pypi sqlfluff 4.4.0
[4.4.0] - 2026-10-02
on Python PyPI

24 days ago

Highlights

This minor release introduces SQLMesh integration, new linting and formatting options, and a refreshed documentation site, alongside a broad set of dialect improvements and bug fixes.

  • SQLMesh projects can now use SQLFluff through the new sqlfluff-templater-sqlmesh plugin. The optional templater brings SQLFluff linting to SQLMesh models, with support for model definitions and macro rendering.
  • Two recent rules offer additional opt-in conventions. CV13 encourages a pure SELECT * FROM <cte> passthrough as the final query in a CTE-based model, following the dbt convention. TQ04 prefers expression AS alias over T-SQL's alias = expression syntax. Both are currently disabled by default and can be enabled with force_enable = True.
  • The CLI gains GitLab Code Quality output through --format gitlab, allowing findings to appear directly in GitLab merge requests. This joins the existing SARIF and GitHub annotation formats for integrating SQLFluff into CI.
  • The Jinja templater now supports custom delimiters, including the <% ... %> syntax used by Snowflake CLI. LT15 also gains an optional minimum number of blank lines between statements.
  • docs.sqlfluff.com has migrated to a new documentation format. The site is more responsive, easier to search, and supports dark mode, with a design that fits the rest of the SQLFluff project.

Beyond that, this release expands parsing support across Snowflake, MariaDB/MySQL, ClickHouse, PostgreSQL, Oracle, Databricks, Flink, Athena, SQLite, Trino, and other dialects. Rule and templater fixes improve reliability, reduce false positives, and prevent unsafe autofixes.

This release also includes first-time contributions from forty-three new contributors. Thank you all for your contributions. 🏆

