wesm commented on a change in pull request #7789:
URL: https://github.com/apache/arrow/pull/7789#discussion_r465310364
##########
File path: cpp/src/arrow/util/compression.cc
##########
@@ -131,7 +131,7 @@ Result<std::unique_ptr<Codec>>
Codec::Create(Compression::type codec_type,
if (compression_level_set) {
return Status::Invalid("LZ4 doesn't support setting a compression
level.");
}
- codec = internal::MakeLz4RawCodec();
+ codec = internal::MakeLz4HadoopRawCodec();
Review comment:
No. Per below we now have 3 codec implementations:
* LZ4 raw
* LZ4 raw with Hadoop prefix
* LZ4 frame
This change makes "LZ4 raw" inaccessible. When we see "LZ4" in the Parquet
metadata we should always instantiate the LZ4_HADOOP codec which has the logic
to infer raw vs. raw w/ hadoop prefix. Does that make sense?
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
