chenjunjiedada commented on a change in pull request #830:
URL: https://github.com/apache/iceberg/pull/830#discussion_r432219756
##########
File path:
parquet/src/main/java/org/apache/iceberg/parquet/ParquetDictionaryRowGroupFilter.java
##########
@@ -108,11 +118,15 @@ private boolean eval(MessageType fileSchema,
BlockMetaData rowGroup,
}
for (ColumnChunkMetaData meta : rowGroup.getColumns()) {
- PrimitiveType colType =
fileSchema.getType(meta.getPath().toArray()).asPrimitiveType();
- if (colType.getId() != null) {
- int id = colType.getId().intValue();
- isFallback.put(id, ParquetUtil.hasNonDictionaryPages(meta));
- mayContainNulls.put(id, mayContainNull(meta));
+ try {
+ PrimitiveType colType =
fileSchema.getType(meta.getPath().toArray()).asPrimitiveType();
+ if (colType.getId() != null) {
+ int id = colType.getId().intValue();
+ isFallback.put(id, ParquetUtil.hasNonDictionaryPages(meta));
+ mayContainNulls.put(id, mayContainNull(meta));
+ }
+ } catch (org.apache.parquet.io.InvalidRecordException e) {
Review comment:
@rdblue , do you think we need to change both `PruneColumns` and
`SparkParquetReader#ReadBuilder` to satisfy the assumption? It may change a lot.
----------------------------------------------------------------
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