What’s Changed

  • Make stable the default docs landing version (#8631) @alanmcruickshank
  • build(deps): bump github/issue-labeler from 3.4 to 3.5 in the github-actions group (#8618) @dependabot[bot]
  • fix(AM02): preserve UNION's existing case instead of forcing a hardcoded one (#8534) @udsy19
  • Redshift: parse array accessors in SUPER paths in the FROM clause (#8526) @adhavan18
  • Prepare docs URL parity and production indexing (#8629) @alanmcruickshank
  • ci: fix mypy type annotation for "subclass" (#8621) @keraion
  • docs: refine historical versions index (#8609) @alanmcruickshank
  • Databricks: support Unity Catalog volume privileges in GRANT and REVOKE (#8512) @taslater, @claude
  • fix(mysql): parse DO statements (#8582) @rfanth, @claude
  • Oracle: allow DELETE without the optional FROM keyword (#8576) @MrDKOz, @cubic-dev-ai[bot]
  • StarRocks: support PARTITION BY LIST (#8555) (#8595) @mishra-prince
  • feat(clickhouse): add support for projections (#8585) @AndrewKurg
  • Hive/Impala: support the NULL-safe equals operator <=> (#8575) @remilepriol, @claude
  • fix(RF01): allow Trino ROW access for single table targets (#8475) @dennisimoo
  • Trino: support SHOW, DESCRIBE and EXPLAIN ANALYZE statements (#8540) @adhavan18
  • Snowflake: Support ROW_TIMESTAMP_DEFAULT and trailing commas in table column definitions (#8472) @itzzdev09
  • Postgres: support SET ACCESS METHOD on tables and materialized views (#8436) @Mynk11
  • Apply an inner .sqlfluffignore below its own directory on a relative path (#8429) @devYRPauli
  • Snowflake: ALTER USER policy attachment (AUTHENTICATION/PASSWORD/SESSION POLICY) and WORKLOAD_IDENTITY (#8367) @cotsupa, @claude
  • Snowflake: inherited grants (GRANT/REVOKE INHERITED), CREATE SEMANTIC VIEW and CREATE DATABASE ROLE privileges (#8366) @cotsupa, @claude
  • Trino: accept the DISTINCT qualifier on INTERSECT and EXCEPT (#8532) @maksimtech, Claude Opus 5
  • fix(cli): respect output_line_length in the rules listing (#8529) @ddy314
  • ClickHouse: accept the DISTINCT qualifier on INTERSECT and EXCEPT (#8533) @maksimtech, Claude Opus 5
  • Placeholder: ignore embedded colons in colon_optional_quotes (#8566) @devYRPauli, @claude
  • ST04: keep the space before the outer END when flattening (#8568) @devYRPauli, @claude
  • fix(snowflake): stop scripting IF terminators leaking into CASE expressions (#8468) @michal-swiatowy, @claude
  • Oracle: parse the SQL*Plus SHOW command (#8484) @itzzdev09, @claude
  • feat(clickhouse): add support for underscores in numeric literals and for hex and binary numeric literals (#8537) @AndrewKurg
  • feat(clickhouse): add support for IS NOT DISTINCT FROM operator (<=>) (#8528) @AndrewKurg
  • Databricks: require SEQUENCE BY in a CREATE FLOW REPLACE USING spec (#8460) @devYRPauli, @claude
  • fix: raise a clear error instead of crashing on a non-table rules value in pyproject.toml (#8493) @udsy19
  • fix(mysql): support multiple variables and expression defaults in DEC… (#8538) @rfanth, @claude
  • fix(mysql): allow CALL without parentheses (#8539) @rfanth, @claude
  • docs(parser): drop the stale tf arg from RustParser.parse's docstring (#8530) @simpleqt
  • Databricks: support MANAGED LOCATION on CREATE CATALOG (#8511) @taslater, @claude
  • SparkSQL: allow DESCRIBE HISTORY and DESCRIBE DETAIL as relations (#8514) @taslater, @claude
  • Snowflake: add CREATE / ALTER SESSION POLICY statements (#8487) @cotsupa, @claude
  • Support ClickHouse materialized CTE (#8497) @dolik-rce
  • fix(snowflake): parse key=>value USING params in EXECUTE IMMEDIATE FROM (#8518) @udsy19
  • Oracle: support PERCENT in the FETCH clause (#8489) @itzzdev09, @claude
  • SparkSQL: accept DESC as an abbreviation for DESCRIBE HISTORY and DETAIL (#8510) @taslater, @claude
  • Postgres: support COPY options added in PostgreSQL 15-18 (#8488) @itzzdev09, @claude
  • Snowflake: support EXECUTE DBT PROJECT statement (#8504) @davis0178
  • Databricks: allow a magic cell body after a single-line magic directive (#8507) @taslater
  • fix(doris): parse aggregate-type column definitions with DEFAULT/COMMENT/NOT NULL (#8500) @udsy19
  • Databricks: allow a materialized view to declare only expectations (#8508) @taslater, @claude
  • Snowflake: support empty DEFAULT_SECONDARY_ROLES list and ALTER USER SET DEFAULT_SECONDARY_ROLES (#8486) @cotsupa, @claude
  • fix(snowflake): allow arbitrary expressions in EXECUTE IMMEDIATE (#8470) @michal-swiatowy, @claude
  • fix(snowflake): allow CREATE STREAM APPEND_ONLY/SHOW_INITIAL_ROWS in either order (#8469) @michal-swiatowy, @claude
  • Databricks: support PRIVATE streaming tables and CREATE FLOW append flows (#8509) @taslater, @claude
