chore: expose PruneConfig in CLI args (#10639)

Co-authored-by: garwah <garwah@garwah>
Co-authored-by: Emilia Hane <elsaemiliaevahane@gmail.com>

Cargo.lock

@@ -7673,6 +7673,7 @@ dependencies = [
  "shellexpand",
  "strum",
  "tempfile",
+ "thiserror",
  "tokio",
  "toml",
  "tracing",

book/cli/reth/node.md

@@ -580,7 +580,60 @@ Dev testnet:
 
 Pruning:
       --full
-          Run full node. Only the most recent [`MINIMUM_PRUNING_DISTANCE`] block states are stored. This flag takes priority over pruning configuration in reth.toml
+          Run full node. Only the most recent [`MINIMUM_PRUNING_DISTANCE`] block states are stored
+
+      --block-interval <BLOCK_INTERVAL>
+          Minimum pruning interval measured in blocks
+
+          [default: 0]
+
+      --prune.senderrecovery.full
+          Prunes all sender recovery data
+
+      --prune.senderrecovery.distance <BLOCKS>
+          Prune sender recovery data before the `head-N` block number. In other words, keep last N + 1 blocks
+
+      --prune.senderrecovery.before <BLOCK_NUMBER>
+          Prune sender recovery data before the specified block number. The specified block number is not pruned
+
+      --prune.transactionlookup.full
+          Prunes all transaction lookup data
+
+      --prune.transactionlookup.distance <BLOCKS>
+          Prune transaction lookup data before the `head-N` block number. In other words, keep last N + 1 blocks
+
+      --prune.transactionlookup.before <BLOCK_NUMBER>
+          Prune transaction lookup data before the specified block number. The specified block number is not pruned
+
+      --prune.receipts.full
+          Prunes all receipt data
+
+      --prune.receipts.distance <BLOCKS>
+          Prune receipts before the `head-N` block number. In other words, keep last N + 1 blocks
+
+      --prune.receipts.before <BLOCK_NUMBER>
+          Prune receipts before the specified block number. The specified block number is not pruned
+
+      --prune.accounthistory.full
+          Prunes all account history
+
+      --prune.accounthistory.distance <BLOCKS>
+          Prune account before the `head-N` block number. In other words, keep last N + 1 blocks
+
+      --prune.accounthistory.before <BLOCK_NUMBER>
+          Prune account history before the specified block number. The specified block number is not pruned
+
+      --prune.storagehistory.full
+          Prunes all storage history data
+
+      --prune.storagehistory.distance <BLOCKS>
+          Prune storage history before the `head-N` block number. In other words, keep last N + 1 blocks
+
+      --prune.storagehistory.before <BLOCK_NUMBER>
+          Prune storage history before the specified block number. The specified block number is not pruned
+
+      --prune.receiptslogfilter <FILTER_CONFIG>
+          Configure receipts log filter. Format: <`address`>:<`prune_mode`>[,<`address`>:<`prune_mode`>...] Where <`prune_mode`> can be 'full', 'distance:<`blocks`>', or 'before:<`block_number`>'
 
 Engine:
       --engine.experimental

crates/node/builder/src/launch/common.rs

@@ -979,8 +979,29 @@ mod tests {
     fn test_save_prune_config() {
         with_tempdir("prune-store-test", |config_path| {
             let mut reth_config = Config::default();
-            let node_config =
-                NodeConfig { pruning: PruningArgs { full: true }, ..NodeConfig::test() };
+            let node_config = NodeConfig {
+                pruning: PruningArgs {
+                    full: true,
+                    block_interval: 0,
+                    sender_recovery_full: false,
+                    sender_recovery_distance: None,
+                    sender_recovery_before: None,
+                    transaction_lookup_full: false,
+                    transaction_lookup_distance: None,
+                    transaction_lookup_before: None,
+                    receipts_full: false,
+                    receipts_distance: None,
+                    receipts_before: None,
+                    account_history_full: false,
+                    account_history_distance: None,
+                    account_history_before: None,
+                    storage_history_full: false,
+                    storage_history_distance: None,
+                    storage_history_before: None,
+                    receipts_log_filter: vec![],
+                },
+                ..NodeConfig::test()
+            };
             LaunchContext::save_pruning_config_if_full_node(
                 &mut reth_config,
                 &node_config,

crates/node/core/Cargo.toml

@@ -57,6 +57,7 @@ derive_more.workspace = true
 toml.workspace = true
 serde.workspace = true
 strum = { workspace = true, features = ["derive"] }
+thiserror.workspace = true
 
