Merge pull request #2904 from kuzudb/parquet-null-fix

Fix parquet null
kuzudb · Feb 16, 2024 · 851a6e7 · 851a6e7
2 parents abf48f5 + 7055328
commit 851a6e7
Show file tree

Hide file tree

Showing 3 changed files with 4 additions and 3 deletions.
diff --git a/src/include/processor/operator/persistent/reader/parquet/column_reader.h b/src/include/processor/operator/persistent/reader/parquet/column_reader.h
@@ -72,6 +72,7 @@ class ColumnReader {
                 result->setNull(i + resultOffset, true);
                 continue;
             }
+            result->setNull(i + resultOffset, false);
             if (filter[i + resultOffset]) {
                 VALUE_TYPE val = CONVERSION::plainRead(*plainData, *this);
                 result->setValue(i + resultOffset, val);

diff --git a/src/processor/operator/persistent/reader/parquet/list_column_reader.cpp b/src/processor/operator/persistent/reader/parquet/list_column_reader.cpp
@@ -95,10 +95,12 @@ uint64_t ListColumnReader::read(uint64_t numValues, parquet_filter_t& /*filter*/
                 break;
             }
             if (childDefinesPtr[childIdx] >= maxDefine) {
+                resultOut->setNull(resultOffset, false);
                 // value has been defined down the stack, hence its NOT NULL
                 resultPtr[resultOffset].offset = childIdx + currentChunkOffset;
                 resultPtr[resultOffset].size = 1;
             } else if (childDefinesPtr[childIdx] == maxDefine - 1) {
+                resultOut->setNull(resultOffset, false);
                 resultPtr[resultOffset].offset = childIdx + currentChunkOffset;
                 resultPtr[resultOffset].size = 0;
             } else {

diff --git a/src/processor/operator/persistent/reader/parquet/struct_column_reader.cpp b/src/processor/operator/persistent/reader/parquet/struct_column_reader.cpp
@@ -58,9 +58,7 @@ uint64_t StructColumnReader::read(uint64_t numValuesToRead, parquet_filter_t& fi
         }
     }
     for (auto i = 0u; i < numValuesRead; i++) {
-        if (define_out[i] < maxDefine) {
-            result->setNull(i, true);
-        }
+        result->setNull(i, define_out[i] < maxDefine);
     }
 
     return numValuesRead;