danielcweeks commented on code in PR #16614:
URL: https://github.com/apache/iceberg/pull/16614#discussion_r3382926644
##########
parquet/src/main/java/org/apache/iceberg/parquet/Parquet.java:
##########
@@ -1595,6 +1573,53 @@ public <D> CloseableIterable<D> build() {
}
}
+ @VisibleForTesting
+ static ParquetReadOptions buildReadOptions(
+ InputFile file,
+ Map<String, String> properties,
+ Long start,
+ Long length,
+ FileDecryptionProperties fileDecryptionProperties) {
+ ParquetReadOptions.Builder optionsBuilder;
+ HadoopInputFile hadoopInputFile = null;
+ if (HadoopInputFile.class.isInstance(file)) {
+ hadoopInputFile = HadoopInputFile.class.cast(file);
+ }
+
+ if (hadoopInputFile != null) {
+ // remove read properties already set that may conflict with this read
+ Configuration conf = new Configuration(hadoopInputFile.getConf());
+ for (String property : READ_PROPERTIES_TO_REMOVE) {
+ conf.unset(property);
+ }
+ optionsBuilder = HadoopReadOptions.builder(conf);
+ } else {
+ optionsBuilder = ParquetReadOptions.builder(new
PlainParquetConfiguration());
+ }
+
+ for (Map.Entry<String, String> entry : properties.entrySet()) {
+ optionsBuilder.set(entry.getKey(), entry.getValue());
+ }
+
+ if (start != null) {
+ optionsBuilder.withRange(start, start + length);
+ }
+
+ if (fileDecryptionProperties != null) {
+ optionsBuilder.withDecryption(fileDecryptionProperties);
+ }
+
+ if (properties.containsKey(ParquetInputFormat.HADOOP_VECTORED_IO_ENABLED))
{
+ optionsBuilder.withUseHadoopVectoredIo(
Review Comment:
Have you traced this to ensure this actually turns off Vectored IO behavior
from the Iceberg side? I think this will disable it from parquet, but the IO
pattern from the surrounding Iceberg implementation may still go through the
Vectored IO api.
--
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]
