retroflow 0.8.0

Generate beautiful ASCII and PNG flowcharts from simple text descriptions

RetroFlow

 ___________________________________________________________
|                                                           |
|   _____      _             ______ _                       |
|  |  __ \    | |           |  ____| |                      |
|  | |__) |___| |_ _ __ ___ | |__  | | _____      __        |
|  |  _  // _ \ __| '__/ _ \|  __| | |/ _ \ \ /\ / /        |
|  | | \ \  __/ |_| | | (_) | |    | | (_) \ V  V /         |
|  |_|  \_\___|\__|_|  \___/|_|    |_|\___/ \_/\_/          |
|                                                           |
|            [ Beautiful ASCII Flowcharts ]                 |
|___________________________________________________________|
               ||                           ||

A Python library for generating beautiful ASCII and PNG flowcharts from simple text descriptions.

Overview

This is a project to help engineers, researchers, project managers, and others create beautiful, retro-looking ASCII flow diagrams. ASCII diagrams are pretty, and harken back to the mid-20th century technical documentation. They also have real advantages:

1. They work wonderfully in the age of agentic AI, which can easily read and parse these small representations.
2. ASCII diagrams can live inline with: PRs, Markdown files, Slack threads, etc.
3. ASCII diagrams optimize for thinking speed, not presentation quality. It encourages iteration and deletion instead of premature refinement.
4. Minimalist diagrams reduce visual noise (although they do still look retro and pretty)
5. They're tool agnostic and can be rendered anywhere

    ┌─────────┐
    │  START  │░
    └─────────┘░
     ░░░░░░░░░░░
         │
         ▼
    ┌─────────┐
    │ PROCESS │░
    └─────────┘░
     ░░░░░░░░░░░
         │
         ▼
    ┌─────────┐
    │DECISION │░
    └─────────┘░
     ░░░░░░░░░░░
         │
         ▼
    ┌─────────┐
    │   END   │░
    └─────────┘░
     ░░░░░░░░░░░

Installation

pip install retroflow

Or with uv:

uv add retroflow

Quick Start

from retroflow import FlowchartGenerator

# Create a generator
generator = FlowchartGenerator()

# Define your flowchart using simple arrow syntax
flowchart = generator.generate("""
    Start -> Process
    Process -> Decision
    Decision -> End
""")

print(flowchart)

Features

• Simple syntax: Define flowcharts using intuitive A -> B arrow notation
• ASCII output: Generate text-based flowcharts for terminals and documentation
• PNG export: Save high-resolution PNG images with customizable fonts
• Intelligent layout: Automatic node positioning using NetworkX with barycenter heuristic
• Smart edge routing: Edges automatically route around intermediate boxes to avoid visual overlap
• Cycle detection: Handles cyclic graphs gracefully with back-edge routing
• Customizable: Adjust text width, box sizes, spacing, shadows, and fonts
• Unicode box-drawing: Beautiful boxes with optional shadow effects
• Title banners: Optional double-line bordered titles with automatic word wrapping
• Horizontal flow: Left-to-right layout mode for compact linear diagrams

Usage

Basic Generation

from retroflow import FlowchartGenerator

generator = FlowchartGenerator()

# Simple linear flow
result = generator.generate("""
    A -> B
    B -> C
    C -> D
""")
print(result)

Branching Flows

# Branching and merging
result = generator.generate("""
    Start -> Validate
    Validate -> Process
    Validate -> Error
    Process -> End
    Error -> End
""")
print(result)

Cyclic Flows

# Loops and cycles are supported
result = generator.generate("""
    Init -> Check
    Check -> Process
    Process -> Check
    Check -> Done
""")
print(result)

Export to Text File

generator = FlowchartGenerator()

generator.save_txt("""
    A -> B
    B -> C
""", "flowchart.txt")

Export to PNG

Save your flowchart as a high-resolution PNG image:

generator = FlowchartGenerator()

generator.save_png("""
    A -> B
    B -> C
""", "flowchart.png")

Customize the PNG output with various options:

generator.save_png(
    "A -> B\nB -> C",
    "flowchart.png",
    font_size=24,           # Base font size in points (default: 16)
    bg_color="#1a1a2e",     # Background color (default: "#FFFFFF")
    fg_color="#00ff00",     # Text color (default: "#000000")
    padding=40,             # Padding around diagram in pixels (default: 20)
    scale=2,                # Resolution multiplier for crisp output (default: 2)
)

The scale parameter controls the output resolution. With the default scale=2, images render at 2x resolution for crisp display on high-DPI/retina screens. Use scale=1 for smaller file sizes, or scale=3 for even sharper output.

