sql-glider 0.1.8__py3-none-any.whl

sql_glider-0.1.8.dist-info/METADATA

@@ -0,0 +1,893 @@
+Metadata-Version: 2.4
+Name: sql-glider
+Version: 0.1.8
+Summary: SQL Utility Toolkit for better understanding, use, and governance of your queries in a native environment.
+Project-URL: Homepage, https://github.com/rycowhi/sql-glider/
+Project-URL: Repository, https://github.com/rycowhi/sql-glider/
+Project-URL: Documentation, https://github.com/rycowhi/sql-glider/
+Project-URL: Issues, https://github.com/rycowhi/sql-glider/issues
+Author-email: Ryan Whitcomb <ryankwhitcomb@gmail.com>
+License-Expression: Apache-2.0
+License-File: LICENSE
+Keywords: data-governance,data-lineage,lineage,sql,sqlglot
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: SQL
+Classifier: Topic :: Database
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Classifier: Typing :: Typed
+Requires-Python: >=3.11
+Requires-Dist: jinja2>=3.0.0
+Requires-Dist: pydantic>=2.0.0
+Requires-Dist: rich>=13.0.0
+Requires-Dist: rustworkx>=0.15.0
+Requires-Dist: sqlglot[rs]>=25.0.0
+Requires-Dist: typer>=0.9.0
+Provides-Extra: databricks
+Requires-Dist: databricks-sdk>=0.20.0; extra == 'databricks'
+Description-Content-Type: text/markdown
+
+# SQL Glider
+
+SQL Utility Toolkit for better understanding, use, and governance of your queries in a native environment.
+
+## Overview
+
+SQL Glider provides powerful column-level and table-level lineage analysis for SQL queries using SQLGlot. It operates on standalone SQL files without requiring a full project setup, making it perfect for ad-hoc analysis, data governance, and understanding query dependencies.
+
+## Features
+
+- **Forward Lineage:** Trace output columns back to their source tables and columns
+- **Reverse Lineage:** Impact analysis - find which output columns are affected by a source column
+- **Query Dissection:** Decompose SQL into components (CTEs, subqueries, UNION branches) for unit testing
+- **Table Extraction:** List all tables in SQL files with usage type (INPUT/OUTPUT) and object type (TABLE/VIEW/CTE)
+- **Multi-level Tracing:** Automatically handles CTEs, subqueries, and complex expressions
+- **Graph-Based Lineage:** Build and query lineage graphs across thousands of SQL files
+- **Multiple Output Formats:** Text (human-readable), JSON (machine-readable), CSV (spreadsheet-ready)
+- **Dialect Support:** Works with Spark, PostgreSQL, Snowflake, BigQuery, MySQL, and many more SQL dialects
+- **File Export:** Save lineage results to files for documentation or further processing
+
+## Installation
+
+SQL Glider is available on PyPI and can be installed with pip or uv. Python 3.11+ is required.
+
+```bash
+# Install with pip
+pip install sql-glider
+
+# Or install with uv
+uv pip install sql-glider
+```
+
+After installation, the `sqlglider` command is available:
+
+```bash
+sqlglider lineage query.sql
+```
+
+### Development Setup
+
+If you want to contribute or run from source:
+
+```bash
+# Clone the repository
+git clone https://github.com/ryanholmdahl/sql-glider.git
+cd sql-glider
+
+# Install dependencies with uv
+uv sync
+
+# Run from source
+uv run sqlglider lineage <sql_file>
+```
+
+## Quick Start
+
+### Forward Lineage (Source Tracing)
+
+Find out where your output columns come from:
+
+```bash
+# Analyze all output columns
+uv run sqlglider lineage query.sql
+
+# Analyze a specific output column
+uv run sqlglider lineage query.sql --column customer_name
+```
+
+**Example Output:**
+```
+       Query 0: SELECT customer_name, o.order_total FROM customers c JOIN orders o ...
++-----------------------------------------------------------------------------+
+| Output Column   | Source Column                                             |
+|-----------------+------------------------------------------------------------|
+| customer_name   | c.customer_name                                           |
++-----------------------------------------------------------------------------+
+Total: 1 row(s)
+```
+
+This shows that the output column `customer_name` in Query 0 comes from `c.customer_name` (the `customer_name` column in table `c`).
