor::new converts inputs given in channel.rs to InputOwned. Instead of converting FundingTxInput in channel.rs to the type needed by that constructor, pass in FundingTxInput and have it converted directly to InputOwned there. This saves one conversion and Vec allocation. It also lets us use the satisfaction_weight in an upcoming refactor.

Instead of estimating the input weight, track the satisfaction_weight in InteractiveTxInput. This then can be used compute the transaction weight without storing individual weights in NegotiatedTxInput in an upcoming commit.

The weight is no longer needed once a ConstructedTransaction is created, so it doesn't need to be persisted.

Instead of defining a custom function for calculating a transaction's weight, have ConstructedTransaction hold a Transaction so that its weight method can be re-used. As a result NegotiatedTxInput no longer needs to store the transaction inputs.

jkczyz · 2025-09-19T20:06:11Z

Pushed rustfmt fixes.

ldk-reviews-bot · 2025-09-22T00:00:42Z

🔔 1st Reminder

Hey @wpaulino! This PR has been waiting for your review.
Please take a look when you have a chance. If you're unable to review, please let us know so we can find another reviewer.

wpaulino · 2025-09-22T18:38:59Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


SGTM, we already store the local ones separately

ldk-reviews-bot · 2025-09-22T18:44:16Z

👋 The first review has been submitted!

Do you think this PR is ready for a second reviewer? If so, click here to assign a second reviewer.

Now that NegotiatedTxInput no longer holds a TxIn, rename it accordingly.

Now that ConstructedTransaction holds the actual Transaction, there's no need to keep track of and persist the outputs separately. However, the serial IDs are still needed to later reconstruct which outputs were contributed.

The output values are never used so there's no need to persist them. If needed, they can be re-computed though the shared output's local and remote amounts would be lost.

The local and remote input values are used to determine which node sends tx_signatures first. Instead of persisting these values, compute them only when needed from the input metadata. The spec states that the entire shared input value is to be included for the node sending the corresponding tx_add_input, so it isn't necessary to know the local and remote balances which the metadata does not contain.

InteractiveTxSigningSession currently persists holder witnesses directly, but persists counterparty witnesses as part of its unsigned ConstructedTransaction. This makes the ConstructedTransaction actually partially signed even though it is held in a field named unsigned_tx. Instead, persists the counterparty witnesses alongside the holder witnesses directly in InteractiveTxSigningSession, leaving the transaction it holds unsigned.

TheBlueMatt

Code changes look fine to me. I'm not deep enough in the interactive negotiation world to say whether its all a good idea, but seems fine to me.

wpaulino · 2025-09-23T21:08:16Z

 .as_ref()
 .map(|shared_input| SharedInputSignature {
 holder_signature_first: shared_input.holder_sig_first,
- counterparty_signature: None,


Should probably rename SharedInputSignature since it doesn't contain a signature anymore

wpaulino · 2025-09-23T21:10:06Z

+ if let Some(shared_input_index) = self.shared_input_index {
+ let holder_shared_input_sig =
+ holder_tx_signatures.shared_input_signature.or_else(|| {
+ debug_assert!(false);


I think it'd be better to debug_assert_eq!(shared_input_index.is_some(), holder_tx_signatures.shared_input_signature.is_some()) to catch the other possible cases as well

jkczyz · 2025-09-19T19:26:05Z

ConstructedTransaction tracks the result of the InteractiveTxConstruction and is used during an InteractiveTxSigningSession. This PR refactors it such that it holds an actual Transaction rather than the contributed parts, though some metadata in the form of NegotiatedTxInput and NegotiatedTxOutput need to be maintained.

This allows us to re-use the weight calculation provided by Transaction. It also facilitates reconstructing contributed inputs and outputs to produce a SpliceFailed event in #4077.

ldk-reviews-bot · 2025-09-19T19:26:08Z

👋 Thanks for assigning @wpaulino as a reviewer!
I'll wait for their review and will help manage the review process.
Once they submit their review, I'll check if a second reviewer would be helpful.

codecov · 2025-09-19T19:37:15Z

Codecov Report

❌ Patch coverage is 92.20183% with 17 lines in your changes missing coverage. Please review.
✅ Project coverage is 88.71%. Comparing base (3564646) to head (86ed012).
⚠️ Report is 109 commits behind head on main.

