Fabricator

A modern, enterprise-grade CSV data generator for system-of-record testing

Fabricator is a powerful command-line tool that generates realistic CSV test data for system-of-record (SOR) platforms. Built with a robust pipeline architecture and comprehensive validation, it transforms YAML definitions into consistent, relationship-aware CSV datasets.

🚀 Quick Start

# Download the latest release for your platform
curl -L https://github.com/SGNL-ai/fabricator/releases/latest/download/fabricator-linux -o fabricator
chmod +x fabricator

# Generate test data from a YAML definition
./fabricator -f examples/sample.yaml -n 1000 -o ./test-data

# Output:
# ✓ Generated 16 CSV files with 1000 rows each
# ✓ All relationships consistent across files
# ✓ Entity-relationship diagram created

🏗️ Architecture

┌─────────────────┐    ┌──────────────────┐    ┌─────────────────┐
│   YAML Input    │───▶│   Validation     │───▶│   Pipeline      │
│                 │    │                  │    │                 │
│ • Entities      │    │ • JSON Schema    │    │ • Phase 1: IDs  │
│ • Attributes    │    │ • Business Logic │    │ • Phase 2: Rels │
│ • Relationships │    │ • 96% Template   │    │ • Phase 3: Data │
└─────────────────┘    │   Compatibility  │    └─────────────────┘
                       └──────────────────┘             │
                                                        ▼
┌─────────────────┐    ┌──────────────────┐    ┌─────────────────┐
│   CSV Output    │◀───│    Validation    │◀───│  Data Model     │
│                 │    │                  │    │                 │
│ • Multi-file    │    │ • Referential    │    │ • Graph         │
│ • Consistent    │    │   Integrity      │    │ • Entities      │
│ • Realistic     │    │ • Uniqueness     │    │ • Relationships │
└─────────────────┘    └──────────────────┘    └─────────────────┘

✨ Features

🏗️ Robust Architecture

• Pipeline-based processing with clean separation of concerns
• Comprehensive validation with JSON Schema + business logic layers
• Graph-based dependency resolution with topological sorting

📊 Data Generation

• Realistic test data with type-aware field generation
• Relationship consistency across all CSV files
• Variable cardinalities (1:1, 1:N, N:1, N:N) with auto-detection
• Configurable data volume from small samples to large datasets

🔍 Validation & Quality

• YAML schema validation using industry-standard JSON Schema
• Relationship integrity checking across entities
• Uniqueness constraint validation
• Production template compatibility (96% of SGNL catalog templates supported)

🎨 User Experience

• Colorful CLI output with progress indicators
• SVG diagram generation for entity-relationship visualization
• Detailed error messages with actionable guidance
• Multiple operation modes (generate, validate-only, diagram-only)

Installation

Requires Go 1.24.3 or higher.

From Source

# Clone the repository
git clone https://github.com/SGNL-ai/fabricator.git
cd fabricator

# Build the project
make build

The binary will be built to build/fabricator.

From GitHub Releases

Pre-built binaries for Linux, macOS (Intel and Apple Silicon), and Windows are automatically generated for each release and can be downloaded from the GitHub Releases page.

# For macOS Intel
curl -L https://github.com/SGNL-ai/fabricator/releases/latest/download/fabricator-macos-intel -o fabricator
chmod +x fabricator
./fabricator --version

# For macOS Apple Silicon (M1/M2/M3)
curl -L https://github.com/SGNL-ai/fabricator/releases/latest/download/fabricator-macos-apple-silicon -o fabricator
chmod +x fabricator
./fabricator --version

# For Linux
curl -L https://github.com/SGNL-ai/fabricator/releases/latest/download/fabricator-linux -o fabricator
chmod +x fabricator
./fabricator --version

For Windows users, download the fabricator-windows.exe file from the releases page.

