notion2pg-bulk 0.1.0

A Python package that automatically finds and migrates to PostgreSQL all databases in a Notion workspace.

notion2pg_bulk

• Automatically discovers and migrates all databases in a workspace regardless of their views or position in the page structure
• Preserves database descriptions and property descriptions
• Supports relation properties using PostgreSQL arrays
• Select field support: Single and multi-select options stored in separate lookup tables with foreign key constraints
• Complies with Notion API limits (3 requests/second average)

Installation

pip install notion2pg-bulk

Setup and Usage

Requirements

• a Notion API key connected to all the databased that you wish to migrate (note: Notion now allows to connect databases ditectly from the integration configuration page, under the 'Access' tab. For quickly connecting all databases in a workspace, just tick all top level pages.
• PostgreSQL database connection URL

Interactive Mode

On (Default):

• Dispays detected databases and fields and asks for confirmation before running the migration
• Progress bar
• Shows post-migration notes

Off: (--quiet flag in CLI or set interactive_mode=False in Python API)

• .run() runs migration directly and silently

CLI

Options:

• --notion-token: Notion integration token (or set NOTION_TOKEN env var)
• --database-url: PostgreSQL connection string (or set DATABASE_URL env var)
• --quiet: Run in non-interactive mode (skip validation steps and progress bars)
• --extract-page-content: Extract free-form content from page bodies (slower migration)

Examples:

# Interactive mode (default)
notion2pg-bulk --notion-token "your_token" --database-url "postgresql://..."

# Non-interactive mode (skip validation steps)
notion2pg-bulk --notion-token "your_token" --database-url "postgresql://..." --quiet

# Extract page content (slower but includes free-form content)
notion2pg-bulk --notion-token "your_token" --database-url "postgresql://..." --extract-page-content

# Non-interactive with page content extraction
notion2pg-bulk --notion-token "your_token" --database-url "postgresql://..." --quiet --extract-page-content

Python API:

from notion2pg_bulk import NotionMigrator
import sqlalchemy as sa

engine = sa.create_engine('postgresql://user:password@localhost/dbname')
migrator = NotionMigrator(
    notion_token="your_notion_integration_token",
    db_connection=engine,
    interactive_mode=True,
    extract_page_content=True,
)
migrator.run()

Property Type Mapping

Notion Property	PostgreSQL Type	Notes
Title	TEXT	Primary identifier
Rich Text	TEXT	Preserved as markdown with formatting
Number	NUMERIC
Select	TEXT	Single option value + lookup table with foreign key
Multi-select	TEXT[]	Array of option values + lookup table with check constraints
Date	TIMESTAMP
Checkbox	BOOLEAN
URL	TEXT
Email	TEXT
Phone	TEXT
Relation	TEXT[]	Array of related page IDs
People	TEXT[]	Array of user IDs
Files	TEXT[]	Array of file URLs
Created time	TIMESTAMP
Created by	TEXT	User ID
Last edited time	TIMESTAMP
Last edited by	TEXT	User ID
Formula	-	⚠️ Skipped (not supported by Notion API)
Rollup	-	⚠️ Skipped (not supported by Notion API)

Additional Page Content Mode

The migrator can optionally extract free-form content from database pages. Extracted content is stored in the additional_page_content column as plain text.

migrator = NotionMigrator(
    notion_token=token,
    db_connection=engine,
    extract_page_content=True  # Default is false
)

Supported blocks:

• Paragraph text
• Headings (all levels)
• Bulleted and numbered lists
• Code blocks
• To-do items
• Callouts
• Toggle blocks
• Dividers
• Embedded databases (ID reference only)

Unsupported blocks:

• Images and media files
• Complex block types (equations, embeds, etc.)
• Block formatting and styling
• Page hierarchies and relationships

Post-migration analysis: When page content extraction is enabled, the migrator provides a comprehensive summary of:

• Unsupported block types found and their locations (grouped by record)
• Embedded databases that were referenced in page content but not connected to the integration (with full database IDs)