Fix Iceberg read optimization returning NULLs for stats-less manifests

src/Storages/ObjectStorage/StorageObjectStorageSource.cpp

@@ -905,40 +905,43 @@ StorageObjectStorageSource::ReaderHolder StorageObjectStorageSource::createReade
                     }
                 }
             }
-            for (const auto & column : requested_columns_list)
+            if (file_meta_data.has_value() && !file_meta_data.value()->columns_info.empty())
             {
-                const auto & column_name = column.first;
+                for (const auto & column : requested_columns_list)
+                {
+                    const auto & column_name = column.first;
 
-                if (file_meta_data.value()->columns_info.contains(column_name))
-                    continue;
+                    if (file_meta_data.value()->columns_info.contains(column_name))
+                        continue;
 
-                if (!column.second.second.type->isNullable())
-                    continue;
+                    if (!column.second.second.type->isNullable())
+                        continue;
 
-                /// With View over Iceberg table we have someting like 'materialize(time)' as column_name
-                /// Simple cheap check
-                if (column_name.starts_with("materialize(") && column_name.ends_with(")"))
-                    continue;
+                    /// With View over Iceberg table we have someting like 'materialize(time)' as column_name
+                    /// Simple cheap check
+                    if (column_name.starts_with("materialize(") && column_name.ends_with(")"))
+                        continue;
 
-                /// Skip columns produced by prewhere or row-level filter expressions —
-                /// they are computed at read time, not stored in the file.
-                if (format_filter_info
-                    && ((format_filter_info->prewhere_info && column_name == format_filter_info->prewhere_info->prewhere_column_name)
-                        || (format_filter_info->row_level_filter && column_name == format_filter_info->row_level_filter->column_name)))
-                    continue;
+                    /// Skip columns produced by prewhere or row-level filter expressions —
+                    /// they are computed at read time, not stored in the file.
+                    if (format_filter_info
+                        && ((format_filter_info->prewhere_info && column_name == format_filter_info->prewhere_info->prewhere_column_name)
+                            || (format_filter_info->row_level_filter && column_name == format_filter_info->row_level_filter->column_name)))
+                        continue;
 
-                /// Column is nullable and absent in file
-                constant_columns_with_values[column.second.first] =
-                    ConstColumnWithValue{
-                        column.second.second,
-                        Field()
-                    };
-                constant_columns.insert(column_name);
-
-                LOG_DEBUG(log, "In file {} constant column '{}' type '{}' with value 'NULL'",
-                    object_info->getPath(),
-                    column_name,
-                    column.second.second.type);
+                    /// Column is nullable and absent in file
+                    constant_columns_with_values[column.second.first] =
+                        ConstColumnWithValue{
+                            column.second.second,
+                            Field()
+                        };
+                    constant_columns.insert(column_name);
+
+                    LOG_DEBUG(log, "In file {} constant column '{}' type '{}' with value 'NULL'",
+                        object_info->getPath(),
+                        column_name,
+                        column.second.second.type);
+                }
             }
         }
 

tests/integration/helpers/iceberg_avro_schemas.py

@@ -0,0 +1,49 @@
+import json
+
+# Iceberg v2 manifest list schema (only the fields ClickHouse reads).
+MANIFEST_LIST_SCHEMA_STR = json.dumps({
+    "type": "record",
+    "name": "manifest_file",
+    "fields": [
+        {"name": "manifest_path",        "type": "string"},
+        {"name": "manifest_length",      "type": "long"},
+        {"name": "partition_spec_id",    "type": "int"},
+        {"name": "content",              "type": "int"},
+        {"name": "sequence_number",      "type": "long"},
+        {"name": "min_sequence_number",  "type": "long"},
+        {"name": "added_snapshot_id",    "type": "long"},
+        {"name": "added_files_count",    "type": "int"},
+        {"name": "existing_files_count", "type": "int"},
+        {"name": "deleted_files_count",  "type": "int"},
+        {"name": "added_rows_count",     "type": "long"},
+        {"name": "existing_rows_count",  "type": "long"},
+        {"name": "deleted_rows_count",   "type": "long"},
+    ],
+})
+
+# Stats-less manifest entry: no optional stats fields -> empty columns_info.
+MANIFEST_ENTRY_NO_STATS_SCHEMA_STR = json.dumps({
+    "type": "record",
+    "name": "manifest_entry",
+    "fields": [
+        {"name": "status",               "type": "int"},
+        {"name": "snapshot_id",          "type": ["null", "long"]},
+        {"name": "sequence_number",      "type": ["null", "long"]},
+        {"name": "file_sequence_number", "type": ["null", "long"]},
+        {
+            "name": "data_file",
+            "type": {
+                "type": "record",
+                "name": "r2",
+                "fields": [
+                    {"name": "content",            "type": "int"},
+                    {"name": "file_path",          "type": "string"},
+                    {"name": "file_format",        "type": "string"},
+                    {"name": "partition", "type": {"type": "record", "name": "r102", "fields": []}},
+                    {"name": "record_count",       "type": "long"},
+                    {"name": "file_size_in_bytes", "type": "long"},
+                ],
+            },
+        },
+    ],
+})