  • fix(config): preserve nested TOML templater context arrays (#8477) @dennisimoo
  • fix(diff-quality): skip linting when no eligible files remain (#8476) @dennisimoo
  • fix: preserve conversions in dotted Python template variables (#8464) @Yukibei
  • Parse redundant brackets around a join used as a join target (#8393) @chiruu12
  • Snowflake: DCM DEFINE gaps (DEFINE PIPE, DEFINE FILE FORMAT, external stages in DEFINE STAGE) (#8365) @cotsupa, @claude
  • fix(CV11): support MySQL-family CONVERT(expr, type) argument order (#8171) (#8449) @eminsk
  • TSQL: support touching table alias column lists (#8451) @be-student
  • fix(ST06): don't reorder select targets when it would displace comments (#8296) @sjh9714
  • fix: map deprecated semicolon_newline onto CV06 multiline_newline (#8295) @dgvj-work
  • Snowflake: Fix SEMANTIC_VIEW bracket spacing (#8459) @jaysobel
  • feat: Support GitLab Code Quality output format (#8372) @sam-mosleh
  • fix(tests): expand XDG config home variable (#8461) @rubytobi
  • Postgres: support MERGE ... RETURNING (PostgreSQL 17) (#8424) @Mynk11
  • Locate the where clause in CV12 instead of assuming it is last (#8431) @devYRPauli
  • fix(oracle): support IF NOT EXISTS on CREATE SEQUENCE (#8441) @Kunal8954, @claude
  • Snowflake: Parse SEMANTIC_VIEW queries (#8447) @jaysobel
  • Snowflake: fix EXCEPTION handler indentation (#8453) @saime428, @claude
  • DuckDB: parse INSTALL and LOAD extension statements (#8438) @devYRPauli
  • AM04: anchor violations on the wildcard, not the whole SELECT statement (#8410) @rbeilvert
  • Don't assert-crash the analyzer on a non-lambda -> operator (#8399) @winklemad
  • fix(docsv): resolve RST migration rendering artifacts (#8340) @alanmcruickshank, Alan Cruickshank, @cursoragent
  • LT15: add minimum_empty_lines_between_statements (#8282) @adhavan18
  • LT01: don't glue sign indicators into an inline comment marker (#8395) @dylanpulver
  • Oracle: add missing constraint modifiers (#8434) @saulotoledo
  • Snowflake: RESOURCE MONITOR gaps (IF NOT EXISTS position, quoted monitor names) (#8370) @cotsupa, @claude
  • Oracle: Add support for IF NOT EXISTS clause in CREATE INDEX statement (#8445) @saulotoledo
  • SQLite: add ATTACH, DETACH, VACUUM, REINDEX and ANALYZE statements (#8427) @Mynk11
  • fix(reflow): anchor duplicate-whitespace fixes on the whitespace removed (#8419) @winklemad
  • ci: pass newVersionNumber through env in the release workflow (#8375) @WAHIB-EL-KHADIRI, @claude
  • feat(docs): permalink redirects and mixed-builder version picker (#8412) @alanmcruickshank, Alan Cruickshank, @claude
  • Redshift: Scoped permissions (#8402) @adam-eljasiak
  • Fix RF05 crash and false positive on an all-underscore identifier (#8422) @devYRPauli
  • Oracle: add column-level USING INDEX support (#8426) @saulotoledo
  • Snowflake: event table WITH-less clauses, external table gaps (DELTA, USING TEMPLATE, row access ON), SHOW/DESC table variants (#8353) @davidfierro, @claude
  • build(deps): bump MatteoGabriele/agentscan-action from 2.4.0 to 2.5.0 in the github-actions group (#8421) @dependabot[bot]
  • Snowflake: adaptive and interactive warehouse actions, name-less ALTER WAREHOUSE, ALTER SHARE SET COMMENT, CREATE OR ALTER SHARE (#8360) @davidfierro, @claude
  • fix(parser): Bracketed no longer matches empty brackets when content is required (#8406) @cotsupa, @claude
  • build(deps): bump the github-actions group with 16 updates (#8405) @dependabot[bot], @rubytobi
  • feat(snowflake): add grammar for alert DDL (#8420) @adimalkar, @claude
  • docs: fix PyPI publishing workflow link (#8413) @cky0331
  • Snowflake: CREATE DATA METRIC FUNCTION, TABLE(...) parameter signatures, MEMOIZABLE and other CREATE FUNCTION options (#8352) @davidfierro, @claude
  • fix(reflow): ignore alignment predecessors from other lines (#8411) @adimalkar, @claude
  • Snowflake: custom incremental dynamic tables (CUSTOM_INCREMENTAL, unbounded CHANGES, BACKFILL FROM, START AT) (#8403) @davidfierro, @claude
  • fix(CV11): preserve the full operand of a shorthand cast (#8392) @winklemad
  • Clickhouse: support CONSTRAINT ... CHECK/ASSUME table constraints (#8388) @pelovett, @cubic-dev-ai[bot]
  • feat: support custom Jinja delimiters for Snowflake CLI (#8277) @JoeJoeflyn
  • fix: Parse MySQL IF statements as ordered compound statements (#8380) @mvanhorn
  • fix(ci): update actions and configure dependabot (#8400) @rubytobi
  • docs(rust): refresh AGENTS.md status and fix three stale symbol names (#8374) @davidfierro, @claude
  • Snowflake: accept the exception name on RAISE (#8386) @nsparks-stars, @claude
  • correct typo: seleect -> select (#8381) @aecoleman
