emkornfield commented on code in PR #17877:
URL: https://github.com/apache/arrow/pull/17877#discussion_r1066601116
##########
cpp/src/parquet/column_reader.h:
##########
@@ -278,9 +278,13 @@ class PARQUET_EXPORT RecordReader {
/// \brief Attempt to read indicated number of records from column chunk
/// Note that for repeated fields, a record may have more than one value
- /// and all of them are read.
+ /// and all of them are read. If read_dense_for_nullable is true, it will
+ /// not leave any space for null values. Otherwise, it will read spaced.
+ /// Readers must call Reset() before switching between reading dense and
+ /// spaced since reading dense will not update the valid_bits_.
/// \return number of records read
- virtual int64_t ReadRecords(int64_t num_records) = 0;
+ virtual int64_t ReadRecords(int64_t num_records,
+ bool read_dense_for_nullable = false) = 0;
Review Comment:
Thinking about this API, is there a reason you can think of for someone
between to switch between Dense/Non-Dense? Would it make sense pass this as an
option to Make instead?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
