Struct ExecutionContext 

pub struct ExecutionContext<'p> {
    pub processor: &'p QueryProcessor,
    pub function_registry: Option<Arc<Mutex<FunctionRegistry>>>,
    pub table_catalog: Option<Arc<TableCatalog>>,
    pub vfs: Option<Arc<VirtualFileSystemRegistry>>,
    pub standalone_call_handler: Option<Arc<dyn StandaloneCallHandler>>,
    pub sequence_fn: Option<SequenceFn>,
    pub subquery_fn: Option<SubqueryFn>,
    pub schema_ddl_fn: Option<SchemaDdlFn>,
    pub snapshot_ts: Option<u64>,
    pub commit_history: Vec<(u64, u64)>,
    pub written_rows: Vec<(u64, u64)>,
    pub txn_id: Option<u64>,
}

Shared state threaded through the mapper functions

Fields

processor: &'p QueryProcessor

function_registry: Option<Arc<Mutex<FunctionRegistry>>>

table_catalog: Option<Arc<TableCatalog>>

vfs: Option<Arc<VirtualFileSystemRegistry>>

standalone_call_handler: Option<Arc<dyn StandaloneCallHandler>>

sequence_fn: Option<SequenceFn>

subquery_fn: Option<SubqueryFn>

schema_ddl_fn: Option<SchemaDdlFn>

snapshot_ts: Option<u64>

MVCC snapshot timestamp. When Some(ts), reads are isolated to data committed at or before ts. None means read最新 (no isolation).

commit_history: Vec<(u64, u64)>

Commit history for MVCC visibility checks: (txn_id, commit_ts) pairs.

written_rows: Vec<(u64, u64)>

Row-level write set for OCC conflict detection. Populated by the mapper after each write operation (SET, DELETE, INSERT). The connection layer reads this after execution and calls record_write().

txn_id: Option<u64>

Active transaction id threaded into write operators so inserts/deletes use MVCC-aware storage variants (P52.18).

Implementations

impl<'p> ExecutionContext<'p>

pub fn execute_children( &mut self, operators: &[LogicalOperator], ) -> Result<Vec<DataChunk>, ProcessorError>

pub fn resolve_scan_data<'b>( &self, table_name: &str, predicate: Option<(usize, &'b str, &'b Value)>, ) -> (Option<Vec<Vec<Value>>>, Vec<ColumnDefinition>, u64)

Resolve table data and column definitions for a scan node. When snapshot_ts is set on the context, uses MVCC-aware scan.

pub fn resolve_scan_arrow_data( &self, table_name: &str, ) -> (Option<Vec<ArrayRef>>, Vec<ColumnDefinition>, u64)

Resolve scan data directly into Arrow arrays, bypassing the Vec<Vec<Value>> intermediate materialization.

Reads from NodeTable’s NodeGroup column chunks, converts each ColumnChunk to an Arrow array, then concatenates per-group arrays into one array per column.

When snapshot_ts is set, falls back to the Vec<Vec> path since Arrow arrays don’t support MVCC version chain traversal.