+
+### Reverse Lineage (Impact Analysis)
+
+Find out which output columns are affected by a source column:
+
+```bash
+# Find outputs affected by a source column
+uv run sqlglider lineage query.sql --source-column orders.customer_id
+```
+
+**Example Output:**
+```
+          Query 0: SELECT customer_id, segment FROM ...
++---------------------------------------------------------+
+| Output Column      | Source Column                      |
+|--------------------+------------------------------------|
+| orders.customer_id | orders.customer_id                 |
++---------------------------------------------------------+
+Total: 1 row(s)
+```
+
+This shows that if `orders.customer_id` changes, it will impact the output column `customer_id` in Query 0.
+
+## Usage Examples
+
+### Basic Column Lineage
+
+```bash
+# Forward lineage for all columns
+uv run sqlglider lineage query.sql
+
+# Forward lineage for specific column
+uv run sqlglider lineage query.sql --column order_total
+
+# Reverse lineage (impact analysis)
+uv run sqlglider lineage query.sql --source-column orders.customer_id
+```
+
+### Different Output Formats
+
+```bash
+# JSON output
+uv run sqlglider lineage query.sql --output-format json
+
+# CSV output
+uv run sqlglider lineage query.sql --output-format csv
+
+# Export to file
+uv run sqlglider lineage query.sql --output-format json --output-file lineage.json
+```
+
+### Table-Level Lineage
+
+```bash
+# Show which tables are used
+uv run sqlglider lineage query.sql --level table
+```
+
+### Table Extraction
+
+List all tables involved in SQL files with usage and type information:
+
+```bash
+# List all tables in a SQL file
+uv run sqlglider tables overview query.sql
+
+# JSON output with detailed table info
+uv run sqlglider tables overview query.sql --output-format json
+
+# Export to CSV
+uv run sqlglider tables overview query.sql --output-format csv --output-file tables.csv
+```
+
+### Pull DDL from Remote Catalogs
+
+Fetch DDL definitions from remote data catalogs (e.g., Databricks Unity Catalog):
+
+```bash
+# Pull DDL for all tables used in a SQL file (outputs to stdout)
+uv run sqlglider tables pull query.sql --catalog-type databricks
+
+# Save DDL files to a folder (one file per table)
+uv run sqlglider tables pull query.sql -c databricks -o ./ddl/
+
+# List available catalog providers
+uv run sqlglider tables pull --list
+```
+
+**Note:** Requires optional dependencies. Install with: `pip install sql-glider[databricks]`
+
+**Example Output (JSON):**
+```json
+{
+  "queries": [{
+    "query_index": 0,
+    "tables": [
+      {"name": "customers", "usage": "INPUT", "object_type": "UNKNOWN"},
+      {"name": "orders", "usage": "INPUT", "object_type": "UNKNOWN"}
+    ]
+  }]
+}
+```
+
+**Table Usage Types:**
+- `INPUT`: Table is read from (SELECT, JOIN, subqueries)
+- `OUTPUT`: Table is written to (INSERT, CREATE TABLE/VIEW, UPDATE)
+- `BOTH`: Table is both read from and written to
+
+**Object Types:**
+- `TABLE`: CREATE TABLE or DROP TABLE statement
+- `VIEW`: CREATE VIEW or DROP VIEW statement
+- `CTE`: Common Table Expression (WITH clause)
+- `UNKNOWN`: Cannot determine type from SQL alone
+
+### Query Dissection
+
+Decompose SQL queries into constituent parts for unit testing and analysis:
+
+```bash
+# Dissect a SQL file (text output)
+uv run sqlglider dissect query.sql
+
+# JSON output with full component details
+uv run sqlglider dissect query.sql --output-format json
+
+# CSV output for spreadsheet analysis
+uv run sqlglider dissect query.sql --output-format csv
+
+# Export to file
+uv run sqlglider dissect query.sql -f json -o dissected.json
+
+# With templating support
+uv run sqlglider dissect query.sql --templater jinja --var schema=analytics
+
+# From stdin
+echo "WITH cte AS (SELECT id FROM users) SELECT * FROM cte" | uv run sqlglider dissect
+```
+
+**Example Input:**
+```sql
+WITH order_totals AS (
+    SELECT customer_id, SUM(amount) AS total
+    FROM orders
+    GROUP BY customer_id
+)
