Skip to content

Commit 73507c3

Browse files
authored
organize ExternalSorter fields (#13447)
1 parent 06db9ed commit 73507c3

File tree

1 file changed

+29
-14
lines changed
  • datafusion/physical-plan/src/sorts

1 file changed

+29
-14
lines changed

datafusion/physical-plan/src/sorts/sort.rs

Lines changed: 29 additions & 14 deletions
Original file line numberDiff line numberDiff line change
@@ -203,39 +203,54 @@ impl ExternalSorterMetrics {
203203
/// in_mem_batches
204204
/// ```
205205
struct ExternalSorter {
206-
/// schema of the output (and the input)
206+
// ========================================================================
207+
// PROPERTIES:
208+
// Fields that define the sorter's configuration and remain constant
209+
// ========================================================================
210+
/// Schema of the output (and the input)
207211
schema: SchemaRef,
212+
/// Sort expressions
213+
expr: Arc<[PhysicalSortExpr]>,
214+
/// If Some, the maximum number of output rows that will be produced
215+
fetch: Option<usize>,
216+
/// The target number of rows for output batches
217+
batch_size: usize,
218+
/// If the in size of buffered memory batches is below this size,
219+
/// the data will be concatenated and sorted in place rather than
220+
/// sort/merged.
221+
sort_in_place_threshold_bytes: usize,
222+
223+
// ========================================================================
224+
// STATE BUFFERS:
225+
// Fields that hold intermediate data during sorting
226+
// ========================================================================
208227
/// Potentially unsorted in memory buffer
209228
in_mem_batches: Vec<RecordBatch>,
210229
/// if `Self::in_mem_batches` are sorted
211230
in_mem_batches_sorted: bool,
231+
212232
/// If data has previously been spilled, the locations of the
213233
/// spill files (in Arrow IPC format)
214234
spills: Vec<RefCountedTempFile>,
215-
/// Sort expressions
216-
expr: Arc<[PhysicalSortExpr]>,
235+
236+
// ========================================================================
237+
// EXECUTION RESOURCES:
238+
// Fields related to managing execution resources and monitoring performance.
239+
// ========================================================================
217240
/// Runtime metrics
218241
metrics: ExternalSorterMetrics,
219-
/// If Some, the maximum number of output rows that will be
220-
/// produced.
221-
fetch: Option<usize>,
242+
/// A handle to the runtime to get spill files
243+
runtime: Arc<RuntimeEnv>,
222244
/// Reservation for in_mem_batches
223245
reservation: MemoryReservation,
246+
224247
/// Reservation for the merging of in-memory batches. If the sort
225248
/// might spill, `sort_spill_reservation_bytes` will be
226249
/// pre-reserved to ensure there is some space for this sort/merge.
227250
merge_reservation: MemoryReservation,
228-
/// A handle to the runtime to get spill files
229-
runtime: Arc<RuntimeEnv>,
230-
/// The target number of rows for output batches
231-
batch_size: usize,
232251
/// How much memory to reserve for performing in-memory sort/merges
233252
/// prior to spilling.
234253
sort_spill_reservation_bytes: usize,
235-
/// If the in size of buffered memory batches is below this size,
236-
/// the data will be concatenated and sorted in place rather than
237-
/// sort/merged.
238-
sort_in_place_threshold_bytes: usize,
239254
}
240255

241256
impl ExternalSorter {

0 commit comments

Comments
 (0)