This VS Code extension adds basic Language Server Protocol (LSP) features for working with SQL code.
It is fast and lightweight, works on plain .sql files, and does not require a database connection.
It uses a regex-based indexer instead of a full SQL parser — so results won’t be perfect in every scenario, but it provides a solid, practical foundation for SQL development in VS Code.
✨ Features
Go to Definition
Jump to the definition of a table, type, or column in your workspace.
Works with both plain names (Employee) and aliases (e.EmployeeId).
Find All References
Locate table and column references across SQL files.
Skips matches in comments and unrelated identifiers to reduce false positives.
IntelliSense / Auto-completion
Suggests table names and types.
After typing alias. or TableName., column suggestions appear.
Schema prefixes are handled (dbo.TableName is treated the same as TableName).
Workspace Indexing
Automatically indexes all .sql files in the workspace.
Updates instantly as you type or save.
🚀 Getting Started
Install the extension.
Open a folder or workspace containing .sql files.
Start editing — features like definitions, references, and completions activate automatically.
💡 Works best when your schema objects (tables, types, procedures) are defined in .sql files within your workspace.
⚠️ Preview Notice
This is a Preview release.
Optimized for T-SQL / SQL Server style DDL & DML.
Dialects like Postgres or MySQL may partially work but are not fully supported yet.
Column and reference detection is heuristic — complex scripts may produce misses or false positives.
We’re releasing early to gather real-world feedback before expanding the feature set.
🛠 Planned Improvements
Outline view for procedures, tables, and columns.
Workspace symbol search (Ctrl+T).
Diagnostics for undefined tables/columns.
Hover information and quick documentation.
Smarter reference resolution (scoped by FROM / JOIN context).
🧑💻 Contributing
We welcome feedback and bug reports!
Open issues with examples of SQL code that doesn’t behave as expected.
Pull requests are encouraged — especially for dialect support or smarter parsing.
Known Limitations
This extension is intentionally lightweight and does not do full SQL parsing or semantic analysis.
Be aware of these trade-offs:
Column References
Column references are matched globally by name.
If multiple tables have the same column (e.g. Id), all may appear in references.
(Teams using explicit naming standards like EmployeeId, DepartmentId will be less affected.)
Bare Columns
Bare columns (SELECT EmployeeId) are treated as global — not resolved to a specific table.
Aliased usage (e.EmployeeId) works better.
Schemas
By default, dbo.TableName and TableName are treated as the same.
Multiple schemas with the same table name are not yet distinguished.
Cross-File Consistency
Accuracy depends on having all schema files in your workspace.
Missing files = missing definitions.
(For SSDT projects, this typically isn’t an issue since all objects are included.)