+INSERT INTO analytics.summary
+SELECT * FROM order_totals WHERE total > 100
+```
+
+**Example Output (JSON):**
+```json
+{
+  "queries": [{
+    "query_index": 0,
+    "statement_type": "INSERT",
+    "total_components": 3,
+    "components": [
+      {
+        "component_type": "CTE",
+        "component_index": 0,
+        "name": "order_totals",
+        "sql": "SELECT customer_id, SUM(amount) AS total FROM orders GROUP BY customer_id",
+        "is_executable": true,
+        "dependencies": [],
+        "location": "WITH clause"
+      },
+      {
+        "component_type": "TARGET_TABLE",
+        "component_index": 1,
+        "name": "analytics.summary",
+        "sql": "analytics.summary",
+        "is_executable": false,
+        "location": "INSERT INTO target"
+      },
+      {
+        "component_type": "SOURCE_QUERY",
+        "component_index": 2,
+        "sql": "SELECT * FROM order_totals WHERE total > 100",
+        "is_executable": true,
+        "dependencies": ["order_totals"],
+        "location": "INSERT source SELECT"
+      }
+    ]
+  }]
+}
+```
+
+**Extracted Component Types:**
+- `CTE`: Common Table Expressions from WITH clause
+- `MAIN_QUERY`: The primary SELECT statement
+- `SUBQUERY`: Nested SELECT in FROM clause
+- `SCALAR_SUBQUERY`: Single-value subquery in SELECT list, WHERE, HAVING
+- `TARGET_TABLE`: Output table for INSERT/CREATE/MERGE (not executable)
+- `SOURCE_QUERY`: SELECT within DML/DDL statements
+- `UNION_BRANCH`: Individual SELECT in UNION/UNION ALL
+
+**Use Cases:**
+- Unit test CTEs and subqueries individually
+- Extract DQL from CTAS, CREATE VIEW, INSERT statements
+- Analyze query structure and component dependencies
+- Break apart complex queries for understanding
+
+### Different SQL Dialects
+
+```bash
+# PostgreSQL
+uv run sqlglider lineage query.sql --dialect postgres
+
+# Snowflake
+uv run sqlglider lineage query.sql --dialect snowflake
+
+# BigQuery
+uv run sqlglider lineage query.sql --dialect bigquery
+```
+
+### Multi-Query Files
+
+SQL Glider automatically detects and analyzes multiple SQL statements in a single file:
+
+```bash
+# Analyze all queries in a file
+uv run sqlglider lineage multi_query.sql
+
+# Filter to only queries that reference a specific table
+uv run sqlglider lineage multi_query.sql --table customers
+
+# Analyze specific column across all queries
+uv run sqlglider lineage multi_query.sql --column customer_id
+
+# Reverse lineage across all queries (impact analysis)
+uv run sqlglider lineage multi_query.sql --source-column orders.customer_id
+```
+
+**Example multi-query file:**
+```sql
+-- multi_query.sql
+SELECT customer_id, customer_name FROM customers;
+
+SELECT order_id, customer_id, order_total FROM orders;
+
+INSERT INTO customer_orders
+SELECT c.customer_id, c.customer_name, o.order_id
+FROM customers c
+JOIN orders o ON c.customer_id = o.customer_id;
+```
+
+**Output includes query index for each statement:**
+```
+   Query 0: SELECT customer_id, customer_name FROM customers
++---------------------------------------------------+
+| Output Column           | Source Column           |
+|-------------------------+-------------------------|
+| customers.customer_id   | customers.customer_id   |
+| customers.customer_name | customers.customer_name |
++---------------------------------------------------+
+Total: 2 row(s)
+
+       Query 1: SELECT order_id, customer_id, order_total FROM orders
++---------------------------------------------+
+| Output Column      | Source Column          |
+|--------------------+------------------------|
+| orders.customer_id | orders.customer_id     |
+| orders.order_id    | orders.order_id        |
+| orders.order_total | orders.order_total     |
++---------------------------------------------+
+Total: 3 row(s)
+
+       Query 2: INSERT INTO customer_orders ...
++---------------------------------------------+
+| Output Column      | Source Column          |
+|--------------------+------------------------|
+...