 # io
 dirs-next = "2.0.0"

crates/node/core/src/args/error.rs

@@ -0,0 +1,22 @@
+use std::num::ParseIntError;
+
+/// Error while parsing a `ReceiptsLogPruneConfig`
+#[derive(thiserror::Error, Debug)]
+#[allow(clippy::enum_variant_names)]
+pub(crate) enum ReceiptsLogError {
+    /// The format of the filter is invalid.
+    #[error("invalid filter format: {0}")]
+    InvalidFilterFormat(String),
+    /// Address is invalid.
+    #[error("address is invalid: {0}")]
+    InvalidAddress(String),
+    /// The prune mode is not one of full, distance, before.
+    #[error("prune mode is invalid: {0}")]
+    InvalidPruneMode(String),
+    /// The distance value supplied is invalid.
+    #[error("distance is invalid: {0}")]
+    InvalidDistance(ParseIntError),
+    /// The block number supplied is invalid.
+    #[error("block number is invalid: {0}")]
+    InvalidBlockNumber(ParseIntError),
+}

crates/node/core/src/args/mod.rs

@@ -58,4 +58,5 @@ pub use benchmark_args::BenchmarkArgs;
 
 pub mod utils;
 
+mod error;
 pub mod types;

crates/node/core/src/args/pruning.rs

@@ -1,51 +1,250 @@
 //! Pruning and full node arguments
 
+use crate::args::error::ReceiptsLogError;
 use clap::Args;
 use reth_chainspec::ChainSpec;
 use reth_config::config::PruneConfig;
+use reth_primitives::{Address, BlockNumber};
 use reth_prune_types::{PruneMode, PruneModes, ReceiptsLogPruneConfig, MINIMUM_PRUNING_DISTANCE};
+use std::collections::BTreeMap;
 