Files with missing lines	Patch %	Lines
lightning/src/ln/interactivetxs.rs	93.33%	11 Missing and 3 partials ⚠️
lightning/src/ln/channel.rs	25.00%	3 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #4097 +/- ##
==========================================
+ Coverage 87.81% 88.71% +0.89% 
==========================================
Files 176 177 +1 Lines 131770 133237 +1467 Branches 131770 133237 +1467 ==========================================
+ Hits 115719 118196 +2477 + Misses 13416 12326 -1090 - Partials 2635 2715 +80

Flag	Coverage Δ
fuzzing	`21.74% <0.54%> (+0.14%)`	⬆️
tests	`88.55% <92.20%> (+0.89%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jkczyz · 2025-09-19T19:43:19Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


Wondering if we should have InteractiveTxSigningSession store both local and remote signatures until finalize_funding_tx is called. That way, ConstructedTransaction would always contain an unsigned transaction instead of one that may be partially signed.

SGTM, we already store the local ones separately

InteractiveTxConstructor::new converts inputs given in channel.rs to InputOwned. Instead of converting FundingTxInput in channel.rs to the type needed by that constructor, pass in FundingTxInput and have it converted directly to InputOwned there. This saves one conversion and Vec allocation. It also lets us use the satisfaction_weight in an upcoming refactor.

Instead of estimating the input weight, track the satisfaction_weight in InteractiveTxInput. This then can be used compute the transaction weight without storing individual weights in NegotiatedTxInput in an upcoming commit.

The weight is no longer needed once a ConstructedTransaction is created, so it doesn't need to be persisted.

Instead of defining a custom function for calculating a transaction's weight, have ConstructedTransaction hold a Transaction so that its weight method can be re-used. As a result NegotiatedTxInput no longer needs to store the transaction inputs.

jkczyz · 2025-09-19T20:06:11Z

Pushed rustfmt fixes.

ldk-reviews-bot · 2025-09-22T00:00:42Z

🔔 1st Reminder

Hey @wpaulino! This PR has been waiting for your review.
Please take a look when you have a chance. If you're unable to review, please let us know so we can find another reviewer.

wpaulino · 2025-09-22T18:38:59Z

 	fn add_local_witnesses(&mut self, witnesses: Vec<Witness>) {
- self.inputs
+ self.tx
+ .input
 .iter_mut()
+ .zip(self.inputs.iter())
 .enumerate()
- .filter(|(_, input)| input.is_local(self.holder_is_initiator))
+ .filter(|(_, (_, input))| input.is_local(self.holder_is_initiator))
 .filter(|(index, _)| {
 self.shared_input_index
 .map(|shared_index| *index != shared_index as usize)
 .unwrap_or(true)
 })
- .map(|(_, input)| &mut input.txin)
+ .map(|(_, (txin, _))| txin)
 .zip(witnesses)
 .for_each(|(input, witness)| input.witness = witness);
 	}


SGTM, we already store the local ones separately

ldk-reviews-bot · 2025-09-22T18:44:16Z

👋 The first review has been submitted!

Do you think this PR is ready for a second reviewer? If so, click here to assign a second reviewer.

Now that NegotiatedTxInput no longer holds a TxIn, rename it accordingly.

Now that ConstructedTransaction holds the actual Transaction, there's no need to keep track of and persist the outputs separately. However, the serial IDs are still needed to later reconstruct which outputs were contributed.

The output values are never used so there's no need to persist them. If needed, they can be re-computed though the shared output's local and remote amounts would be lost.

The local and remote input values are used to determine which node sends tx_signatures first. Instead of persisting these values, compute them only when needed from the input metadata. The spec states that the entire shared input value is to be included for the node sending the corresponding tx_add_input, so it isn't necessary to know the local and remote balances which the metadata does not contain.