Usage

# Basic usage (short options)
./build/fabricator -f <yaml-file> [-o <dir>] [-n <count>] [-a]

# Basic usage (long options)
./build/fabricator --file <yaml-file> [--output <dir>] [--num-rows <count>] [--auto-cardinality]

# View version information
./build/fabricator -v

Command Line Options

Short Flag	Long Flag	Description	Default
-f	--file	Path to the YAML definition file (required)	-
-o	--output	Directory to store generated CSV files	"output"
-n	--num-rows	Number of rows to generate for each entity	100
-a	--auto-cardinality	Enable automatic cardinality detection	false
-d	--diagram	Generate Entity-Relationship diagram	true
	--validate	Validate relationships in CSV files	true
	--validate-only	Validate existing CSV files without generation	false
-v	--version	Display version information	-

Examples

# Generate CSV files from example.yaml with 500 rows per entity
./build/fabricator -f example.yaml -n 500 -o data/sgnl

# Using long-form options
./build/fabricator --file example.yaml --num-rows 1000 --output export/data

# Generate CSV files with automatic cardinality detection for relationships
./build/fabricator -f example.yaml -n 200 -a

# Using long-form options with auto-cardinality
./build/fabricator --file example.yaml --num-rows 500 --auto-cardinality --output data/variable-cardinality

# Generate CSV files but disable ER diagram generation
./build/fabricator -f example.yaml --diagram=false

# Validate existing CSV files without generating new data
./build/fabricator -f example.yaml -o existing/csv/data --validate-only

# Validate existing CSV files and generate an ER diagram
./build/fabricator -f example.yaml -o existing/csv/data --validate-only --diagram

YAML Format

The YAML file should define a system-of-record structure, including:

• Entities with attributes
• Relationships between entities
• External IDs that will be used for CSV filenames

Each entity in the YAML file will result in a corresponding CSV file, with the filename derived from the entity's externalId.

Generated Data & Validation

The tool provides the following functionality:

1. CSV Generation:

   • CSV files named after each entity's external ID (without the namespace prefix)
   • Headers matching the entity's attribute external IDs
   • Consistent data across relationships between entities
   • Variable cardinality relationships (with the -a flag)
   • Realistic test data based on attribute names and types

2. CSV Validation (via --validate-only):

   • Checks existing CSV files against a YAML definition
   • Validates relationship consistency across entities
   • Verifies unique constraint requirements are met
   • Helpful for validating production or manually-created data exports
   • Use with the existing output directory containing CSV files

3. Entity-Relationship Diagram (enabled by default):

   • SVG visualization of all entities and their relationships
   • Color-coded entities with attributes listed
   • Primary keys (uniqueId attributes) highlighted
   • Relationship cardinality indicators (1:1, 1:N, N:1, N:M)
   • Can be disabled with --diagram=false
   • Works in both generation and validation-only modes

The data generator intelligently creates appropriate values based on field names:

• ID fields get unique identifiers
• Name fields get contextual names based on entity types (e.g., person names for users, company names for organizations)
• Date fields get properly formatted dates
• Email fields get valid email addresses
• Boolean fields get true/false values
• Numeric fields get appropriate numbers

Relationship Cardinality

When the auto-cardinality feature is enabled (-a flag), Fabricator automatically detects and generates appropriate cardinality for entity relationships:

• 1:1 relationships - Simple one-to-one mappings between entities
• 1:N relationships - One entity related to multiple instances of another entity
• N:1 relationships - Multiple entities related to a single instance of another entity

Cardinality detection is based on:

1. Entity metadata (primary detection method)

   • Fields with uniqueId: true are used to identify key relationships
   • When a relationship links a unique ID to a non-unique field, cardinality is automatically determined

2. Field naming patterns (fallback method)

   • Field names ending with "Id" typically indicate N:1 relationships
   • Plural field names or names ending with "Ids" suggest 1:N relationships

Without the -a flag, all relationships default to 1:1 cardinality.

📈 Performance

Fabricator is designed for efficiency and can handle large datasets:

Dataset Size	Entities	Time	Memory
Small	5	<1s	<50MB
Medium	16	2-5s	<100MB
Large	50	10-30s	<500MB

Benchmarks (16 entities, complex relationships):

• 1,000 rows/entity: ~3 seconds, 16 CSV files
• 10,000 rows/entity: ~15 seconds, consistent relationships
• 100,000 rows/entity: ~2 minutes, 1.6M total records

🛠️ Development

Prerequisites for Development

• Go 1.23+ (tested with 1.23 and 1.24)
• golangci-lint for code quality
• Pre-commit hooks (optional but recommended)

Development Commands

# Run tests
make test

# Run tests with coverage
make coverage

# Format code
make fmt

# Static analysis
make vet

# Run linter
make lint

# Run all checks (CI pipeline)
make ci

# Security scanning
gosec ./...
govulncheck ./...

Contributing

See CONTRIBUTING.md for detailed development guidelines, architecture documentation, and contribution workflow.

License

This project is licensed under the MIT License - see the LICENSE file for details.