+```
+
+### Graph-Based Lineage (Cross-File Analysis)
+
+For analyzing lineage across multiple SQL files, SQL Glider provides graph commands:
+
+```bash
+# Build a lineage graph from a single file
+uv run sqlglider graph build query.sql -o graph.json
+
+# Build from multiple files
+uv run sqlglider graph build query1.sql query2.sql query3.sql -o graph.json
+
+# Build from a directory (recursively finds all .sql files)
+uv run sqlglider graph build ./queries/ -r -o graph.json
+
+# Build from a manifest CSV file
+uv run sqlglider graph build --manifest manifest.csv -o graph.json
+
+# Merge multiple graphs into one
+uv run sqlglider graph merge graph1.json graph2.json -o merged.json
+
+# Query upstream dependencies (find all sources for a column)
+uv run sqlglider graph query graph.json --upstream orders.customer_id
+
+# Query downstream dependencies (find all columns affected by a source)
+uv run sqlglider graph query graph.json --downstream customers.id
+```
+
+**Example Upstream Query Output:**
+```
+                              Sources for 'order_totals.total'
++--------------------------------------------------------------------------------------------+
+| Column | Table  | Hops | Root | Leaf | Paths                              | File           |
+|--------+--------+------+------+------+------------------------------------+----------------|
+| amount | orders |    1 |  Y   |  N   | orders.amount -> order_totals.total| test_graph.sql |
++--------------------------------------------------------------------------------------------+
+
+Total: 1 column(s)
+```
+
+**Example Downstream Query Output:**
+```
+                             Affected Columns for 'orders.amount'
++--------------------------------------------------------------------------------------------+
+| Column | Table        | Hops | Root | Leaf | Paths                              | File           |
+|--------+--------------+------+------+------+------------------------------------+----------------|
+| total  | order_totals |    1 |  N   |  Y   | orders.amount -> order_totals.total| test_graph.sql |
++--------------------------------------------------------------------------------------------+
+
+Total: 1 column(s)
+```
+
+**Output Fields:**
+- **Root**: `Y` if the column has no upstream dependencies (source column)
+- **Leaf**: `Y` if the column has no downstream dependencies (final output)
+- **Paths**: All paths from the dependency to the queried column
+
+**Manifest File Format:**
+```csv
+file_path,dialect
+queries/orders.sql,spark
+queries/customers.sql,postgres
+queries/legacy.sql,
+```
+
+The graph feature is designed for scale - it can handle thousands of SQL files and provides efficient upstream/downstream queries using rustworkx.
+
+## Use Cases
+
+### Data Governance
+
+**Impact Assessment:**
+```bash
+# Before modifying a source column, check its impact
+uv run sqlglider lineage analytics_dashboard.sql --source-column orders.revenue
+```
+
+This helps you understand which downstream outputs will be affected by schema changes.
+
+### Query Understanding
+
+**Source Tracing:**
+```bash
+# Understand where a metric comes from
+uv run sqlglider lineage metrics.sql --column total_revenue
+```
+
+Quickly trace complex calculations back to their source tables.
+
+### Documentation
+
+**Export Lineage:**
+```bash
+# Generate documentation for your queries
+uv run sqlglider lineage query.sql --output-format csv --output-file docs/lineage.csv
+```
+
+Create machine-readable lineage documentation for data catalogs.
+
+### Literal Value Handling
+
+When analyzing UNION queries, SQL Glider identifies literal values (constants) as sources and displays them clearly:
+
+```sql
+-- query.sql
+SELECT customer_id, last_order_date FROM active_customers
+UNION ALL
+SELECT customer_id, NULL AS last_order_date FROM prospects
+UNION ALL
+SELECT customer_id, 'unknown' AS status FROM legacy_data
+```
+
+```bash
+uv run sqlglider lineage query.sql
+```
+
+**Example Output:**
+```
+       Query 0: SELECT customer_id, last_order_date FROM active_customers ...