 /// Parameters for pruning and full node
 #[derive(Debug, Clone, Args, PartialEq, Eq, Default)]
 #[command(next_help_heading = "Pruning")]
 pub struct PruningArgs {
     /// Run full node. Only the most recent [`MINIMUM_PRUNING_DISTANCE`] block states are stored.
-    /// This flag takes priority over pruning configuration in reth.toml.
     #[arg(long, default_value_t = false)]
     pub full: bool,
+
+    /// Minimum pruning interval measured in blocks.
+    #[arg(long, default_value_t = 0)]
+    pub block_interval: u64,
+
+    // Sender Recovery
+    /// Prunes all sender recovery data.
+    #[arg(long = "prune.senderrecovery.full", conflicts_with_all = &["sender_recovery_distance", "sender_recovery_before"])]
+    pub sender_recovery_full: bool,
+    /// Prune sender recovery data before the `head-N` block number. In other words, keep last N +
+    /// 1 blocks.
+    #[arg(long = "prune.senderrecovery.distance", value_name = "BLOCKS", conflicts_with_all = &["sender_recovery_full", "sender_recovery_before"])]
+    pub sender_recovery_distance: Option<u64>,
+    /// Prune sender recovery data before the specified block number. The specified block number is
+    /// not pruned.
+    #[arg(long = "prune.senderrecovery.before", value_name = "BLOCK_NUMBER", conflicts_with_all = &["sender_recovery_full", "sender_recovery_distance"])]
+    pub sender_recovery_before: Option<BlockNumber>,
+
+    // Transaction Lookup
+    /// Prunes all transaction lookup data.
+    #[arg(long = "prune.transactionlookup.full", conflicts_with_all = &["transaction_lookup_distance", "transaction_lookup_before"])]
+    pub transaction_lookup_full: bool,
+    /// Prune transaction lookup data before the `head-N` block number. In other words, keep last N
+    /// + 1 blocks.
+    #[arg(long = "prune.transactionlookup.distance", value_name = "BLOCKS", conflicts_with_all = &["transaction_lookup_full", "transaction_lookup_before"])]
+    pub transaction_lookup_distance: Option<u64>,
+    /// Prune transaction lookup data before the specified block number. The specified block number
+    /// is not pruned.
+    #[arg(long = "prune.transactionlookup.before", value_name = "BLOCK_NUMBER", conflicts_with_all = &["transaction_lookup_full", "transaction_lookup_distance"])]
+    pub transaction_lookup_before: Option<BlockNumber>,
+
+    // Receipts
+    /// Prunes all receipt data.
+    #[arg(long = "prune.receipts.full", conflicts_with_all = &["receipts_distance", "receipts_before"])]
+    pub receipts_full: bool,
+    /// Prune receipts before the `head-N` block number. In other words, keep last N + 1 blocks.
+    #[arg(long = "prune.receipts.distance", value_name = "BLOCKS", conflicts_with_all = &["receipts_full", "receipts_before"])]
+    pub receipts_distance: Option<u64>,
+    /// Prune receipts before the specified block number. The specified block number is not pruned.
+    #[arg(long = "prune.receipts.before", value_name = "BLOCK_NUMBER", conflicts_with_all = &["receipts_full", "receipts_distance"])]
+    pub receipts_before: Option<BlockNumber>,
+
+    // Account History
+    /// Prunes all account history.
+    #[arg(long = "prune.accounthistory.full", conflicts_with_all = &["account_history_distance", "account_history_before"])]
+    pub account_history_full: bool,
+    /// Prune account before the `head-N` block number. In other words, keep last N + 1 blocks.
+    #[arg(long = "prune.accounthistory.distance", value_name = "BLOCKS", conflicts_with_all = &["account_history_full", "account_history_before"])]
+    pub account_history_distance: Option<u64>,
+    /// Prune account history before the specified block number. The specified block number is not
+    /// pruned.
+    #[arg(long = "prune.accounthistory.before", value_name = "BLOCK_NUMBER", conflicts_with_all = &["account_history_full", "account_history_distance"])]
+    pub account_history_before: Option<BlockNumber>,
+
+    // Storage History
+    /// Prunes all storage history data.
+    #[arg(long = "prune.storagehistory.full", conflicts_with_all = &["storage_history_distance", "storage_history_before"])]
+    pub storage_history_full: bool,
+    /// Prune storage history before the `head-N` block number. In other words, keep last N + 1
+    /// blocks.
+    #[arg(long = "prune.storagehistory.distance", value_name = "BLOCKS", conflicts_with_all = &["storage_history_full", "storage_history_before"])]
+    pub storage_history_distance: Option<u64>,
+    /// Prune storage history before the specified block number. The specified block number is not
+    /// pruned.
+    #[arg(long = "prune.storagehistory.before", value_name = "BLOCK_NUMBER", conflicts_with_all = &["storage_history_full", "storage_history_distance"])]
+    pub storage_history_before: Option<BlockNumber>,
+
+    // Receipts Log Filter
+    /// Configure receipts log filter. Format:
+    /// <`address`>:<`prune_mode`>[,<`address`>:<`prune_mode`>...] Where <`prune_mode`> can be
+    /// 'full', 'distance:<`blocks`>', or 'before:<`block_number`>'
+    #[arg(long = "prune.receiptslogfilter", value_name = "FILTER_CONFIG", value_delimiter = ',', value_parser = parse_receipts_log_filter)]
+    pub receipts_log_filter: Vec<String>,
 }
 
 impl PruningArgs {
     /// Returns pruning configuration.
     pub fn prune_config(&self, chain_spec: &ChainSpec) -> Option<PruneConfig> {
-        if !self.full {
-            return None
-        }
+        // Initialise with a default prune configuration.
+        let mut config = PruneConfig::default();
 
