Expand description
Bulk part encoder/decoder.
Structsยง
- Batch
Pruning ๐Stats - Adapter implementing
PruningStatisticsforBatchStats. - Batch
Stats ๐ - Per-batch min/max statistics for the first tag column in a
MultiBulkPart. - Bulk
Part - A raw bulk part in the memtable.
- Bulk
Part Converter - Converter that converts structs into BulkPart.
- Bulk
Part Encode Metrics - Metrics for encoding a part.
- Bulk
Part Encoder - Bulk
Part Meta - Encoded
Bulk Part - Multi
Bulk Part - A collection of ordered RecordBatches representing a bulk part without parquet encoding.
- Unordered
Part - A collection of small unordered bulk parts. Used to batch small parts together before merging them into a sorted part.
Enumsยง
- Primary
KeyColumn ๐Builder - Primary key column builder for handling strings specially.
Constantsยง
Functionsยง
- align_
parts ๐ - Align the JSON columns in BulkParts, to unified Arrow arrays. So that we can compute (concat, sort, etc.) on them.
- align_
schema_ ๐with_ json_ array - convert_
bulk_ part - Converts a
BulkPartthat is unordered and without encoded primary keys into aBulkPartwith the same format as produced by BulkPartConverter. - new_
primary_ ๐key_ column_ builders - predicate_
references_ ๐column - Returns true if the predicate references the given column name.
- record_
batch_ estimated_ size - More accurate estimation of the size of a record batch.
- should_
prune_ ๐bulk_ part - Returns true if the batch should be pruned (skipped) based on the first-tag min/max statistics and the predicate in the context. Returns false if no pruning is possible (no primary key, no predicate, or the batch matches the predicate).
- sort_
primary_ key_ record_ batch - Sorts the record batch with primary key format.