Support cast decimal data type with different precision and sale (#839)

Author: harveyyue

native-engine/datafusion-ext-plans/src/scan/mod.rs

@@ -171,7 +171,9 @@ fn schema_adapter_cast_column(col: &ArrayRef, data_type: &DataType) -> Result<Ar
             DataType::Int16 => handle_decimal!(Int16, Decimal128, i128, *prec, *scale),
             DataType::Int32 => handle_decimal!(Int32, Decimal128, i128, *prec, *scale),
             DataType::Int64 => handle_decimal!(Int64, Decimal128, i128, *prec, *scale),
-            DataType::Decimal128(p, s) if p == prec && s == scale => Ok(col.clone()),
+            DataType::Decimal128(p, s) => {
+                datafusion_ext_commons::arrow::cast::cast_scan_input_array(col.as_ref(), data_type)
+            }
             _ => df_execution_err!(
                 "schema_adapter_cast_column unsupported type: {:?} => {:?}",
                 col.data_type(),