Allow for the overriding of stringify_dict for json export format o…

…n BaseSQLToGCSOperator (#26277)
apache · Sep 18, 2022 · b4f8a06 · b4f8a06
1 parent 706a618
commit b4f8a06
Showing 1 changed file with 5 additions and 1 deletion.
diff --git a/airflow/providers/google/cloud/transfers/sql_to_gcs.py b/airflow/providers/google/cloud/transfers/sql_to_gcs.py
@@ -53,6 +53,8 @@ class BaseSQLToGCSOperator(BaseOperator):
  file size of the splits. Check https://cloud.google.com/storage/quotas
  to see the maximum allowed file size for a single object.
  :param export_format: Desired format of files to be exported. (json, csv or parquet)
+ :param stringify_dict: Whether to dump Dictionary type objects
+ (such as JSON columns) as a string. Applies only to JSON export format.
  :param field_delimiter: The delimiter to be used for CSV files.
  :param null_marker: The null marker to be used for CSV files.
  :param gzip: Option to compress file for upload (does not apply to schemas).
@@ -99,6 +101,7 @@ def __init__(
  schema_filename: str | None = None,
  approx_max_file_size_bytes: int = 1900000000,
  export_format: str = 'json',
+ stringify_dict: bool = False,
  field_delimiter: str = ',',
  null_marker: str | None = None,
  gzip: bool = False,
@@ -121,6 +124,7 @@ def __init__(
  self.schema_filename = schema_filename
  self.approx_max_file_size_bytes = approx_max_file_size_bytes
  self.export_format = export_format.lower()
+ self.stringify_dict = stringify_dict
  self.field_delimiter = field_delimiter
  self.null_marker = null_marker
  self.gzip = gzip
@@ -243,7 +247,7 @@ def _write_local_data_files(self, cursor):
  tbl = pa.Table.from_pydict(row_pydic, parquet_schema)
  parquet_writer.write_table(tbl)
  else:
- row = self.convert_types(schema, col_type_dict, row, stringify_dict=False)
+ row = self.convert_types(schema, col_type_dict, row, stringify_dict=self.stringify_dict)
  row_dict = dict(zip(schema, row))
 
  tmp_file_handle.write(