10.5.1

Bugfixes

• Fix error with ident.* syntax #430
• Add missing || operator to default SQL dialect #433

10.5.0

Enhancements

• Add language: transactsql as alias for tsql #427

10.4.0

Bugfixes

• Fix parse error in BETWEEN expressions that use tbl.col syntax #423
• Don't support backslash escapes when SQL dialect only supports repeated-quote escapes #308

Mildly breaking changes

• All strings and identifiers no more support all kinds of escaping (e.g. both \" and "") #425
• Unclosed strings/quoted-identifiers (like "...) are no more supported #425
• All possible single-char operators are no more supported in all dialects (only operators specific to the dialect) #426

10.3.0

Improvements

• Support nested block comments in PostgreSQL and Transact-SQL #422 #411

Mildly breaking changes

• Unclosed block comments (like /* ...) will now produce an error

10.2.0

Bugfixes

• Fix formatting of TIMESTAMP WITH TIME ZONE in PostgreSQL #391
• Fix formatting of FOR UPDATE clause in PL/SQL #340

Internal changes

• Reduce coupling between Tokenizer and TokenizerEngine #416

10.1.1

Bugfixes

• Fix crash caused by ROWS BETWEEN in window definitions #417 #415

10.1.0

Improvements

• Add support for @@variables in MariaDB, MySQL, SingleStoreDB #410
• Remove support of prefixes from double-quoted strings inMariaDB, MySQL, SingleStoreDB (shouldn't break unless you're for some reason formatting invalid code) #408

Bugfixes

• Fix tabular style formatting of LIMIT clause #412

Internal changes

• Remove unnecessary type imports, improving compatibility with older TypeScript versions #404
• Remove Tokens from AST #403

10.0.0

Breaking changes

• Encountering braces {} or brackets [] in SQL dialect not supporting these will now throw error #400
• Encountering unbalanced parenthesis () (also {} and []) will now produce an error #277

Performance

• Tokenizer is now faster, thanks to removal of case-insensitive regex transform #394
• Parser is now slower, because of the switch to nearley #277
• Overall performance is more impacted by the parser, so the net result is a slower formatting. Though this should only affect you if you're formatting megabytes worth of SQL.

Internal changes

• Switch from hand-written parser to Nearley parser-generator #277
• Added more token types #398 #395
• Replace open- and closeParens configs with extraParens #405

10.0.0-beta.1

Breaking changes

• Encountering braces {} or brackets [] in SQL dialect not supporting these will now throw error #400
• Encountering unbalanced parenthesis () (also {} and []) will now produce an error #277

Performance

• Tokenizer is now faster, thanks to removal of case-insensitive regex transform #394
• Parser is now slower, because of the switch to nearley #277
• Overall performance is more impacted by the parser, so the net result is a slower formatting. Though this should only effect you if you're formatting megabytes worth of SQL.

Internal changes

• Switch from hand-written parser to Nearley parser-generator #277
• Added more token types #398 #395

9.2.0

New features:

• Added paramTypes config option to allow user-configured parameter placeholder types #367
• Extended language option to allow for passing in a user-defined Formatter class #306

Improvements:

• Better handling of line-comments, preserving their position either at the end of line or between lines #365