tustvold commented on a change in pull request #1228:
URL: https://github.com/apache/arrow-rs/pull/1228#discussion_r795044493
##########
File path: parquet/src/arrow/bit_util.rs
##########
@@ -15,40 +15,33 @@
// specific language governing permissions and limitations
// under the License.
-use arrow::util::bit_chunk_iterator::BitChunks;
+use arrow::util::bit_chunk_iterator::UnalignedBitChunk;
use std::ops::Range;
/// Counts the number of set bits in the provided range
pub fn count_set_bits(bytes: &[u8], range: Range<usize>) -> usize {
- let mut count = 0_usize;
- let chunks = BitChunks::new(bytes, range.start, range.end - range.start);
- chunks.iter().for_each(|chunk| {
- count += chunk.count_ones() as usize;
- });
- count += chunks.remainder_bits().count_ones() as usize;
- count
+ let unaligned = UnalignedBitChunk::new(bytes, range.start, range.end -
range.start);
Review comment:
The slice of bits being iterated on isn't even byte aligned, as range is
in bits not bytes. The reason UnalignedBitChunk is faster here is it doesn't
try to produce aligned output, which doesn't matter for counting set bits
--
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]