  • Snowflake Scripting: WHILE/LOOP/REPEAT/CASE control flow (#8344) @davidfierro, @claude
  • fix(redshift): parse SELECT EXCLUDE after the full select list (#8270) @dgvj-work
  • fix(CV12): preserve SQLite outer join semantics (#8203) @dev-willbird1936
  • Snowflake: managed MCP servers (CREATE/DROP/SHOW/DESCRIBE MCP SERVER, USAGE grants) (#8369) @cotsupa, @claude
  • Snowflake: governance policies on views and ALTER VIEW gaps (#8342) @davidfierro, @claude
  • Snowflake: mixed GROUPING SETS in GROUP BY, PIVOT aggregate alias, UNPIVOT IN-list aliases (#8351) @davidfierro, @claude
  • Add rule CV13: final SELECT of a CTE model should be SELECT * FROM ... (dbt convention) (#7976) @youdie006, @alanmcruickshank
  • fix(rust): Parser parity across Ref errors, arena-tree panics, regex case folding, and indentation config (#8244) @rubytobi
  • Snowflake: DYNAMIC TABLE column level actions and governance clauses (#8343) @davidfierro, @claude
  • Snowflake: PROCEDURE gaps (TEMP, argument direction, RESTRICTED CALLER, anonymous procedures, CALL INTO) (#8311) @davidfierro, @claude
  • docs: explain how to find layout segment types (#8299) @Himanshuagrawal4, Alan Cruickshank, @claude
  • CV11: leave CONVERT alone on mysql, where the arguments are reversed (#8283) @adhavan18
  • feat(postgres): support GENERATED ALWAYS AS (...) VIRTUAL (#8139) (#8146) @aznikline
  • Add SECOND(S) unreserved keyword to dialect_databricks_keywords.py (#8303) @ridgeamaro-1
  • Snowflake: data governance policies on tables (projection/aggregation/join policy, contacts, COPY TAGS, storage lifecycle) (#8305) @davidfierro, @claude
  • refactor(docsv): replace usage/index stub with sidebar section group (#8337) @alanmcruickshank, Alan Cruickshank, @claude
  • fix(docsv): inline starter config to remove Sphinx cross-dependency (#8338) @alanmcruickshank, Alan Cruickshank, @claude
  • fix(docsv): update stale content across 5 pages (#8339) @alanmcruickshank, Alan Cruickshank, @claude
  • Athena: parse USING EXTERNAL FUNCTION UDF declarations (#8263) @davidpavlovschi
  • docs(vitepress): populate missing sections and add internal API auto-docs (#8320) @alanmcruickshank, Alan Cruickshank, @claude
  • Improve versioned docs publishing: manual republish, pre-upload validation (#8319) @alanmcruickshank, Alan Cruickshank
  • MariaDB: support IF [NOT] EXISTS on CREATE/ALTER/DROP SEQUENCE (#8288) @mrsinham, Julien LEFEVRE
  • MariaDB: support IF EXISTS on standalone DROP INDEX (#8290) @mrsinham, Alan Cruickshank, @alanmcruickshank, Julien LEFEVRE
  • Snowflake: new DYNAMIC TABLE clauses (SCHEDULER, FROZEN WHERE, REFRESH USING, EXECUTE AS USER, iceberg options) (#8308) @davidfierro, @claude
  • Snowflake: MATERIALIZED VIEW gaps (CLUSTER BY, INTERACTIVE, UNSET COMMENT/TAG fixes) (#8307) @davidfierro, @claude
  • MariaDB: support NEXT VALUE FOR / PREVIOUS VALUE FOR sequence value expressions (#8318) @mrsinham, Julien LEFEVRE
  • Dialect/flink improvements (#7333) @kalanyuz
  • feat: Add SQLMesh templater (#7722) @barradasCouto, @claude, @keraion, @WittierDinosaur
  • Snowflake: CREATE/ALTER TABLE gaps (schema evolution, generated columns, constraint actions, search optimization ON) (#8306) @davidfierro, @claude
  • docs(vitepress): rework the version picker and warn on stale docs (#8300) @alanmcruickshank, Alan Cruickshank
  • MariaDB: support PARTITION BY SYSTEM_TIME on system-versioned tables (#8313) @mrsinham, Julien LEFEVRE
  • Snowflake: TASK gaps (identifier valued parameters, ALTER TASK actions, EXECUTE TASK options) (#8310) @davidfierro, @claude
  • Snowflake: STREAM gaps (WITH TAG, AT(STREAM =>), ON EVENT TABLE, CREATE OR ALTER, clone COPY GRANTS) (#8309) @davidfierro, @claude
  • fix(mysql): parse charset and collation names as dedicated segments (#8314) @mrsinham, Julien LEFEVRE
  • Migrate deprecated rules:max_line_length into the core section (#8271) @Eljees
  • MariaDB: support DELETE HISTORY on system-versioned tables (#8312) @mrsinham, Julien LEFEVRE
  • docs(vitepress): adopt the shared SQLFluff design system (#8289) @alanmcruickshank, Alan Cruickshank, @claude
  • fix(clickhouse): a couple of fixes for the CREATE DICTIONARY statement (#8298) @AndrewKurg
  • MariaDB: support IF EXISTS on ALTER TABLE (table level and DROP PARTITION) (#8287) @mrsinham, Julien LEFEVRE
  • MariaDB: support IF NOT EXISTS on CREATE VIEW (#8291) @mrsinham, Julien LEFEVRE
  • MariaDB: support temporal tables (system-versioned and application-time periods) (#8293) @mrsinham, Julien LEFEVRE

New Contributors

Check out latest releases or
releases around sqlfluff 4.4.0

Don't miss a new sqlfluff release

NewReleases is sending notifications on new releases.

Get notifications