feat(chain)!: Add ability to modify canonicalization algorithm

Introduce `CanonicalizationParams` which is passed in to
`CanonicalIter::new`.

`CanonicalizationParams::assume_canonical` is the only field right now.
This contains a list of txids that we assume to be canonical,
superceding any other canonicalization rules.

Author: evanlinjin

crates/bitcoind_rpc/examples/filter_iter.rs

@@ -92,6 +92,7 @@ fn main() -> anyhow::Result<()> {
         .filter_chain_unspents(
             &chain,
             chain.tip().block_id(),
+            Default::default(),
             graph.index.outpoints().clone(),
         )
         .collect();

crates/bitcoind_rpc/tests/test_emitter.rs

@@ -5,7 +5,7 @@ use bdk_chain::{
     bitcoin::{Address, Amount, Txid},
     local_chain::{CheckPoint, LocalChain},
     spk_txout::SpkTxOutIndex,
-    Balance, BlockId, IndexedTxGraph, Merge,
+    Balance, BlockId, CanonicalizationParams, IndexedTxGraph, Merge,
 };
 use bdk_testenv::{anyhow, TestEnv};
 use bitcoin::{hashes::Hash, Block, OutPoint, ScriptBuf, WScriptHash};
@@ -306,9 +306,13 @@ fn get_balance(
 ) -> anyhow::Result<Balance> {
     let chain_tip = recv_chain.tip().block_id();
     let outpoints = recv_graph.index.outpoints().clone();
-    let balance = recv_graph
-        .graph()
-        .balance(recv_chain, chain_tip, outpoints, |_, _| true);
+    let balance = recv_graph.graph().balance(
+        recv_chain,
+        chain_tip,
+        CanonicalizationParams::default(),
+        outpoints,
+        |_, _| true,
+    );
     Ok(balance)
 }
 

crates/chain/benches/canonicalization.rs

@@ -1,3 +1,4 @@
+use bdk_chain::CanonicalizationParams;
 use bdk_chain::{keychain_txout::KeychainTxOutIndex, local_chain::LocalChain, IndexedTxGraph};
 use bdk_core::{BlockId, CheckPoint};
 use bdk_core::{ConfirmationBlockTime, TxUpdate};
@@ -90,16 +91,19 @@ fn setup<F: Fn(&mut KeychainTxGraph, &LocalChain)>(f: F) -> (KeychainTxGraph, Lo
 }
 
 fn run_list_canonical_txs(tx_graph: &KeychainTxGraph, chain: &LocalChain, exp_txs: usize) {
-    let txs = tx_graph
-        .graph()
-        .list_canonical_txs(chain, chain.tip().block_id());
+    let txs = tx_graph.graph().list_canonical_txs(
+        chain,
+        chain.tip().block_id(),
+        CanonicalizationParams::default(),
+    );
     assert_eq!(txs.count(), exp_txs);
 }
 
 fn run_filter_chain_txouts(tx_graph: &KeychainTxGraph, chain: &LocalChain, exp_txos: usize) {
     let utxos = tx_graph.graph().filter_chain_txouts(
         chain,
         chain.tip().block_id(),
+        CanonicalizationParams::default(),
         tx_graph.index.outpoints().clone(),
     );
     assert_eq!(utxos.count(), exp_txos);
@@ -109,6 +113,7 @@ fn run_filter_chain_unspents(tx_graph: &KeychainTxGraph, chain: &LocalChain, exp
     let utxos = tx_graph.graph().filter_chain_unspents(
         chain,
         chain.tip().block_id(),
+        CanonicalizationParams::default(),
         tx_graph.index.outpoints().clone(),
     );
     assert_eq!(utxos.count(), exp_utxos);

crates/chain/src/canonical_iter.rs

@@ -11,12 +11,23 @@ use bitcoin::{Transaction, Txid};
 type CanonicalMap<A> = HashMap<Txid, (Arc<Transaction>, CanonicalReason<A>)>;
 type NotCanonicalSet = HashSet<Txid>;
 
+/// Modifies the canonicalization algorithm.
+#[derive(Debug, Default, Clone)]
+pub struct CanonicalizationParams {
+    /// Transactions that will supercede all other transactions.
+    ///
+    /// In case of conflicting transactions within `assume_canonical`, transactions that appear
+    /// later in the list (have higher index) have precedence.
+    pub assume_canonical: Vec<Txid>,
+}
+
 /// Iterates over canonical txs.
 pub struct CanonicalIter<'g, A, C> {
     tx_graph: &'g TxGraph<A>,
     chain: &'g C,
     chain_tip: BlockId,
 
