MariaDB SQL Formatter

Why MariaDB has its own setting

MariaDB began as a MySQL fork and still reads text the same way: identifiers in backticks, strings in single or double quotes, backslash escapes such as 'can\'t', and # line comments. Where the two differ is grammar. MariaDB has had DELETE … RETURNING since 10.0 and INSERT … RETURNING since 10.5, and the MariaDB setting treats RETURNING as a clause. With the MySQL setting the same statement still formats, but returning cart_id, sku is glued onto the end of the WHERE condition and left in lower case.

So choose by server, not by habit. A query written for MariaDB 10.6 or 11.x belongs under MariaDB; one that must also run on MySQL 8 is better checked under MySQL, since that is the stricter target for RETURNING and sequences.

Sequences, temporal tables and JSON

NEXTVAL(invoice_seq), LASTVAL(invoice_seq) and NEXT VALUE FOR invoice_seq format as ordinary expressions inside SELECT and INSERT lists. CREATE SEQUENCE itself is laid out less neatly: START WITH 1000 is split so that WITH begins a new line, as if a CTE followed. The statement is unchanged apart from whitespace, so it still runs.

For system-versioned tables, FOR SYSTEM_TIME AS OF TIMESTAMP '2026-03-01 00:00:00', BETWEEN … AND … and ALL stay on the same line as the table name, which keeps a point-in-time query easy to scan. WITH SYSTEM VERSIONING at the end of a CREATE TABLE lands on a WITH line of its own. Application-time periods are less lucky: in FOR PORTION OF valid_period FROM … TO …, the FROM is mistaken for a new FROM clause.

JSON_TABLE(items, '$[*]' COLUMNS (…)) is expanded with one column definition per line, and INTERSECT and EXCEPT separate their two queries cleanly.

Upserts, procedures and executable comments

MariaDB has no AS new row alias for ON DUPLICATE KEY UPDATE, so upserts refer to the incoming row with VALUES(col) or its synonym VALUE(col). Prefer VALUE(qty) when formatting: the plural form is read as the start of a VALUES clause and gets broken across lines, while the singular stays inline.

Stored routines format best without DELIMITER. With DELIMITER $$, the closing END $$ and the next DELIMITER ; end up on one line, so remove those client commands first. The body statements between BEGIN and END are each laid out, though not indented under BEGIN.

One caveat for Minify: the generic /*! … */ executable comment is preserved, but MariaDB’s own /*M!100500 … */ form is treated as an ordinary comment and removed. Keep such statements in formatted form, or check the minified output before using it.

Useful options

Commas → Leading suits long INSERT column lists and multi-row VALUES. Function case → UPPER makes JSON_VALUE and DATE_FORMAT stand out from column names. To look for unclosed quotes or brackets without changing the layout, open the SQL validator; to squeeze a query onto one line, the SQL minifier does only that.

Examples

insert into customers (email, country, signup_source) values ('ana@example.com', 'PT', 'web'), ('li@example.com', 'SG', 'app') returning id, email, created_at;

INSERT INTO
  customers (email, country, signup_source)
VALUES
  ('ana@example.com', 'PT', 'web')
, ('li@example.com', 'SG', 'app')
RETURNING
  id
, email
, created_at;

select p.sku, p.amount, p.row_start from prices for system_time as of timestamp '2026-03-01 00:00:00' as p where p.sku like 'SKU-1%' order by p.sku;

SELECT
  p.sku,
  p.amount,
  p.row_start
FROM
  prices FOR system_time AS of TIMESTAMP '2026-03-01 00:00:00' AS p
WHERE
  p.sku LIKE 'SKU-1%'
ORDER BY
  p.sku;

select o.id as order_id, jt.sku, jt.qty from orders o, json_table(o.items, '$[*]' columns (sku varchar(32) path '$.sku', qty int path '$.qty')) as jt where o.created_at >= '2026-01-01' and jt.qty > 1;

SELECT
  o.id AS order_id,
  jt.sku,
  jt.qty
FROM
  orders o,
  json_table (
    o.items,
    '$[*]' columns (
      sku VARCHAR(32) path '$.sku',
      qty INT path '$.qty'
    )
  ) AS jt
WHERE
  o.created_at >= '2026-01-01'
  AND jt.qty > 1;

# nightly stock sync
insert into inventory (sku, warehouse_id, qty) values ('SKU-1001', 3, 40), ('SKU-1002', 3, 12) on duplicate key update qty = qty + value(qty), updated_at = now();

# nightly stock sync
INSERT INTO
  inventory (sku, warehouse_id, qty)
VALUES
  ('SKU-1001', 3, 40),
  ('SKU-1002', 3, 12)
ON DUPLICATE KEY UPDATE
  qty = qty + value (qty),
  updated_at = now();

Common errors and how to fix them

Related tools

• SQL formatter
• sql formatter/mysql
• sql formatter/postgresql
• sql validator
• sql minifier
• JSON to SQL
• sql style guide
• sql unbalanced parentheses