-        Some(PruneConfig {
-            block_interval: 5,
-            segments: PruneModes {
-                sender_recovery: Some(PruneMode::Full),
-                transaction_lookup: None,
-                // prune all receipts if chain doesn't have deposit contract specified in chain spec
-                receipts: chain_spec
-                    .deposit_contract
-                    .as_ref()
-                    .map(|contract| PruneMode::Before(contract.block))
-                    .or(Some(PruneMode::Distance(MINIMUM_PRUNING_DISTANCE))),
-                account_history: Some(PruneMode::Distance(MINIMUM_PRUNING_DISTANCE)),
-                storage_history: Some(PruneMode::Distance(MINIMUM_PRUNING_DISTANCE)),
-                receipts_log_filter: ReceiptsLogPruneConfig(
-                    chain_spec
+        // If --full is set, use full node defaults.
+        if self.full {
+            config = PruneConfig {
+                block_interval: 5,
+                segments: PruneModes {
+                    sender_recovery: Some(PruneMode::Full),
+                    transaction_lookup: None,
+                    // prune all receipts if chain doesn't have deposit contract specified in chain
+                    // spec
+                    receipts: chain_spec
                         .deposit_contract
                         .as_ref()
-                        .map(|contract| (contract.address, PruneMode::Before(contract.block)))
-                        .into_iter()
-                        .collect(),
-                ),
-            },
-        })
+                        .map(|contract| PruneMode::Before(contract.block))
+                        .or(Some(PruneMode::Distance(MINIMUM_PRUNING_DISTANCE))),
+                    account_history: Some(PruneMode::Distance(MINIMUM_PRUNING_DISTANCE)),
+                    storage_history: Some(PruneMode::Distance(MINIMUM_PRUNING_DISTANCE)),
+                    receipts_log_filter: ReceiptsLogPruneConfig(
+                        chain_spec
+                            .deposit_contract
+                            .as_ref()
+                            .map(|contract| (contract.address, PruneMode::Before(contract.block)))
+                            .into_iter()
+                            .collect(),
+                    ),
+                },
+            }
+        }
+
+        // Override with any explicitly set prune.* flags.
+        if let Some(mode) = self.sender_recovery_prune_mode() {
+            config.segments.sender_recovery = Some(mode);
+        }
+        if let Some(mode) = self.transaction_lookup_prune_mode() {
+            config.segments.transaction_lookup = Some(mode);
+        }
+        if let Some(mode) = self.receipts_prune_mode() {
+            config.segments.receipts = Some(mode);
+        }
+        if let Some(mode) = self.account_history_prune_mode() {
+            config.segments.account_history = Some(mode);
+        }
+        if let Some(mode) = self.storage_history_prune_mode() {
+            config.segments.storage_history = Some(mode);
+        }
+
+        Some(config)
     }
+    const fn sender_recovery_prune_mode(&self) -> Option<PruneMode> {
+        if self.sender_recovery_full {
+            Some(PruneMode::Full)
+        } else if let Some(distance) = self.sender_recovery_distance {
+            Some(PruneMode::Distance(distance))
+        } else if let Some(block_number) = self.sender_recovery_before {
+            Some(PruneMode::Before(block_number))
+        } else {
+            None
+        }
+    }
+
+    const fn transaction_lookup_prune_mode(&self) -> Option<PruneMode> {
+        if self.transaction_lookup_full {
+            Some(PruneMode::Full)
+        } else if let Some(distance) = self.transaction_lookup_distance {
+            Some(PruneMode::Distance(distance))
+        } else if let Some(block_number) = self.transaction_lookup_before {
+            Some(PruneMode::Before(block_number))
+        } else {
+            None
+        }
+    }
+
+    const fn receipts_prune_mode(&self) -> Option<PruneMode> {
+        if self.receipts_full {
+            Some(PruneMode::Full)
+        } else if let Some(distance) = self.receipts_distance {
+            Some(PruneMode::Distance(distance))
+        } else if let Some(block_number) = self.receipts_before {
+            Some(PruneMode::Before(block_number))
+        } else {
+            None
+        }
+    }
+
+    const fn account_history_prune_mode(&self) -> Option<PruneMode> {
+        if self.account_history_full {
+            Some(PruneMode::Full)
+        } else if let Some(distance) = self.account_history_distance {
+            Some(PruneMode::Distance(distance))
+        } else if let Some(block_number) = self.account_history_before {
+            Some(PruneMode::Before(block_number))
+        } else {
+            None
+        }
+    }
+
+    const fn storage_history_prune_mode(&self) -> Option<PruneMode> {
+        if self.storage_history_full {
+            Some(PruneMode::Full)
+        } else if let Some(distance) = self.storage_history_distance {
+            Some(PruneMode::Distance(distance))
+        } else if let Some(block_number) = self.storage_history_before {
+            Some(PruneMode::Before(block_number))
+        } else {
+            None
+        }
+    }
+}
+
+pub(crate) fn parse_receipts_log_filter(
+    value: &str,
+) -> Result<ReceiptsLogPruneConfig, ReceiptsLogError> {
+    let mut config = BTreeMap::new();
+    // Split out each of the filters.
+    let filters = value.split(',');
+    for filter in filters {
+        let parts: Vec<&str> = filter.split(':').collect();
+        if parts.len() < 2 {
+            return Err(ReceiptsLogError::InvalidFilterFormat(filter.to_string()));
+        }
+        // Parse the address
+        let address = parts[0]
+            .parse::<Address>()
+            .map_err(|_| ReceiptsLogError::InvalidAddress(parts[0].to_string()))?;
+
+        // Parse the prune mode
+        let prune_mode = match parts[1] {
+            "full" => PruneMode::Full,
+            s if s.starts_with("distance") => {
+                if parts.len() < 3 {
+                    return Err(ReceiptsLogError::InvalidFilterFormat(filter.to_string()));
+                }
+                let distance =
+                    parts[2].parse::<u64>().map_err(ReceiptsLogError::InvalidDistance)?;
+                PruneMode::Distance(distance)
+            }
+            s if s.starts_with("before") => {
+                if parts.len() < 3 {
+                    return Err(ReceiptsLogError::InvalidFilterFormat(filter.to_string()));
+                }
+                let block_number = parts[2]
+                    .parse::<BlockNumber>()
+                    .map_err(ReceiptsLogError::InvalidBlockNumber)?;
+                PruneMode::Before(block_number)
+            }
+            _ => return Err(ReceiptsLogError::InvalidPruneMode(parts[1].to_string())),
+        };
+        config.insert(address, prune_mode);
+    }
+    Ok(ReceiptsLogPruneConfig(config))
 }
 
