Merge pull request #29123 from depressed-pho/csv-custom-null

docs/en/operations/settings/settings.md

@@ -2833,6 +2833,43 @@ Possible values:
 
 Default value: `1`.
 
+## output_format_csv_null_representation {#output_format_csv_null_representation}
+
+Defines the representation of `NULL` for [CSV](../../interfaces/formats.md#csv) output format. User can set any string as a value, for example, `My NULL`.
+
+Default value: `\N`.
+
+**Examples**
+
+Query
+
+```sql
+SELECT * from csv_custom_null FORMAT CSV;
+```
+
+Result
+
+```text
+788
+\N
+\N
+```
+
+Query
+
+```sql
+SET output_format_csv_null_representation = 'My NULL';
+SELECT * FROM csv_custom_null FORMAT CSV;
+```
+
+Result
+
+```text
+788
+My NULL
+My NULL
+```
+
 ## output_format_tsv_null_representation {#output_format_tsv_null_representation}
 
 Defines the representation of `NULL` for [TSV](../../interfaces/formats.md#tabseparated) output format. User can set any string as a value, for example, `My NULL`.

src/Core/Settings.h

@@ -577,6 +577,7 @@ class IColumn;
     M(String, output_format_avro_codec, "", "Compression codec used for output. Possible values: 'null', 'deflate', 'snappy'.", 0) \
     M(UInt64, output_format_avro_sync_interval, 16 * 1024, "Sync interval in bytes.", 0) \
     M(Bool, output_format_tsv_crlf_end_of_line, false, "If it is set true, end of line in TSV format will be \\r\\n instead of \\n.", 0) \
+    M(String, output_format_csv_null_representation, "\\N", "Custom NULL representation in CSV format", 0) \
     M(String, output_format_tsv_null_representation, "\\N", "Custom NULL representation in TSV format", 0) \
     M(Bool, output_format_decimal_trailing_zeros, false, "Output trailing zeros when printing Decimal values. E.g. 1.230000 instead of 1.23.", 0) \
     \

src/DataTypes/Serializations/SerializationNullable.cpp

@@ -327,7 +327,7 @@ void SerializationNullable::serializeTextCSV(const IColumn & column, size_t row_
     const ColumnNullable & col = assert_cast<const ColumnNullable &>(column);
 
     if (col.isNullAt(row_num))
-        writeCString("\\N", ostr);
+        writeString(settings.csv.null_representation, ostr);
     else
         nested->serializeTextCSV(col.getNestedColumn(), row_num, ostr, settings);
 }

src/Formats/FormatFactory.cpp

@@ -60,6 +60,7 @@ FormatSettings getFormatSettings(ContextPtr context, const Settings & settings)
     format_settings.csv.delimiter = settings.format_csv_delimiter;
     format_settings.csv.empty_as_default = settings.input_format_defaults_for_omitted_fields;
     format_settings.csv.input_format_enum_as_number = settings.input_format_csv_enum_as_number;
+    format_settings.csv.null_representation = settings.output_format_csv_null_representation;
     format_settings.csv.unquoted_null_literal_as_null = settings.input_format_csv_unquoted_null_literal_as_null;
     format_settings.csv.input_format_arrays_as_nested_csv = settings.input_format_csv_arrays_as_nested_csv;
     format_settings.custom.escaping_rule = settings.format_custom_escaping_rule;

src/Formats/FormatSettings.h

@@ -76,6 +76,7 @@ struct FormatSettings
         bool crlf_end_of_line = false;
         bool input_format_enum_as_number = false;
         bool input_format_arrays_as_nested_csv = false;
+        String null_representation = "\\N";
     } csv;
 
     struct Custom

src/Storages/PostgreSQL/MaterializedPostgreSQLConsumer.cpp

@@ -436,7 +436,7 @@ void MaterializedPostgreSQLConsumer::processReplicationMessage(const char * repl
 
                 if (new_relation_definition)
                 {
-                    current_schema_data.column_identifiers.emplace_back(std::make_tuple(data_type_id, type_modifier));
+                    current_schema_data.column_identifiers.emplace_back(std::make_pair(data_type_id, type_modifier));
                 }
                 else
                 {

tests/queries/0_stateless/02029_output_csv_null_representation.reference

@@ -0,0 +1,4 @@
+# output_format_csv_null_representation should initially be \\N
+"val1",\N,"val3"
+# Changing output_format_csv_null_representation
+"val1",∅,"val3"

tests/queries/0_stateless/02029_output_csv_null_representation.sql

@@ -0,0 +1,16 @@
+DROP TABLE IF EXISTS test_data;
+CREATE TABLE test_data (
+    col1 Nullable(String),
+    col2 Nullable(String),
+    col3 Nullable(String)
+) ENGINE = Memory;
+
+INSERT INTO test_data VALUES ('val1', NULL, 'val3');
+
+SELECT '# output_format_csv_null_representation should initially be \\N';
+SELECT * FROM test_data FORMAT CSV;
+
+SELECT '# Changing output_format_csv_null_representation';
+SET output_format_csv_null_representation = '∅';
+SELECT * FROM test_data FORMAT CSV;
+SET output_format_csv_null_representation = '\\N';