✨ Features
- Allow returning >1 statements using macros (#3004) @georgesittas
- Add Databricks Serverless support (#3001) @eakmanrq
- Make MacroEvaluator.columns_to_types more lenient (#3006) @georgesittas
- Automatically load signals from the
signals/subfolder (#3010) @plaflamme - Add support for dbt external location in sources meta (#2999) @Themiscodes
🪲 Bug fixes
- Make lookback conditional on missing intervals closes #2985 (#3002) @tobymao
- Builtin pattern matching audits operators (#3011) @Themiscodes
- Reading boolean columns from CSV files (#3014) @Themiscodes
- Toby/perf (#3008) @tobymao
- Properly handle panda columns scd type 2 (#3009) @eakmanrq
- Exponentially multiplying removed interval records (#3016) @izeigerman
- Slack user warning :: the top level
textargument is missing in the request payload (#3020) @zzztimbo - Snapshot parsing during migration (#3022) @izeigerman
- Quote table names closes #3021 (#3023) @tobymao
- Include models selected for backfill in the virtual layer even if they haven't been changed directly (#3024) @izeigerman
- Serialization regression with sqlmesh.core.dialect.normalize_model_name (#3025) @z3z1ma
🧹 Chore
- Remove len == 0 from snowflake _create_column_comments (#2993) @treysp
- Fix sushi waiters default_catalog capitalization (#2994) @treysp
- Fix docstring description -> project typo in _Node (#2997) @georgesittas
- Mention that python model default kind is FULL in model/overview.md (#2996) @georgesittas
- Remove dead snowflake comment code (#2995) @treysp
- Fix python model pre/post-statement docs example (#3007) @georgesittas
- Improve the data model and performance of the state sync (#3005) @izeigerman
- Set interval end per model in plan (#3013) @izeigerman
- Introduce the snapshot cache to the state sync (#3018) @izeigerman
- Bump sqlglot to v25.14.0 (#3003) @georgesittas
- Bump sqlglot to v25.15.0 (#3026) @georgesittas