Use the following SELECT INTO JSON syntax to export the query results into a CSV file.

SELECT ... INTO OUTFILE WITH PARAMETERS
'file_URI_JSON "file":
  begin-array ("uri": "OCIFS_URI") | ("uri": "PAR_URI") | ("uri": "NATIVE_URI"),
  [, "dialect": begin-object dialect_option [, dialect_option]... end-object ]
end-object'
  
'begin-object_PAR_JSON "file": 
  begin-array "par": "par_url" [, "prefix": "prefix_value" ] end-array 
  [, "dialect": begin-object dialect_option [, dialect_option]... end-object ]
end-object'
  
'begin-object_RP_JSON "file": 
  begin-array "bucket": "bucket_name" [, "prefix": "prefix_value" [, "namespace: "namespace_name" ] [,"region": "region_name" ] ] end-array 
  [,"dialect": begin-object dialect_option [, dialect_option]... end-object ]
end-object'

dialect_option_key_value:
  {
    {"format": "csv" } 
    | {"field_delimiter": ","} 
    | {"record_delimiter": "\\n"}
    | {"encoding": "utf8mb4"} 
    | {"escape_character": "\\"} 
    | {"quotation_marks": "\""}
    | {"null_value": "\\N"} 
    | {"has_header": {true | false} }
    | {"compression": {"uncompressed" | "snappy" | "gzip" | "brotli" | "zstd" | "lz4"} }
  }

Supported as of MySQL 9.3.1. The following parameters are included in the JSON syntax to export the query results to the Object Storage location represented as a URI.

The following parameters are included in the JSON syntax to export the query results to the Object Storage location represented as a PAR URL.

The following parameters are included in the JSON syntax to export the query results to the Object Storage location based on the assigned resource principal.

The dialect parameters help you to structure the export query results data in the exported CSV file. The dialect parameters are applicable when you export the query results using JSON syntax. These parameters are based on the dialect parameters required to create the external table.

You can use the following dialect parameters to export the query results to CSV files.

The following example uses an OCIFS URI to export query results. It uses default delimiters and specifies a header row in the file.

SELECT * FROM example
INTO OUTFILE WITH PARAMETERS
'{"file":[{
     "uri":"oci://bucket_1@tenancy1/data_files_1/"}], 
     "dialect": 
               {"has_header": true, 
                "format": "csv"}
}';

The following example uses a PAR URL to export query results. It uses default delimiters and specifies a header row in the file.

SELECT * FROM example
INTO OUTFILE WITH PARAMETERS
'{"file":[{
     "par":"https://objectstorage.region1.oraclecloud.com/p/1P4yKjnwXLEPh9Ol0Fbg2r81tvgJfh0XRcCcsr45ggh7uu_t_X302Ea-WdYffl5CL1ON/n/tenancy1/b/usr-bucket/o/", "prefix":"summary/2025"}], 
     "dialect": 
               {"has_header": true, 
                "format": "csv"}
}';

The following example uses resource principals to export query results as a CSV file. No dialect parameters are defined, so default values are used.

SELECT sum(l_quantity) from lineitem limit 1000
INTO OUTFILE WITH PARAMETERS
'{"file":[{"region": "region1", 
           "namespace": "org_namespace", 
           "bucket": "usr-bucket",
           "prefix": "export/summary/"}]
}';

The following example uses resource principals to export query results as a CSV file. A header row is defined in the file.

SELECT sum(l_quantity) from lineitem limit 1000
INTO OUTFILE WITH PARAMETERS
'{"file":[{"region": "region1", 
           "namespace": "org_namespace", 
           "bucket": "usr-bucket",
           "prefix": "export/summary/"}]
           "dialect": {"has_header": true, "format": "csv"}
}';