RyoFaker 1.0.0

Enterprise data generation library extending Faker

RyoFaker 🎲

Enterprise data generation library extending Faker with schema-driven generation, referential integrity management, and data quality features specifically designed for QA engineers, data scientists, and ETL developers.

---

Why RyoFaker?

RyoFaker solves real-world data generation challenges that standard Faker doesn't address:

✅ Schema-driven generation from JSON/YAML (no repetitive code)
✅ Referential integrity with automatic FK management
✅ Data quality testing (inject nulls, duplicates, constraint violations)
✅ Time-series & CDC event simulation
✅ India-specific providers (PAN, Aadhaar, GSTIN, IFSC, IMEI, IMSI)
✅ Multi-format output (Pandas, PySpark, Parquet, CSV, SQL, Cloud)
✅ 100% Faker-compatible — drop-in replacement

---

Installation

Basic installation pip install RyoFaker

With PySpark support pip install RyoFaker[spark]

With cloud storage support (OCI, Azure, S3) pip install RyoFaker[cloud]

Install everything pip install RyoFaker[all]

text

---

Quick Start

Basic Usage (Faker-compatible)

from ryofaker import RyoFaker

All Faker methods work identically rf = RyoFaker('en_IN') print(rf.name()) # "Aryan Mullick" print(rf.address()) # "Plot 45, Sector 12, New Delhi 110001"

text

India-Specific Providers

Indian identity documents print(rf.pan()) # "ABCDE1234F" print(rf.aadhaar()) # "1234 5678 9012" print(rf.gstin()) # "27ABCDE1234F1Z5" print(rf.ifsc()) # "SBIN0001234"

Telecom identifiers print(rf.imei()) # "123456789012345" print(rf.imsi()) # "404011234567890" print(rf.msisdn()) # "[phone]"

text

Schema-Driven Generation

Define schema once (customer_order.json):

{ "tables": { "customers": { "rows": 100, "columns": { "customer_id": {"provider": "uuid4", "primary_key": true}, "first_name": {"provider": "first_name"}, "last_name": {"provider": "last_name"}, "email": {"provider": "email", "unique": true}, "pan": {"provider": "pan"}, "created_at": {"provider": "date_between", "args": {"start_date": "-2y"}} } }, "orders": { "rows": 500, "columns": { "order_id": {"provider": "uuid4", "primary_key": true}, "customer_id": {"foreign_key": "customers.customer_id"}, "order_date": {"provider": "date_time_this_year"}, "amount": {"provider": "pydecimal", "args": {"left_digits": 5, "right_digits": 2}}, "status": {"provider": "random_element", "args": {"elements": ["pending", "shipped", "delivered"]}} } } }, "relationships": [ {"parent": "customers", "child": "orders", "ratio": 5} ] }

text

Generate with one line:

from ryofaker import RyoFaker

rf = RyoFaker() data = rf.from_schema('customer_order.json', format='pandas')

Returns: {'customers': DataFrame, 'orders': DataFrame} FK integrity guaranteed: all orders.customer_id exist in customers.customer_id text

Referential Integrity

Manual FK management tables = rf.with_relationships({ 'users': { 'rows': 50, 'schema': {'user_id': 'uuid4', 'name': 'name'} }, 'orders': { 'rows': 200, 'schema': {'order_id': 'uuid4', 'user_id': 'fk:users.user_id', 'amount': 'pydecimal'} } })

Guaranteed: Every orders.user_id exists in users.user_id text

Data Quality Testing

import pandas as pd

Generate clean data df = rf.from_schema('schema.json', format='pandas')['customers']

Inject nulls for testing df_with_nulls = rf.inject_nulls(df, columns=['phone', 'email'], rate=0.1)

Inject duplicates for deduplication testing df_with_dupes = rf.inject_duplicates(df, rate=0.05)

