wgtmac commented on code in PR #3526:
URL: https://github.com/apache/parquet-java/pull/3526#discussion_r3510148378
##########
parquet-hadoop/src/main/java/org/apache/parquet/hadoop/CodecFactory.java:
##########
@@ -252,10 +260,22 @@ public CompressionCodecName getCodecName() {
@Override
public BytesCompressor getCompressor(CompressionCodecName codecName) {
- BytesCompressor comp = compressors.get(codecName);
+ String key = cacheKey(codecName);
+ BytesCompressor comp = compressors.get(key);
if (comp == null) {
comp = createCompressor(codecName);
- compressors.put(codecName, comp);
+ compressors.put(key, comp);
+ }
+ return comp;
+ }
+
+ @Override
+ public BytesCompressor getCompressor(CompressionCodecName codecName, int
level) {
Review Comment:
This path bypasses `DirectCodecFactory#createCompressor(...)`. With a direct
codec factory, per-column levels for ZSTD/SNAPPY can fall back to the
heap/Hadoop compressor path; please add a direct-factory override or test.
(reviewed by Codex)
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