Configuration

FlowchartGenerator Options

generator = FlowchartGenerator(
    max_text_width=22,      # Max width for text inside boxes before wrapping (default: 22)
    min_box_width=10,       # Minimum box width (default: 10)
    horizontal_spacing=12,  # Space between boxes horizontally (default: 12)
    vertical_spacing=6,     # Space between boxes vertically (default: 6)
    shadow=True,            # Whether to draw box shadows (default: True)
    font="Cascadia Code",   # Font for PNG output (default: None, uses system font)
    title="My Diagram",     # Optional title banner with double-line border (default: None)
    direction="TB",         # Flow direction: "TB" (top-to-bottom) or "LR" (left-to-right)
)

Font Configuration

You can specify a font for PNG output either at initialization or per-call:

# Set font for all PNG exports from this generator
generator = FlowchartGenerator(font="Cascadia Code")
generator.save_png("A -> B", "flowchart.png")

# Or override the font for a specific export
generator.save_png("A -> B", "flowchart.png", font="Monaco")

Common monospace fonts that work well:

• Windows: Cascadia Code, Consolas, Courier New
• macOS: Monaco, Menlo, SF Mono
• Linux: DejaVu Sans Mono, Ubuntu Mono, Liberation Mono

If the specified font isn't found, RetroFlow automatically falls back to available system fonts.

Title Banner

Add a title banner with a double-line border above your flowchart:

generator = FlowchartGenerator(title="CI/CD Pipeline")
result = generator.generate("""
    Build -> Test
    Test -> Deploy
""")

Output:

╔══════════════════╗
║  CI/CD Pipeline  ║
╚══════════════════╝

    ┌─────────┐
    │  Build  │░
    └────┬────┘░
     ░░░░│░░░░░░
         │
         ▼
    ┌─────────┐
    │  Test   │░
    └────┬────┘░
     ░░░░│░░░░░░
         │
         ▼
    ┌──────────┐
    │  Deploy  │░
    └──────────┘░
     ░░░░░░░░░░░░

Title Wrapping: Long titles automatically wrap at word boundaries (approximately every 15 characters) to keep the title box compact:

generator = FlowchartGenerator(title="My Very Long Project Title Here")

╔═════════════════╗
║  My Very Long   ║
║  Project Title  ║
║      Here       ║
╚═════════════════╝

You can also override the title per-call:

generator.generate("A -> B", title="Override Title")

Horizontal Flow Mode

Use direction="LR" for left-to-right layouts instead of top-to-bottom:

generator = FlowchartGenerator(direction="LR")
result = generator.generate("""
    Start -> Process
    Process -> End
""")

This produces a horizontal flowchart where nodes flow from left to right, which can be more compact for linear processes.

Input Format

The input format uses a simple arrow syntax:

# Comments start with #
NodeA -> NodeB
NodeB -> NodeC

# Node names can contain spaces
User Login -> Validate Credentials
Validate Credentials -> Access Granted

Rules

• Each line defines one connection: Source -> Target
• Node names are trimmed of whitespace
• Empty lines are ignored
• Lines starting with # are comments
• Node names can contain spaces

API Reference

FlowchartGenerator

The main class for generating flowcharts.

Methods

Method	Description
generate(input_text)	Generate ASCII flowchart string
save_txt(input_text, filename)	Save flowchart to a text file
save_png(input_text, filename, ...)	Save flowchart as a PNG image

Convenience Functions

from retroflow import parse_flowchart

# Parse without rendering
connections = parse_flowchart("A -> B\nB -> C")
# Returns: [('A', 'B'), ('B', 'C')]

Lower-Level API

For advanced use cases, you can access the individual components:

from retroflow import (
    Parser,
    SugiyamaLayout,  # Alias for NetworkXLayout
    Canvas,
    BoxRenderer,
)

# Parse input
parser = Parser()
connections = parser.parse("A -> B\nB -> C")

# Compute layout
layout_engine = SugiyamaLayout()
layout_result = layout_engine.layout(connections)

# Access layout information
for node_name, node_layout in layout_result.nodes.items():
    print(f"{node_name}: layer={node_layout.layer}, position={node_layout.position}")

Development

Setup

git clone https://github.com/ronikobrosly/retroflow.git
cd retroflow
uv sync --dev

Run Tests

uv run pytest

Run Linting

uvx ruff check .
uvx ruff format --check .

License

MIT License - see LICENSE for details.

Contributing

Contributions are 100% welcome! Please feel free to submit a PR.