Highlights
This minor release contains four particular changes of note:
- The default
render_variant_limitis now5instead of1, so SQLFluff may report new linting violations from templated branches that were previously not inspected in Jinja and dbt projects. Documentation for the feature is now also available in Template Variant Rendering. - This release drops support for Python 3.9, which reached end of life at the end of October 2025.
- Security improvements that protect against resource exhaustion through malicious queries by limiting total parsed nodes. Users can configure the new
max_parse_nodesconfig setting to enable parsing of larger files in their project if necessary. - A new
AL10rule requires aliases onFROMsubqueries, because omitting them causes parse errors in most major dialects.
Beyond that, there are parser improvements for T-SQL, PostgreSQL, Snowflake, BigQuery, DuckDB, ClickHouse, Oracle, Hive/SparkSQL, Databricks, DB2, Athena, Trino, MariaDB/MySQL, StarRocks, Teradata, and Greenplum. There are also rule fixes for ST06, ST11, LT02, and LT09, better handling for placeholder and dbt/Jinja rendering edge cases, and a new option to fail when files are skipped for size.
This release also includes first-time contributions from twenty new contributors. Thank you all for your contributions. 🏆
What’s Changed
- fix(postgres): support CORRELATION in CREATE STATISTICS and IF EXISTS in ALTER STATISTICS (#7831) @jonasboos
- TSQL: allow set expressions in
DECLARE ... CURSOR FOR(#7812) @copilot-swe-agent[bot] - fix: Ensure
pool.join()is called in ParallelRunner (#7686) @peterbud - fix(clickhouse): allow WHERE/PREWHERE after ARRAY JOIN (#7837) @Cayan
- fix(postgres): parse \crosstabview as query buffer terminator (#7833) @SAY-5
- Add max_parse_nodes parser limit (#7816) @alanmcruickshank
- fix(athena): allow START as identifier (not reserved for SELECT) (#7834) @SAY-5
- TSQL: parse
ALTER TABLE ... ENABLE|DISABLE TRIGGER(#7811) @copilot-swe-agent[bot] - Support Teradata CAST character set phrases (#7766) @copilot-swe-agent[bot]
- Oracle: add support for MULTISET operators (#7711) @saulotoledo
- TSQL: Parse issues DatatypeMethodSeg in TableRef and TableExp for Merge using (#7802) @dpurfield
- Enable multi-branch linting by default (#7808) @alanmcruickshank
- fix(tsql): parse SET DATEFORMAT canonical Microsoft syntax (#7819) @Booyaka101
- add regression test for issue 7543 (#7797) @rotempasharel1
- fix: prevent placeholder templater from suppressing lint violations (#7818) @ben-duivenvoorden
- TSQL: Add CONTAINSTABLE support (#7814) @peterbud
- fix(duckdb): tokenize != as a single token to avoid factorial collision (#7742) @barry3406
- fix(tsql): allow Unicode letters in @parameter names (#7810) @Booyaka101
- Drop support for python 3.9 (#7341) @annebelleo
- Add DB2 support for trailing read-only and isolation clauses on SELECT (#7768) @copilot-swe-agent[bot]
- Handle shorthand-casted aggregates correctly in ST06 column ordering (#7792) @copilot-swe-agent[bot]
- Hive: parse SET configs and
${...}parameter references (#7767) @copilot-swe-agent[bot] - Preserve mocked dbt
var()placeholders under Jinja subscripting (#7793) @copilot-swe-agent[bot] - fix: ST11 false positive on struct/nested field access (#7795) @sebastien-irco
- Snowflake: support CREATE OR ALTER for FILE FORMAT, FUNCTION, PROCEDURE, EXTERNAL FUNCTION, AUTHENTICATION POLICY, TAG (#7799) @stevebeck89
- Reject trailing commas after the final CTE in
WITHclauses (#7761) @copilot-swe-agent[bot] - Support colon-delimited
json_object()pairs in Trino and Athena (#7764) @copilot-swe-agent[bot] - feat: Add BigLake syntax support for BigQuery dialect (#7794) @sebastien-irco
- Fix typo: remove duplicate 'the' in documentation (#7782) @Jah-yee
- test: add LT02 autofix fixtures for MERGE INSERT VALUES indentation (ansi, bigquery, sparksql) (#7791) @Scolliq
- Rust: Fix sequence terminators (#7787) @keraion
- fix: Support compound statements in BigQuery WHILE body (#7789) @sebastien-irco
- feat: Support function calls in MERGE USING clause (#7788) @sebastien-irco
- build(deps): bump rustls-webpki from 0.103.10 to 0.103.13 in /sqlfluffrs (#7790) @dependabot[bot]
- fix(7784): create table clone statement grammar for databricks dialect (#7786) @pgrz
- Relax ignore spec typing in discovery module (#7785) @peterbud
- BigQuery: parse
OPTIONSinCREATE TABLE FUNCTION(#7776) @copilot-swe-agent[bot] - Allow MariaDB/MySQL
FROM DUALto parse as a table reference (#7763) @copilot-swe-agent[bot] - Support lateral column alias expressions in ClickHouse, Databricks, and Redshift (#7760) @copilot-swe-agent[bot]
- fix(tsql): keep required OPENROWSET BULK aliases (#7781) @cloudyun888
- feat: Add CREATE AGGREGATE FUNCTION support for BigQuery dialect (#7780) @sebastien-irco
- fix(postgres): preserve begin atomic body indentation (#7777) @cloudyun888