++---------------------------------------------------------------------+
+| Output Column                    | Source Column                    |
+|----------------------------------+----------------------------------|
+| active_customers.customer_id     | active_customers.customer_id     |
+|                                  | prospects.customer_id            |
+| active_customers.last_order_date | <literal: NULL>                  |
+|                                  | active_customers.last_order_date |
++---------------------------------------------------------------------+
+Total: 4 row(s)
+```
+
+Literal values are displayed as `<literal: VALUE>` to clearly distinguish them from actual column sources:
+- `<literal: NULL>` - NULL values
+- `<literal: 0>` - Numeric literals
+- `<literal: 'string'>` - String literals
+- `<literal: CURRENT_TIMESTAMP()>` - Function calls
+
+This helps identify which branches of a UNION contribute actual data lineage versus hardcoded values.
+
+### Multi-Level Analysis
+
+SQL Glider automatically traces through CTEs and subqueries:
+
+```sql
+-- query.sql
+WITH order_totals AS (
+    SELECT customer_id, SUM(order_amount) as total_amount
+    FROM orders
+    GROUP BY customer_id
+),
+customer_segments AS (
+    SELECT
+        ot.customer_id,
+        c.customer_name,
+        CASE
+            WHEN ot.total_amount > 10000 THEN 'Premium'
+            ELSE 'Standard'
+        END as segment
+    FROM order_totals ot
+    JOIN customers c ON ot.customer_id = c.customer_id
+)
+SELECT customer_name, segment, total_amount
+FROM customer_segments
+```
+
+```bash
+# Trace segment back to its ultimate sources
+uv run sqlglider lineage query.sql --column segment
+# Output: orders.order_amount (through the CASE statement and SUM)
+
+# Find what's affected by order_amount
+uv run sqlglider lineage query.sql --source-column orders.order_amount
+# Output: segment, total_amount
+```
+
+## CLI Reference
+
+```
+sqlglider lineage <sql_file> [OPTIONS]
+
+Arguments:
+  sql_file                    Path to SQL file to analyze [required]
+
+Options:
+  --level, -l                 Analysis level: 'column' or 'table' [default: column]
+  --dialect, -d               SQL dialect (spark, postgres, snowflake, etc.) [default: spark]
+  --column, -c                Specific output column for forward lineage [optional]
+  --source-column, -s         Source column for reverse lineage (impact analysis) [optional]
+  --table, -t                 Filter to only queries that reference this table (multi-query files) [optional]
+  --output-format, -f         Output format: 'text', 'json', or 'csv' [default: text]
+  --output-file, -o           Write output to file instead of stdout [optional]
+  --help                      Show help message and exit
+```
+
+**Notes:**
+- `--column` and `--source-column` are mutually exclusive. Use one or the other.
+- `--table` filter is useful for multi-query files to analyze only queries that reference a specific table.
+
+### Tables Command
+
+```
+sqlglider tables overview <sql_file> [OPTIONS]
+
+Arguments:
+  sql_file                    Path to SQL file to analyze [required]
+
+Options:
+  --dialect, -d               SQL dialect (spark, postgres, snowflake, etc.) [default: spark]
+  --table                     Filter to only queries that reference this table [optional]
+  --output-format, -f         Output format: 'text', 'json', or 'csv' [default: text]
+  --output-file, -o           Write output to file instead of stdout [optional]
+  --templater, -t             Templater for SQL preprocessing (e.g., 'jinja', 'none') [optional]
+  --var, -v                   Template variable in key=value format (repeatable) [optional]
+  --vars-file                 Path to variables file (JSON or YAML) [optional]
+  --help                      Show help message and exit
+```
+
+```
+sqlglider tables pull <sql_file> [OPTIONS]
+
+Arguments:
+  sql_file                    Path to SQL file to analyze [optional, reads from stdin if omitted]
+
+Options:
+  --catalog-type, -c          Catalog provider (e.g., 'databricks') [required if not in config]
+  --ddl-folder, -o            Output folder for DDL files [optional, outputs to stdout if omitted]
+  --dialect, -d               SQL dialect (spark, postgres, snowflake, etc.) [default: spark]
+  --templater, -t             Templater for SQL preprocessing (e.g., 'jinja', 'none') [optional]
+  --var, -v                   Template variable in key=value format (repeatable) [optional]
+  --vars-file                 Path to variables file (JSON or YAML) [optional]
+  --list, -l                  List available catalog providers and exit
+  --help                      Show help message and exit
+```
+
+**Databricks Setup:**
+
+Install the optional Databricks dependency:
+```bash
+pip install sql-glider[databricks]
+```
+
+Configure authentication (via environment variables or `sqlglider.toml`):
+```bash
+export DATABRICKS_HOST="https://your-workspace.cloud.databricks.com"
+export DATABRICKS_TOKEN="dapi..."
