Lotio

High-performance Lottie animation frame renderer using Skia. Renders animations to PNG frames for video encoding.

Installation

Homebrew (Recommended)

brew tap matrunchyk/lotio
brew install lotio

This installs:

• Binary: lotio
• Headers: /opt/homebrew/include/lotio/ (or /usr/local/include/lotio/)
• Libraries: /opt/homebrew/lib/ (Skia static libraries)

From Source

Prerequisites (macOS):

brew install fontconfig freetype harfbuzz icu4c libpng ninja python@3.11
xcode-select --install

Build:

# Build lotio (binary build with zero bundled dependencies)
./scripts/build_binary.sh

Usage

Command Line

lotio --data <input.json> [--output-format <format>] [--output <file>] [--debug] [--layer-overrides <config.json>] [--fonts <dir>] [--text-padding <0.0-1.0>] [--text-measurement-mode <fast|accurate|pixel-perfect>] [--fps <fps>] <output_dir>

Options:

• --data <input.json> - Path to input Lottie animation JSON file (required)
• --output-format <format> - Output format: png (default), raw, ffv1, or mov
• --output <file> - Output file path (use - to stream to stdout, required for raw, ffv1, mov formats)
• --debug - Enable debug output
• --layer-overrides <config.json> - Path to layer overrides JSON (for text auto-fit, dynamic text values, and image path overrides)
• --fonts <dir> - Directory containing font files (.ttf); fonts are looked up here first, then in the JSON directory
• --text-padding <0.0-1.0> - Text padding factor (0.0-1.0, default: 0.97 = 3% padding)
• --text-measurement-mode <fast|accurate|pixel-perfect> - Text measurement mode: fast | accurate | pixel-perfect (default: accurate)
• --fps <fps> - Frames per second for output (optional, default: animation fps or 30)
• --version - Print version information and exit
• --help, -h - Show help message

Output Formats:

• png (default) - PNG frames written to directory (or streamed to stdout with --output -)
• raw - Uncompressed RGBA video file (fastest, largest files, preserves alpha)
• ffv1 - Lossless FFV1 codec in Matroska container (good compression, preserves alpha)
• mov - MOV container with QTRLE codec (fast encoding, preserves alpha, widely compatible)

Examples:

# Render to PNG frames (default)
lotio --data animation.json --fps 30 frames/

# Direct video encoding (no ffmpeg binary needed)
lotio --data animation.json --output-format mov --output video.mov --fps 30
lotio --data animation.json --output-format ffv1 --output video.mkv --fps 30
lotio --data animation.json --output-format raw --output video.rgb --fps 30

# Stream PNG to stdout (for piping to ffmpeg)
lotio --data animation.json --output-format png --output - --fps 30 | ffmpeg -f image2pipe -i - output.mp4

# With layer overrides
lotio --data animation.json --layer-overrides layer-overrides.json --fps 30 frames/

# With custom font directory
lotio --data animation.json --fonts ./fonts --fps 30 frames/

Docker

Quick start:

docker run --rm -v $(pwd):/workspace matrunchyk/lotio:latest \
  --data data.json --fps 30 --layer-overrides layer-overrides.json --output-format mov --output video.mov

Available images:

• matrunchyk/lotio:latest - lotio binary with built-in video encoding support

Multi-platform support: The image supports linux/arm64 and linux/amd64.

See Docker Documentation for detailed usage.

Browser (WebAssembly)

Install from npm:

npm install lotio

Basic Usage:

import Lotio, { State, TextMeasurementMode } from 'lotio';

// Load fonts
const fontResponse = await fetch('./fonts/OpenSans-Bold.ttf');
const fontData = new Uint8Array(await fontResponse.arrayBuffer());

// Load animation
const animationResponse = await fetch('./animation.json');
const animationData = await animationResponse.json();

// Create animation instance
const animation = new Lotio({
  fonts: [{ name: 'OpenSans-Bold', data: fontData }],
  fps: 30,
  animation: animationData,
  layerOverrides: { /* optional layer overrides */ },
  textPadding: 0.97,  // Optional: text padding factor (default: 0.97)
  textMeasurementMode: TextMeasurementMode.ACCURATE,  // Optional: TextMeasurementMode.FAST | TextMeasurementMode.ACCURATE | TextMeasurementMode.PIXEL_PERFECT
  wasmPath: './lotio.wasm'
});

// Event handlers (fluent interface)
animation
  .on('error', (error, anim) => {
    console.error('Animation error:', error);
  })
  .on('loaded', (anim) => {
    console.log('Animation loaded');
    anim.start();
  })
  .on('start', (anim) => {
    console.log('Animation started');
  })
  .on('pause', (anim) => {
    console.log('Animation paused');
  })
  .on('stop', (anim) => {
    console.log('Animation stopped');
  })
  .on('end', (anim) => {
    console.log('Animation ended');
  })
  .on('frame', (frameNumber, time, anim) => {
    // Render to canvas
    const canvas = document.getElementById('canvas');
    anim.renderToCanvas(canvas, '#2a2a2a');
  });

