[Fix](parquet-reader) Fix parquet reader crash in set_dict(). (#41307)

## Proposed changes Backport #40643
apache · Sep 26, 2024 · 2292a69 · 2292a69
1 parent 9ae7165
commit 2292a69
Show file tree

Hide file tree

Showing 2 changed files with 6 additions and 0 deletions.
diff --git a/be/src/vec/exec/format/parquet/byte_array_dict_decoder.cpp b/be/src/vec/exec/format/parquet/byte_array_dict_decoder.cpp
@@ -32,6 +32,9 @@ namespace doris::vectorized {
 Status ByteArrayDictDecoder::set_dict(std::unique_ptr<uint8_t[]>& dict, int32_t length,
                                       size_t num_values) {
     _dict = std::move(dict);
+    if (_dict == nullptr) {
+        return Status::Corruption("Wrong dictionary data for byte array type, dict is null.");
+    }
     _dict_items.reserve(num_values);
     uint32_t offset_cursor = 0;
     char* dict_item_address = reinterpret_cast<char*>(_dict.get());

diff --git a/be/src/vec/exec/format/parquet/fix_length_dict_decoder.hpp b/be/src/vec/exec/format/parquet/fix_length_dict_decoder.hpp
@@ -107,6 +107,9 @@ class FixLengthDictDecoder final : public BaseDictDecoder {
             return Status::Corruption("Wrong dictionary data for fixed length type");
         }
         _dict = std::move(dict);
+        if (_dict == nullptr) {
+            return Status::Corruption("Wrong dictionary data for byte array type, dict is null.");
+        }
         char* dict_item_address = reinterpret_cast<char*>(_dict.get());
         _dict_items.resize(num_values);
         _dict_value_to_code.reserve(num_values);