+export DATABRICKS_WAREHOUSE_ID="abc123..."
+```
+
+### Dissect Command
+
+```
+sqlglider dissect [sql_file] [OPTIONS]
+
+Arguments:
+  sql_file                    Path to SQL file to analyze [optional, reads from stdin if omitted]
+
+Options:
+  --dialect, -d               SQL dialect (spark, postgres, snowflake, etc.) [default: spark]
+  --output-format, -f         Output format: 'text', 'json', or 'csv' [default: text]
+  --output-file, -o           Write output to file instead of stdout [optional]
+  --templater, -t             Templater for SQL preprocessing (e.g., 'jinja', 'none') [optional]
+  --var, -v                   Template variable in key=value format (repeatable) [optional]
+  --vars-file                 Path to variables file (JSON or YAML) [optional]
+  --help                      Show help message and exit
+```
+
+**Output Fields:**
+- `component_type`: Type of component (CTE, MAIN_QUERY, SUBQUERY, etc.)
+- `component_index`: Sequential order within the query (0-based)
+- `name`: CTE name, subquery alias, or target table name
+- `sql`: The extracted SQL for this component
+- `is_executable`: Whether the component can run standalone (TARGET_TABLE is false)
+- `dependencies`: List of CTE names this component references
+- `location`: Human-readable context (e.g., "WITH clause", "FROM clause")
+- `depth`: Nesting level (0 = top-level)
+- `parent_index`: Index of parent component for nested components
+
+### Graph Commands
+
+```
+sqlglider graph build <paths> [OPTIONS]
+
+Arguments:
+  paths                       SQL file(s) or directory to process [optional]
+
+Options:
+  --output, -o                Output JSON file path [required]
+  --manifest, -m              Path to manifest CSV file [optional]
+  --recursive, -r             Recursively search directories [default: True]
+  --glob, -g                  Glob pattern for SQL files [default: *.sql]
+  --dialect, -d               SQL dialect [default: spark]
+  --node-format, -n           Node format: 'qualified' or 'structured' [default: qualified]
+```
+
+```
+sqlglider graph merge <inputs> [OPTIONS]
+
+Arguments:
+  inputs                      JSON graph files to merge [optional]
+
+Options:
+  --output, -o                Output file path [required]
+  --glob, -g                  Glob pattern for graph files [optional]
+```
+
+```
+sqlglider graph query <graph_file> [OPTIONS]
+
+Arguments:
+  graph_file                  Path to graph JSON file [required]
+
+Options:
+  --upstream, -u              Find source columns for this column [optional]
+  --downstream, -d            Find affected columns for this source [optional]
+  --output-format, -f         Output format: 'text', 'json', or 'csv' [default: text]
+```
+
+**Notes:**
+- `--upstream` and `--downstream` are mutually exclusive. Use one or the other.
+- Graph queries are case-insensitive for column matching.
+
+## Output Formats
+
+### Text Format (Default)
+
+Human-readable Rich table format showing query index and preview:
+
+```
+       Query 0: SELECT customer_name FROM customers c ...
++---------------------------------------------------+
+| Output Column   | Source Column                   |
+|-----------------+---------------------------------|
+| customer_name   | c.customer_name                 |
++---------------------------------------------------+
+Total: 1 row(s)
+```
+
+### JSON Format
+
+Machine-readable structured format with query metadata:
+
+```json
+{
+  "queries": [
+    {
+      "query_index": 0,
+      "query_preview": "SELECT customer_name FROM customers c ...",
+      "level": "column",
+      "lineage": [
+        {
+          "output_name": "customer_name",
+          "source_name": "c.customer_name"
+        }
+      ]
+    }
+  ]
+}
+```
+
+### CSV Format
+
+Spreadsheet-ready tabular format with query index:
+
+```csv
+query_index,output_column,source_column
+0,customer_name,c.customer_name
+```
+
+**Note:** Each source column gets its own row. If an output column has multiple sources, there will be multiple rows with the same `query_index` and `output_column`.
