neo4j-transfer 0.2.0

Tool for transferring select data from one active Neo4j instance to another.

git s# Neo4j Transfer

A Python tool for efficiently transferring selected data from one Neo4j instance to another, with support for batch processing, progress tracking, and transfer metadata.

Features

• Selective Data Transfer: Transfer specific node labels and relationship types
• Batch Processing: Efficient large-scale transfers with configurable batch sizes
• Progress Tracking: Real-time progress updates during transfers
• Transfer Metadata: Optional augmentation with source element IDs and timestamps
• Undo Capability: Remove transferred data using the original transfer specification
• Database Reset: Option to clear target database before transfer
• Stoppable Transfers: Cancel long-running transfers gracefully

Documentation

📖 View Full API Documentation

Installation

From GitHub (Development)

pip install git+https://github.com/jalakoo/neo4j-transfer.git@main

From PyPI (When Available)

pip install neo4j-transfer

Quick Start

Basic Transfer

from neo4j_transfer import Neo4jCredentials, TransferSpec, transfer

# Define source and target database credentials
source_creds = Neo4jCredentials(
    uri="bolt://localhost:7687",
    username="neo4j",
    password="source_password",
    database="neo4j"
)

target_creds = Neo4jCredentials(
    uri="bolt://localhost:7688", 
    username="neo4j",
    password="target_password",
    database="neo4j"
)

# Specify what to transfer
spec = TransferSpec(
    node_labels=["Person", "Company"],
    relationship_types=["WORKS_FOR", "OWNS"]
)

# Execute the transfer
result = transfer(source_creds, target_creds, spec)
print(f"Transfer completed: {result.nodes_created} nodes, {result.relationships_created} relationships")

Transfer with Progress Tracking

from neo4j_transfer import transfer_generator

# Use the generator for real-time progress updates
for update in transfer_generator(source_creds, target_creds, spec):
    if hasattr(update, 'records_completed'):
        progress = update.float_completed()
        print(f"Progress: {progress:.1%}")

Transfer with Metadata Augmentation

spec = TransferSpec(
    node_labels=["Person"],
    relationship_types=["KNOWS"],
    should_append_data=True,  # Add transfer metadata
    element_id_key="_source_element_id",
    timestamp_key="_transfer_timestamp"
)

Undo a Transfer

from neo4j_transfer import undo

# Remove all data transferred with this spec
undo(target_creds, spec)

Advanced Usage

Database Reset Before Transfer

spec = TransferSpec(
    node_labels=["Person"],
    overwrite_target=True,  # Clear target database first
    batch_size=10000
)

Stoppable Transfer

gen = transfer_generator(source_creds, target_creds, spec)
controller = next(gen)  # Get the controller

# In another thread or after some condition:
controller.request_stop()

# Continue consuming updates
for update in gen:
    # Process updates until stop is requested
    pass

API Reference

Neo4jCredentials

Credentials for accessing a Neo4j database instance.

• uri (str): The URI address of the Neo4j database
• password (str): The password for authentication
• username (str, optional): Username (default: "neo4j")
• database (str, optional): Database name (default: "neo4j")

TransferSpec

Configuration for data transfer between Neo4j instances.

• node_labels (list[str]): Node labels to transfer
• relationship_types (list[str], optional): Relationship types to transfer
• should_append_data (bool): Add transfer metadata to copied data (default: True)
• element_id_key (str): Property key for source element ID (default: "_transfer_element_id")
• timestamp_key (str): Property key for transfer timestamp (default: "_transfer_timestamp")
• overwrite_target (bool): Clear target database before transfer (default: False)
• batch_size (int): Batch size for transfers (default: 10000)

UploadResult

Result object containing transfer statistics.

• nodes_created (int): Number of nodes created
• relationships_created (int): Number of relationships created
• records_total (int): Total records processed
• records_completed (int): Records completed
• was_successful (bool): Whether transfer succeeded
• seconds_to_complete (float): Transfer duration
• float_completed(): Returns completion percentage as float (0.0-1.0)

Requirements

• Python 3.11+
• Neo4j 4.0+
• Access to both source and target Neo4j instances

License

MIT License - see LICENSE.txt for details.