// Control methods
animation
  .setFps(60)           // Change FPS
  .seek(10)             // Seek to frame 10
  .start()              // Start playback
  .pause()              // Pause
  .stop();              // Stop and reset

// Getters
const fps = animation.getFps();
const state = animation.getState(); // 'stopped' | 'paused' | 'loaded' | 'error' | 'playing'
const frame = animation.getCurrentFrame();
const info = animation.getAnimationInfo();

// Render current frame to canvas
const canvas = document.getElementById('canvas');
animation.renderToCanvas(canvas, '#ffffff');

// Cleanup
animation.destroy();

Full Example with Canvas:

<!DOCTYPE html>
<html>
<head>
  <title>Lotio Animation</title>
</head>
<body>
  <canvas id="canvas"></canvas>
  <button id="playBtn">Play</button>
  <button id="pauseBtn">Pause</button>
  <button id="stopBtn">Stop</button>
  
  <script type="module">
    import Lotio from 'lotio';
    
    let animation;
    
    async function init() {
      // Load font
      const fontRes = await fetch('./fonts/OpenSans-Bold.ttf');
      const fontData = new Uint8Array(await fontRes.arrayBuffer());
      
      // Load animation
      const animRes = await fetch('./animation.json');
      const animData = await animRes.json();
      
      // Create animation
      animation = new Lotio({
        fonts: [{ name: 'OpenSans-Bold', data: fontData }],
        fps: 30,
        animation: animData,
        wasmPath: './lotio.wasm'
      });
      
      const canvas = document.getElementById('canvas');
      
      // Render frames
      animation.on('frame', () => {
        animation.renderToCanvas(canvas);
      });
      
      // Controls
      document.getElementById('playBtn').onclick = () => animation.start();
      document.getElementById('pauseBtn').onclick = () => animation.pause();
      document.getElementById('stopBtn').onclick = () => animation.stop();
    }
    
    init();
  </script>
</body>
</html>

Samples

The samples/ directory contains example Lottie animations and configurations:

• samples/sample1/ - Basic animation with layer overrides

  • data.json - Lottie animation file
  • layer-overrides.json - Text and image customization configuration
  • output/ - Rendered frames (run lotio to generate)

• samples/sample2/ - Animation with external images

  • data.json - Lottie animation file with image references
  • images/ - External image assets referenced by the animation
  • output/ - Rendered frames (run lotio to generate)

• samples/sample5/ - Text under track matte (Bodymovin-style, tp omitted); lotio's Skottie build includes a track-matte patch so the matte source is the nearest prior layer with td !== 0 and it renders correctly without editing data.json.

Try the samples:

# Sample 1: Basic animation with text customization
cd samples/sample1
lotio --data data.json --layer-overrides layer-overrides.json --fps 30 output/

# Sample 2: Animation with external images
cd samples/sample2
lotio --data data.json --fps 30 output/

Using as a Library

Headers

Headers are installed at /opt/homebrew/include/lotio/ (or /usr/local/include/lotio/):

#include <lotio/core/animation_setup.h>
#include <lotio/text/text_processor.h>
#include <lotio/utils/logging.h>

Linking

Link with Skia libraries:

g++ -I/opt/homebrew/include -L/opt/homebrew/lib \
    -llotio -lskottie -lskia -lskparagraph -lsksg -lskshaper \
    -lskunicode_icu -lskunicode_core -lskresources -ljsonreader \
    your_app.cpp -o your_app

Or use pkg-config (recommended):

g++ $(pkg-config --cflags --libs lotio) your_app.cpp -o your_app

Using Skia Directly

The lotio package includes Skia headers and libraries, so you can use Skia features directly in your code:

// Use Skia directly
#include <skia/core/SkCanvas.h>
#include <skia/core/SkSurface.h>
#include <skia/modules/skottie/include/Skottie.h>

// Use lotio
#include <lotio/core/animation_setup.h>

int main() {
    // Use Skia API directly
    SkImageInfo info = SkImageInfo::MakeN32(800, 600, kOpaque_SkAlphaType);
    auto surface = SkSurfaces::Raster(info);
    
    // Use lotio functions
    AnimationSetupResult result = setupAndCreateAnimation("input.json", "");
    
    return 0;
}

Compile with:

g++ $(pkg-config --cflags --libs lotio) your_app.cpp -o your_app

The pkg-config file includes all necessary include paths:

• -I${includedir} - Lotio headers
• -I${includedir}/skia - Skia core headers
• -I${includedir}/skia/gen - Skia generated headers

