SQL LSP Setup

This command will configure your SQL development environment with sql-language-server LSP and sqlfluff for linting and formatting.

Prerequisites Check

Verify you have Node.js and Python installed:

node --version
python --version || python3 --version

Installation Steps

1. Install sql-language-server (LSP)

npm install -g sql-language-server

Verify installation:

sql-language-server --version

2. Install sqlfluff (Linter & Formatter)

pip install sqlfluff

Or with pipx for isolated installation:

pipx install sqlfluff

Verify installation:

sqlfluff --version

3. Create Project Configuration (Optional)

Create a .sqlfluff configuration file in your project root:

cat > .sqlfluff << 'EOF'
[sqlfluff]
# Set your SQL dialect
dialect = postgres
# Options: postgres, mysql, sqlite, bigquery, redshift, snowflake, tsql, etc.

[sqlfluff:rules]
max_line_length = 120
indent_unit = space
tab_space_size = 4

[sqlfluff:rules:L010]
# Keywords should be upper case
capitalisation_policy = upper

[sqlfluff:rules:L030]
# Function names should be upper case
capitalisation_policy = upper

[sqlfluff:rules:L014]
# Unquoted identifiers should be lower case
capitalisation_policy = lower
EOF

4. Verify LSP Configuration

Check that .lsp.json exists and is properly configured:

cat .lsp.json

Expected configuration:

{
    "sql": {
        "command": "sql-language-server",
        "args": ["up", "--method", "stdio"],
        "extensionToLanguage": {
            ".sql": "sql"
        },
        "transport": "stdio"
    }
}

5. Test sqlfluff

Test sqlfluff on a sample SQL file:

# Create test file
cat > test.sql << 'EOF'
select * from users where id = 1;
EOF

# Lint the file
sqlfluff lint test.sql

# Format the file
sqlfluff format test.sql

# Clean up
rm test.sql

Supported SQL Dialects

sqlfluff supports multiple SQL dialects. Configure in .sqlfluff:

• postgres - PostgreSQL
• mysql - MySQL/MariaDB
• sqlite - SQLite
• bigquery - Google BigQuery
• redshift - Amazon Redshift
• snowflake - Snowflake
• tsql - Microsoft SQL Server / T-SQL
• oracle - Oracle SQL
• db2 - IBM DB2
• sparksql - Apache Spark SQL
• hive - Apache Hive
• clickhouse - ClickHouse
• athena - AWS Athena

Quick Install

# Install both tools
npm install -g sql-language-server
pip install sqlfluff

Troubleshooting

sql-language-server not found

1. Ensure Node.js is installed: node --version
2. Check npm global path: npm config get prefix
3. Ensure global npm binaries are in PATH

sqlfluff not found

1. Ensure Python is installed: python --version
2. Check pip installation: pip show sqlfluff
3. Try: python -m pip install sqlfluff

LSP not activating

1. Ensure .sql files exist in your project
2. Restart Claude Code after installation
3. Check LSP logs for errors

Linting/formatting issues

1. Verify dialect in .sqlfluff matches your SQL flavor
2. Run manually: sqlfluff lint yourfile.sql
3. Check for syntax errors in SQL

Verification

After setup, verify everything works:

# Check installations
sql-language-server --version
sqlfluff --version

# Verify LSP config
cat .lsp.json

# Test sqlfluff
echo "SELECT * FROM users;" | sqlfluff lint -

All commands should complete successfully without errors.