Edge case testing print(rf.testing.max_varchar(255)) # String exactly 255 chars print(rf.testing.unicode_emoji()) # "🚀💯😊" print(rf.testing.sql_injection()) # "'; DROP TABLE users; --"

text

Time-Series & CDC Simulation

Generate CDC event stream for event in rf.cdc_stream(schema={'user_id': 'uuid4', 'name': 'name'}, duration_seconds=60): print(event)

{'operation': 'INSERT', 'timestamp': '2025-10-14T10:30:15', 'data': {...}}

{'operation': 'UPDATE', 'timestamp': '2025-10-14T10:30:45', 'data': {...}}

text

Multi-Format Output

Pandas DataFrame (default) df = rf.from_schema('schema.json', format='pandas')

PySpark DataFrame spark_df = rf.from_schema('schema.json', format='pyspark')

Parquet file rf.from_schema('schema.json', format='parquet', output='data.parquet')

CSV file rf.from_schema('schema.json', format='csv', output='data.csv')

SQL INSERT statements sql = rf.from_schema('schema.json', format='sql')

text

---

Use Cases

1. Software Testing & QA

Repeatable regression datasets rf = RyoFaker() rf.seed(42) test_data = rf.from_schema('test_schema.json', rows=1000)

Same seed = identical data every run text

2. Load Testing

Generate 10 million rows at scale rf.from_schema('load_test.json', rows=10_000_000, format='parquet', output='load_data.parquet')

text

3. Privacy & Compliance

Replace production data with synthetic copies synthetic_users = rf.from_schema('user_schema.json', rows=100000)

Preserves patterns but no real identities text

4. ETL Testing

CDC stream testing for event in rf.cdc_stream(schema={...}, duration_seconds=300): kafka_producer.send('cdc-topic', event)

text

---

Custom Providers Available

Enterprise Providers

Provider	Methods	Use Case
india_identity	pan(), aadhaar(), gstin(), ifsc()	Indian KYC, banking, tax
telecom	imei(), imsi(), msisdn()	Mobile network, SIM provisioning
healthcare	mrn(), insurance_id(), icd10_code()	Patient records, claims
retail	sku(), upc(), order_id()	Inventory, e-commerce
banking	account_number(), transaction_id(), swift_code()	Payment systems, ledgers

Testing Providers

Provider	Methods	Use Case
edge_cases	max_varchar(), unicode_emoji(), sql_injection()	Boundary testing, security
stress	random_blob(), deep_json()	Load testing, performance
regression	Seeded fixed datasets	Reproducible test scenarios

---

Advanced Features

Dependency Graph Resolution

RyoFaker uses NetworkX to automatically resolve FK dependencies and generate tables in the correct order:

Define complex multi-table schema with circular references RyoFaker automatically detects and resolves the dependency graph data = rf.from_schema('complex_schema.json')

text

Progress Tracking

Built-in progress bars for large datasets rf.from_schema('big_schema.json', rows=1_000_000, show_progress=True)

Progress: 100%|████████████| 1000000/1000000 [00:45<00:00, 22000rows/s] text

Locale Support

All Faker locales supported + India-specific enhancements rf_india = RyoFaker('en_IN') rf_us = RyoFaker('en_US') rf_multi = RyoFaker(['en_IN', 'hi_IN'], use_weighting=True)

text

---

📚 Documentation

• Full Documentation: https://ryofaker.readthedocs.io
• API Reference: https://ryofaker.readthedocs.io/api/
• Schema Reference: https://ryofaker.readthedocs.io/schema/
• Examples: https://github.com/ada/ryofaker/tree/main/examples

---

🤝 Contributing

We welcome contributions! Please see CONTRIBUTING.md for guidelines.

Clone repository git clone https://github.com/ada/ryofaker.git cd ryofaker

Install in development mode pip install -e ".[dev]"

Run tests pytest tests/ -v

Format code black ryofaker/ tests/

text

---

📄 License

RyoFaker is licensed under the MIT License. See LICENSE for details.

---

🙏 Acknowledgments

RyoFaker is built on top of the excellent Faker library. We extend deep gratitude to the Faker maintainers and contributors.

---

📞 Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Email: [e-mail]

---

Made with ❤️ by the ADA Data Science Team