CI/CD Pipeline

The project uses GitHub Actions workflows for automated building, testing, and deployment:

graph TB
    subgraph triggers["Event Triggers"]
        mainPush["Push to main<br/>(creates semver tag)"]
        tagPush["Tag push v*"]
        lotioChanges["Changes to<br/>src/** or Dockerfile.lotio<br/>or build_binary.sh"]
        docsChanges["Changes to<br/>docs/** or examples/**"]
        manual["Manual<br/>workflow_dispatch"]
    end

    subgraph lotioWorkflow["build-lotio.yml<br/>Concurrency: build-lotio<br/>Cancel in-progress: true"]
        buildLotioImg["Build Skia & lotio<br/>push matrunchyk/lotio"]
    end

    subgraph releaseWorkflow["release.yml<br/>Concurrency: release<br/>Cancel in-progress: true"]
        versionTag["Generate version<br/>& create tag"]
        buildMac["Build macOS<br/>binary & dev package"]
        buildWasm["Build WASM<br/>library"]
        buildHomebrew["Build Homebrew<br/>bottle"]
        buildDocs["Build<br/>documentation"]
        publishRelease["Create GitHub<br/>release"]
        
        versionTag --> buildMac
        versionTag --> buildWasm
        versionTag --> buildHomebrew
        buildWasm --> buildDocs
        buildHomebrew --> buildDocs
        buildDocs --> publishRelease
    end

    subgraph testWorkflow["test.yml<br/>Concurrency: test<br/>Cancel in-progress: true"]
        testDocker["Test Docker<br/>image"]
        testWasm["Test JS/WASM<br/>library"]
        testHomebrew["Test Homebrew<br/>package"]
    end

    subgraph pagesWorkflow["pages.yml<br/>Concurrency: pages<br/>Cancel in-progress: false"]
        buildPages["Build & deploy<br/>documentation"]
    end

    mainPush --> lotioWorkflow
    mainPush --> releaseWorkflow
    mainPush --> pagesWorkflow
    
    tagPush --> lotioWorkflow
    tagPush --> releaseWorkflow
    
    lotioChanges --> lotioWorkflow
    docsChanges --> pagesWorkflow
    
    manual --> lotioWorkflow
    manual --> releaseWorkflow
    manual --> testWorkflow
    manual --> pagesWorkflow

    lotioWorkflow -->|Docker image ready| releaseWorkflow
    releaseWorkflow -->|workflow_run<br/>after completion| testWorkflow

    style lotioWorkflow fill:#e1f5ff
    style releaseWorkflow fill:#fff4e1
    style testWorkflow fill:#e8f5e9
    style pagesWorkflow fill:#f3e5f5

Workflow Descriptions

build-lotio.yml - Builds and publishes matrunchyk/lotio Docker image

• Purpose: Create lotio binary Docker image using pre-built Skia base image
• Triggers: Main branch push, tag pushes, source code changes, Dockerfile.lotio changes, build_binary.sh changes, manual dispatch
• Logic: Uses matrunchyk/skia:latest as base image (Skia pre-built), only compiles lotio source
• Build chain: Dockerfile.skia → Dockerfile.lotio (uses pre-built Skia)
• Concurrency: Single instance per workflow (cancels in-progress runs when new one starts)
• Output: matrunchyk/lotio:latest and matrunchyk/lotio:v1.2.3 (multi-platform: arm64, amd64)
• Architecture tags: Also creates -arm64 and -amd64 tags for clarity

release.yml - Builds all release artifacts and creates GitHub release

• Purpose: Build and package all distribution formats (binaries, WASM, Homebrew, docs)
• Triggers: Push to main (creates semver tag automatically), tag pushes (v*), manual dispatch
• Logic:
  • Generates semver version from tag or creates new tag on main push
  • Builds Skia from scratch using build_binary.sh (zero bundled dependencies, fast build)
  • Builds in parallel: macOS, Linux, WASM, Homebrew
  • Injects version into all artifacts
• Concurrency: Single instance per workflow (cancels in-progress runs when new one starts)
• Output: macOS dev package, WASM package, Homebrew bottle, GitHub release

test.yml - Integration tests for all built artifacts

• Purpose: Validate that all release artifacts work correctly
• Triggers: After release.yml completes successfully, manual dispatch
• Tests:
  • Docker image: --help, --version, library functionality, video generation with --debug
  • JS/WASM library: Load, API functions, frame rendering
  • Homebrew package: Installation, --help, --version, basic functionality
• Concurrency: Single instance per workflow (cancels in-progress runs when new one starts)

pages.yml - Builds and deploys documentation to GitHub Pages

• Purpose: Generate and deploy documentation with version injection
• Triggers: Changes to docs, examples, or build scripts; manual dispatch
• Logic: Installs lotio npm package, injects version from git tag
• Concurrency: Single instance per workflow (does not cancel in-progress runs)
• Output: Deployed to GitHub Pages

Project Structure

src/
├── core/          # Core functionality (argument parsing, animation setup, rendering)
├── text/          # Text processing (configuration, font handling, sizing)
└── utils/         # Utilities (logging, string utils, crash handling)

IDE Setup

The project includes IDE configuration for Cursor/VS Code:

• .vscode/c_cpp_properties.json - C/C++ extension settings
• .clangd - clangd language server settings

Reload Cursor/VS Code after cloning: Cmd+Shift+P → "Reload Window"

Troubleshooting

Skia build fails:

• Ensure all dependencies are installed
• Check sufficient disk space (Skia build is large)
• Review error messages in scripts/build_binary.sh output

Linker errors:

• Verify Skia libraries exist in third_party/skia/skia/out/Release/
• Check library paths in build script

IDE include errors:

• Reload Cursor/VS Code
• Verify .vscode/c_cpp_properties.json has correct paths

License

See individual component licenses:

• Skia: third_party/skia/skia/LICENSE