+    unprocessed_assumed_txs: Box<dyn Iterator<Item = (Txid, Arc<Transaction>)> + 'g>,
     unprocessed_anchored_txs:
         Box<dyn Iterator<Item = (Txid, Arc<Transaction>, &'g BTreeSet<A>)> + 'g>,
     unprocessed_seen_txs: Box<dyn Iterator<Item = (Txid, Arc<Transaction>, u64)> + 'g>,
@@ -30,8 +41,20 @@ pub struct CanonicalIter<'g, A, C> {
 
 impl<'g, A: Anchor, C: ChainOracle> CanonicalIter<'g, A, C> {
     /// Constructs [`CanonicalIter`].
-    pub fn new(tx_graph: &'g TxGraph<A>, chain: &'g C, chain_tip: BlockId) -> Self {
+    pub fn new(
+        tx_graph: &'g TxGraph<A>,
+        chain: &'g C,
+        chain_tip: BlockId,
+        params: CanonicalizationParams,
+    ) -> Self {
         let anchors = tx_graph.all_anchors();
+        let unprocessed_assumed_txs = Box::new(
+            params
+                .assume_canonical
+                .into_iter()
+                .rev()
+                .filter_map(|txid| Some((txid, tx_graph.get_tx(txid)?))),
+        );
         let unprocessed_anchored_txs = Box::new(
             tx_graph
                 .txids_by_descending_anchor_height()
@@ -46,6 +69,7 @@ impl<'g, A: Anchor, C: ChainOracle> CanonicalIter<'g, A, C> {
             tx_graph,
             chain,
             chain_tip,
+            unprocessed_assumed_txs,
             unprocessed_anchored_txs,
             unprocessed_seen_txs,
             unprocessed_leftover_txs: VecDeque::new(),
@@ -190,6 +214,12 @@ impl<A: Anchor, C: ChainOracle> Iterator for CanonicalIter<'_, A, C> {
                 return Some(Ok((txid, tx, reason)));
             }
 
+            if let Some((txid, tx)) = self.unprocessed_assumed_txs.next() {
+                if !self.is_canonicalized(txid) {
+                    self.mark_canonical(txid, tx, CanonicalReason::assumed());
+                }
+            }
+
             if let Some((txid, tx, anchors)) = self.unprocessed_anchored_txs.next() {
                 if !self.is_canonicalized(txid) {
                     if let Err(err) = self.scan_anchors(txid, tx, anchors) {
@@ -232,6 +262,12 @@ pub enum ObservedIn {
 /// The reason why a transaction is canonical.
 #[derive(Debug, Clone, PartialEq, Eq)]
 pub enum CanonicalReason<A> {
+    /// This transaction is explicitly assumed to be canonical by the caller, superceding all other
+    /// canonicalization rules.
+    Assumed {
+        /// Whether it is a descendant that is assumed to be canonical.
+        descendant: Option<Txid>,
+    },
     /// This transaction is anchored in the best chain by `A`, and therefore canonical.
     Anchor {
         /// The anchor that anchored the transaction in the chain.
@@ -250,6 +286,12 @@ pub enum CanonicalReason<A> {
 }
 
 impl<A: Clone> CanonicalReason<A> {
+    /// Constructs a [`CanonicalReason`] for a transaction that is assumed to supercede all other
+    /// transactions.
+    pub fn assumed() -> Self {
+        Self::Assumed { descendant: None }
+    }
+
     /// Constructs a [`CanonicalReason`] from an `anchor`.
     pub fn from_anchor(anchor: A) -> Self {
         Self::Anchor {
@@ -272,6 +314,9 @@ impl<A: Clone> CanonicalReason<A> {
     /// descendant, but is transitively relevant.
     pub fn to_transitive(&self, descendant: Txid) -> Self {
         match self {
+            CanonicalReason::Assumed { .. } => Self::Assumed {
+                descendant: Some(descendant),
+            },
             CanonicalReason::Anchor { anchor, .. } => Self::Anchor {
                 anchor: anchor.clone(),
                 descendant: Some(descendant),
@@ -287,6 +332,7 @@ impl<A: Clone> CanonicalReason<A> {
     /// descendant.
     pub fn descendant(&self) -> &Option<Txid> {
         match self {
+            CanonicalReason::Assumed { descendant, .. } => descendant,
             CanonicalReason::Anchor { descendant, .. } => descendant,
             CanonicalReason::ObservedIn { descendant, .. } => descendant,
         }