+
+## Development
+
+### Setup
+
+```bash
+# Install dependencies
+uv sync
+
+# Run linter
+uv run ruff check
+
+# Auto-fix issues
+uv run ruff check --fix
+
+# Format code
+uv run ruff format
+
+# Type checking
+uv run basedpyright
+```
+
+### Project Structure
+
+See [ARCHITECTURE.md](ARCHITECTURE.md) for detailed technical documentation.
+
+```
+src/sqlglider/
+├── cli.py                    # Typer CLI entry point
+├── dissection/
+│   ├── analyzer.py           # DissectionAnalyzer for query decomposition
+│   ├── formatters.py         # Output formatters (text, JSON, CSV)
+│   └── models.py             # ComponentType, SQLComponent, QueryDissectionResult
+├── graph/
+│   ├── builder.py            # Build graphs from SQL files
+│   ├── merge.py              # Merge multiple graphs
+│   ├── query.py              # Query upstream/downstream lineage
+│   └── models.py             # Graph data models
+├── lineage/
+│   ├── analyzer.py           # Core lineage analysis using SQLGlot
+│   └── formatters.py         # Output formatters (text, JSON, CSV)
+└── utils/
+    └── file_utils.py         # File I/O utilities
+```
+
+## Publishing
+
+SQL Glider is configured for publishing to both TestPyPI and PyPI using `uv`.
+
+### Versioning
+
+SQL Glider uses Git tags for version management via [hatch-vcs](https://github.com/ofek/hatch-vcs). The version is automatically derived from Git:
+
+- **Tagged commits:** Version matches the tag (e.g., `git tag v0.2.0` produces version `0.2.0`)
+- **Untagged commits:** Version includes development info (e.g., `0.1.dev18+g7216a59`)
+
+**Creating a new release:**
+
+```bash
+# Create and push a version tag
+git tag v0.2.0
+git push origin v0.2.0
+
+# Build will now produce version 0.2.0
+uv build
+```
+
+**Tag format:** Use `v` prefix (e.g., `v1.0.0`, `v0.2.1`). The `v` is stripped from the final version number.
+
+### Building the Package
+
+```bash
+# Build the distribution files (wheel and sdist)
+uv build
+```
+
+This creates distribution files in the `dist/` directory.
+
+### Publishing to TestPyPI
+
+Always test your release on TestPyPI first:
+
+```bash
+# Publish to TestPyPI
+uv publish --index testpypi --token <YOUR_TESTPYPI_TOKEN>
+
+# Test installation from TestPyPI
+uv pip install --index-url https://test.pypi.org/simple/ sql-glider
+```
+
+### Publishing to PyPI
+
+Once verified on TestPyPI, publish to production:
+
+```bash
+# Publish to PyPI
+uv publish --index pypi --token <YOUR_PYPI_TOKEN>
+```
+
+### Token Setup
+
+You'll need API tokens from both registries:
+
+1. **TestPyPI Token:** Create at https://test.pypi.org/manage/account/token/
+2. **PyPI Token:** Create at https://pypi.org/manage/account/token/
+
+**Option 1: Pass token directly (shown above)**
+
+**Option 2: Environment variable**
+```bash
+export UV_PUBLISH_TOKEN=pypi-...
+uv publish --index pypi
+```
+
+**Option 3: Store in `.env` file (not committed to git)**
+```bash
+# .env
+UV_PUBLISH_TOKEN=pypi-...
+```
+
+**Security Note:** Never commit API tokens to version control. The `.gitignore` file should include `.env`.
+
+## Dependencies
+
+- **sqlglot[rs]:** SQL parser and lineage analysis library with Rust extensions
+- **typer:** CLI framework with type hints
+- **rich:** Terminal formatting and colored output
+- **pydantic:** Data validation and serialization
+- **rustworkx:** High-performance graph library for cross-file lineage analysis
+
+## References
+
+- [SQLGlot Documentation](https://sqlglot.com/)
+- [UV Documentation](https://docs.astral.sh/uv/)
+- [Typer Documentation](https://typer.tiangolo.com/)
+- [Ruff Documentation](https://docs.astral.sh/ruff/configuration/)
+
+## License
+
+See LICENSE file for details.