vvysotskyi commented on a change in pull request #1667: DRILL-7069: Moved
version checks outside loops in transformBinaryInMetadataCache
URL: https://github.com/apache/drill/pull/1667#discussion_r261519267
##########
File path:
exec/java-exec/src/main/java/org/apache/drill/exec/store/parquet/ParquetReaderUtility.java
##########
@@ -292,30 +294,47 @@ public static void
transformBinaryInMetadataCache(ParquetTableMetadataBase parqu
Set<List<String>> columnsNames =
getBinaryColumnsNames(parquetTableMetadata);
boolean allowBinaryMetadata =
allowBinaryMetadata(parquetTableMetadata.getDrillVersion(), readerConfig);
+ // Setting Min / Max values for ParquetTableMetadata_v1
+ if (new MetadataVersion(1, 0).equals(new
MetadataVersion(parquetTableMetadata.getMetadataVersion()))) {
+ for (ParquetFileMetadata file : parquetTableMetadata.getFiles()) {
+ for (RowGroupMetadata rowGroupMetadata : file.getRowGroups()) {
+ Long rowCount = rowGroupMetadata.getRowCount();
+ for (ColumnMetadata columnMetadata : rowGroupMetadata.getColumns()) {
+ if (columnMetadata.getPrimitiveType() ==
PrimitiveTypeName.BINARY || columnMetadata.getPrimitiveType() ==
PrimitiveTypeName.FIXED_LEN_BYTE_ARRAY) {
+ setMinMaxValues(columnMetadata, rowCount, allowBinaryMetadata,
false);
+ }
+ }
+ }
+ }
+ return;
+ }
+
+ // Variables needed for debugging only
+ Stopwatch timer = logger.isDebugEnabled() ? Stopwatch.createStarted() :
null;
+ int maxRowGroups = 0, minRowGroups = 99999, maxNumColumns = 0;
+
+ // Setting Min / Max values for V2 and V3 versions; for versions V3_3 and
above need to do decoding
+ boolean needDecoding = new
MetadataVersion(parquetTableMetadata.getMetadataVersion()).compareTo(new
MetadataVersion(3, 3)) >= 0;
for (ParquetFileMetadata file : parquetTableMetadata.getFiles()) {
+ maxRowGroups = Math.max(maxRowGroups,file.getRowGroups().size());
+ minRowGroups = Math.min(minRowGroups,file.getRowGroups().size());
for (RowGroupMetadata rowGroupMetadata : file.getRowGroups()) {
Long rowCount = rowGroupMetadata.getRowCount();
+ maxNumColumns =
Math.max(maxNumColumns,rowGroupMetadata.getColumns().size());
for (ColumnMetadata columnMetadata : rowGroupMetadata.getColumns()) {
- // Setting Min / Max values for ParquetTableMetadata_v1
- if (new MetadataVersion(1, 0).equals(new
MetadataVersion(parquetTableMetadata.getMetadataVersion()))) {
- if (columnMetadata.getPrimitiveType() == PrimitiveTypeName.BINARY
- || columnMetadata.getPrimitiveType() ==
PrimitiveTypeName.FIXED_LEN_BYTE_ARRAY) {
- setMinMaxValues(columnMetadata, rowCount, allowBinaryMetadata,
false);
- }
- }
- // Setting Min / Max values for V2 and all V3 versions prior to V3_3
- else if (new
MetadataVersion(parquetTableMetadata.getMetadataVersion()).compareTo(new
MetadataVersion(3, 3)) < 0
- &&
columnsNames.contains(Arrays.asList(columnMetadata.getName()))) {
- setMinMaxValues(columnMetadata, rowCount, allowBinaryMetadata,
false);
- }
- // Setting Min / Max values for V3_3 and all next versions
- else if (new
MetadataVersion(parquetTableMetadata.getMetadataVersion()).compareTo(new
MetadataVersion(3, 3)) >= 0
- &&
columnsNames.contains(Arrays.asList(columnMetadata.getName()))) {
- setMinMaxValues(columnMetadata, rowCount, allowBinaryMetadata,
true);
+ if (columnsNames.contains(Arrays.asList(columnMetadata.getName())))
{
+ setMinMaxValues(columnMetadata, rowCount, allowBinaryMetadata,
needDecoding);
}
}
}
}
+
+ if (timer != null) { // log a debug message and stop the timer
+ String reportRG = 1 == maxRowGroups ? "1 rowgroup" : "between " +
minRowGroups + "-" + maxRowGroups + "rowgroups";
+ logger.debug("Transforming binary in metadata cache took {} ms ({}
files, {} per file, max {} columns)", timer.elapsed(TimeUnit.MILLISECONDS),
+ parquetTableMetadata.getFiles().size(),reportRG,maxNumColumns);
Review comment:
And please add spaces here:
`parquetTableMetadata.getFiles().size(),reportRG,maxNumColumns);` ->
`parquetTableMetadata.getFiles().size(), reportRG, maxNumColumns);`
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