 #[cfg(test)]
@@ -66,4 +265,62 @@ mod tests {
         let args = CommandParser::<PruningArgs>::parse_from(["reth"]).args;
         assert_eq!(args, default_args);
     }
+
+    #[test]
+    fn test_parse_receipts_log_filter() {
+        let filter1 = "0x0000000000000000000000000000000000000001:full";
+        let filter2 = "0x0000000000000000000000000000000000000002:distance:1000";
+        let filter3 = "0x0000000000000000000000000000000000000003:before:5000000";
+        let filters = [filter1, filter2, filter3].join(",");
+
+        // Args can be parsed.
+        let result = parse_receipts_log_filter(&filters);
+        assert!(result.is_ok());
+        let config = result.unwrap();
+        assert_eq!(config.0.len(), 3);
+
+        // Check that the args were parsed correctly.
+        let addr1: Address = "0x0000000000000000000000000000000000000001".parse().unwrap();
+        let addr2: Address = "0x0000000000000000000000000000000000000002".parse().unwrap();
+        let addr3: Address = "0x0000000000000000000000000000000000000003".parse().unwrap();
+
+        assert_eq!(config.0.get(&addr1), Some(&PruneMode::Full));
+        assert_eq!(config.0.get(&addr2), Some(&PruneMode::Distance(1000)));
+        assert_eq!(config.0.get(&addr3), Some(&PruneMode::Before(5000000)));
+    }
+
+    #[test]
+    fn test_parse_receipts_log_filter_invalid_filter_format() {
+        let result = parse_receipts_log_filter("invalid_format");
+        assert!(matches!(result, Err(ReceiptsLogError::InvalidFilterFormat(_))));
+    }
+
+    #[test]
+    fn test_parse_receipts_log_filter_invalid_address() {
+        let result = parse_receipts_log_filter("invalid_address:full");
+        assert!(matches!(result, Err(ReceiptsLogError::InvalidAddress(_))));
+    }
+
+    #[test]
+    fn test_parse_receipts_log_filter_invalid_prune_mode() {
+        let result =
+            parse_receipts_log_filter("0x0000000000000000000000000000000000000000:invalid_mode");
+        assert!(matches!(result, Err(ReceiptsLogError::InvalidPruneMode(_))));
+    }
+
+    #[test]
+    fn test_parse_receipts_log_filter_invalid_distance() {
+        let result = parse_receipts_log_filter(
+            "0x0000000000000000000000000000000000000000:distance:invalid_distance",
+        );
+        assert!(matches!(result, Err(ReceiptsLogError::InvalidDistance(_))));
+    }
+
+    #[test]
+    fn test_parse_receipts_log_filter_invalid_block_number() {
+        let result = parse_receipts_log_filter(
+            "0x0000000000000000000000000000000000000000:before:invalid_block",
+        );
+        assert!(matches!(result, Err(ReceiptsLogError::InvalidBlockNumber(_))));
+    }
 }

crates/node/core/src/node_config.rs

@@ -229,7 +229,7 @@ impl NodeConfig {
     }
 
     /// Set the pruning args for the node
-    pub const fn with_pruning(mut self, pruning: PruningArgs) -> Self {
+    pub fn with_pruning(mut self, pruning: PruningArgs) -> Self {
         self.pruning = pruning;
         self
     }