Merge pull request apache#76 from zhouyuan/wip_optional_dictencode_da…

…taset [C++] allow to optionally disable dict encoding
xuechendi · Jun 19, 2020 · 0b4df23 · 0b4df23
2 parents f8b201f + 7d75580
commit 0b4df23
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/cpp/src/arrow/dataset/file_parquet.cc b/cpp/src/arrow/dataset/file_parquet.cc
@@ -104,7 +104,7 @@ static parquet::ReaderProperties MakeReaderProperties(
 void SetDictionaryColumns(const parquet::ParquetFileReader &reader,
                           parquet::ArrowReaderProperties &properties,
                           const std::unordered_set<std::string> &dict_columns) {
-  if (dict_columns.empty()) {
+  if (dict_columns.empty() && std::getenv("FORCE_ENCODE")) {
     // default: dict-encode all columns
     int num_columns = reader.metadata()->num_columns();
     for (int i = 0; i < num_columns; i++) {