fix parsing array struct when repetition_type is 'REPEATED'

yma11 · Aug 14, 2024 · 097d30f · 097d30f
1 parent 13899de
commit 097d30f
Show file tree

Hide file tree

Showing 2 changed files with 17 additions and 0 deletions.
diff --git a/velox/dwio/parquet/reader/ParquetReader.cpp b/velox/dwio/parquet/reader/ParquetReader.cpp
@@ -464,6 +464,23 @@ std::unique_ptr<ParquetTypeWithId> ReaderBase::getParquetColumnInfo(
               maxDefine,
               isOptional,
               isRepeated);
+        } else {
+          // Row type
+          auto type =
+              createRowType(children, isFileColumnNamesReadAsLowerCase());
+          return std::make_unique<ParquetTypeWithId>(
+              std::move(type),
+              std::move(children),
+              curSchemaIdx,
+              maxSchemaElementIdx,
+              ParquetTypeWithId::kNonLeaf, // columnIdx,
+              std::move(name),
+              std::nullopt,
+              std::nullopt,
+              maxRepeat,
+              maxDefine,
+              isOptional,
+              isRepeated);
         }
       } else {
         // Row type

diff --git a/velox/dwio/parquet/tests/examples/proto-struct-with-array.parquet b/velox/dwio/parquet/tests/examples/proto-struct-with-array.parquet