- chore: make pytest as optional dependency (#7275) @fatelei
- feat: Add option for non-zero exit on large_file_skip (#7731) @zozo123
- fix: support trailing GROUPING SETS in Hive GROUP BY (#7651) @saschabuehrle
- Snowflake: parse
WITH TAGon dynamic table columns without a data type (#7754) @copilot-swe-agent[bot] - Support
COLLATEin Databricks CREATE TABLE column properties (#7749) @copilot-swe-agent[bot] - Fix unparseable Databricks/SparkSQL function parameters with function call DEFAULT values (#7721) @copilot-swe-agent[bot]
- Support
CHECK/NOCHECK CONSTRAINT ALLin T-SQLALTER TABLE(#7748) @copilot-swe-agent[bot] - fix(databricks): RF03 should not flag STRUCT function aliases (#7735) @zozo123
- fix(LT09): Consistent single/multi target newline behavior (#7729) @zozo123
- Snowflake: Support
UPDATE ALL BY NAMEsyntax forMERGE(#7756) @ninazacharia-toast - CI: add dhi.io login to workflow (#7757) @keraion
- feat: move to dhi image (#7752) @hoodadt
- fix: LT02 false positive when Jinja block contains same-line parentheses (#7744) @galsakuri
- build(deps): bump rand from 0.9.2 to 0.9.4 in /sqlfluffrs (#7750) @dependabot[bot]
- Add aliasing.required rule AL10: derived tables must have an alias (#7743) @cherylschaefer
- Snowflake: handle ADD SCORING PROFILE in CORTEX SEARCH + fix typo (#7647) @jakadam2
- refactor(oracle): Move keywords to separate file (#7726) @zozo123
- fix(databricks): Allow _metadata column references (#7732) @zozo123
- fix(LT02): Don't force FROM/VALUES separation when values are inline (#7738) @zozo123
- fix(sparksql): Add regression tests for struct field access alignment (#6913) (#7728) @zozo123
- fix: Apply warning-level fixes even when no errors present (#7727) @zozo123
- MySQL: Add support for role GRANT and SET DEFAULT ROLE statements (#7720) @copilot-swe-agent[bot]
- fix(greenplum): Add support for CREATE/DROP EXTERNAL TABLE (#7725) @zozo123
- fix(snowflake): Exclude NEXTVAL/CURRVAL sequences from references.from checks (#7724) @zozo123
- feat(starrocks): Add support for ARRAY, MAP, and STRUCT types (#7717) @zozo123
- fix(snowflake): Support multiple statements in EXCEPTION WHEN handlers (#7718) @zozo123
- TSQL: Add support for ALTER DATABASE SCOPED CONFIGURATION statements (#7688) @peterbud
- Support dotted option names in Postgres ALTER VIEW / ALTER MATERIALIZED VIEW (#7680) @copilot-swe-agent[bot]
- Snowflake: parse structured OBJECT types in CAST and datatype contexts (#7681) @copilot-swe-agent[bot]
- Allow SQLite FTS5 option assignments in CREATE VIRTUAL TABLE (#7683) @copilot-swe-agent[bot]
- TSQL: use aggregate_order_by in WITHIN GROUP (#7687) @peterbud
- docs: Enhance rule generation (#7707) @peterbud
- T-SQL: parse ALTER / DROP FULLTEXT INDEX and improve CREATE (#7712) @dreik
- fix: align following line position handling (#7702) @mauro-lanza
- Oracle: prevent incorrect fix for bind and substitution variables (#7709) @saulotoledo
- Oracle: add support for DROP USER CASCADE / DROP PROFILE / DROP CLUSTER (#7706) @saulotoledo
- Oracle: add support for no-argument procedure calls (#7701) @saulotoledo
- Support macros-calling-macros in jinja templater (#7685) @mrichards42
- fix: add --no-warn-unused-configs flag to mypyc commands in tox.ini (#7698) @peterbud
- StarRocks: add LOAD LABEL and INSERT OVERWRITE WITH LABEL support (#7682) @andystenhe
- [codex] Fix DuckDB double-quoted struct keys (#7694) @zozo123
- fix(ST05): don't descend into subqueries when checking for value table functions (#7692) @nadjichalal
- fix: ST10 RHS constant expression redundancy check (#7690) @aayushr7
- Add dbt 1.11 function() builtin to Jinja apply_dbt_builtins (#7637) @ReinerBRO
- Add table_options to ALTER TABLE in mysql & mariadb (#7411) @WittierDinosaur
New Contributors
- @nadjichalal made their first contribution in #7692
- @zozo123 made their first contribution in #7694
- @mauro-lanza made their first contribution in #7702
- @jakadam2 made their first contribution in #7647
- @cherylschaefer made their first contribution in #7743
- @galsakuri made their first contribution in #7744
- @hoodadt made their first contribution in #7752
- @saschabuehrle made their first contribution in #7651
- @cloudyun888 made their first contribution in #7777
- @pgrz made their first contribution in #7786
- @Scolliq made their first contribution in #7791
- @Jah-yee made their first contribution in #7782
- @stevebeck89 made their first contribution in #7799
- @Booyaka101 made their first contribution in #7810
- @barry3406 made their first contribution in #7742
- @ben-duivenvoorden made their first contribution in #7818
- @rotempasharel1 made their first contribution in #7797
- @SAY-5 made their first contribution in #7834
- @Cayan made their first contribution in #7837
- @jonasboos made their first contribution in #7831