InteractiveTxSigningSession currently persists holder witnesses directly, but persists counterparty witnesses as part of its unsigned ConstructedTransaction. This makes the ConstructedTransaction actually partially signed even though it is held in a field named unsigned_tx. Instead, persists the counterparty witnesses alongside the holder witnesses directly in InteractiveTxSigningSession, leaving the transaction it holds unsigned.

TheBlueMatt

Code changes look fine to me. I'm not deep enough in the interactive negotiation world to say whether its all a good idea, but seems fine to me.

wpaulino · 2025-09-23T21:08:16Z

 .as_ref()
 .map(|shared_input| SharedInputSignature {
 holder_signature_first: shared_input.holder_sig_first,
- counterparty_signature: None,


Should probably rename SharedInputSignature since it doesn't contain a signature anymore

wpaulino · 2025-09-23T21:10:06Z

+ if let Some(shared_input_index) = self.shared_input_index {
+ let holder_shared_input_sig =
+ holder_tx_signatures.shared_input_signature.or_else(|| {
+ debug_assert!(false);


I think it'd be better to debug_assert_eq!(shared_input_index.is_some(), holder_tx_signatures.shared_input_signature.is_some()) to catch the other possible cases as well

jkczyz requested a review from wpaulino September 19, 2025 19:26

jkczyz commented Sep 19, 2025
View reviewed changes

jkczyz added 4 commits September 19, 2025 14:52

Track satisfaction_weight in InteractiveTxInput
1e510cc
Instead of estimating the input weight, track the satisfaction_weight in InteractiveTxInput. This then can be used compute the transaction weight without storing individual weights in NegotiatedTxInput in an upcoming commit.

Remove NegotiatedTxInput::weight
0a1d445
The weight is no longer needed once a ConstructedTransaction is created, so it doesn't need to be persisted.

jkczyz force-pushed the 2025-09-interactive-tx-refactor branch from f96f304 to 9eeaf52Compare September 19, 2025 20:05

jkczyz mentioned this pull request Sep 19, 2025
Dual-funded channels and Splicing Project Tracking #1621
Closed
7 tasks

wpaulino reviewed Sep 22, 2025
View reviewed changes

jkczyz added 2 commits September 22, 2025 16:07

Rename NegotiatedTxInput to TxInMetadata
d01df5a
Now that NegotiatedTxInput no longer holds a TxIn, rename it accordingly.

Remove duplicate TxOut persistence
0cab3c3
Now that ConstructedTransaction holds the actual Transaction, there's no need to keep track of and persist the outputs separately. However, the serial IDs are still needed to later reconstruct which outputs were contributed.

jkczyz force-pushed the 2025-09-interactive-tx-refactor branch from 9eeaf52 to 7958dd9Compare September 22, 2025 23:37

jkczyz mentioned this pull request Sep 23, 2025
Emit SplicePending and SpliceFailed events #4077
Merged

jkczyz added this to the 0.2 milestone Sep 23, 2025

jkczyz self-assigned this Sep 23, 2025

jkczyz requested a review from TheBlueMatt September 23, 2025 13:35

jkczyz added 4 commits September 23, 2025 11:36

Remove unused fields from ConstructedTransaction
a8d7646
The output values are never used so there's no need to persist them. If needed, they can be re-computed though the shared output's local and remote amounts would be lost.

Reduce visibility of ConstructedTransaction methods
86ed012

jkczyz force-pushed the 2025-09-interactive-tx-refactor branch from 7958dd9 to 86ed012Compare September 23, 2025 16:53

jkczyz requested a review from wpaulino September 23, 2025 16:53

TheBlueMatt approved these changes Sep 23, 2025
View reviewed changes

wpaulino approved these changes Sep 23, 2025
View reviewed changes

wpaulino merged commit 5a84daf into lightningdevkit:mainSep 24, 2025
40 of 45 checks passed