🎵 lastfm-collage-generator

A Python library to generate image collages from a Last.fm user's listening history.

What it does

lastfm-collage-generator fetches top listening data from Last.fm and composites the artwork into a single image grid.

• Albums, Artists & Tracks: Create square or rectangular grids from 1x1 up to 20x20.
• Artist Image Scraping: Because the Last.fm API no longer provides artist photos, the library fetches them directly from artist web pages.
• Themes & Overlays: Choose from built-in themes (dark, light, sunset, neon, glassmorphic) and overlay styles (banner, pill, full_tint, gradient), or disable text for a clean grid.
• Social Media Presets: Ready-made dimension presets for Instagram Stories, Twitter/X headers, and wallpapers with blurred letterbox backdrops.
• Visual Filters & Effects: Apply duotone or custom post-processing filters across the whole collage.
• Local SQLite Caching: Automatically caches downloaded artwork to ~/.cache/lastfm-collage/ to speed up subsequent runs and avoid rate limits.
• Sync & Async: Standard generate() and async generate_async() for integration into scripts, web services, or Discord bots.

📦 Installation

pip install lastfmcollagegenerator

Or with uv:

uv add lastfmcollagegenerator

Note: You need a free Last.fm API key and secret. You can create one on the Last.fm API account page.

⚡ Quickstart

from lastfmcollagegenerator import CollageGenerator

# 1. Initialize with your Last.fm API credentials
generator = CollageGenerator(
    lastfm_api_key="YOUR_API_KEY",
    lastfm_api_secret="YOUR_API_SECRET",
)

# 2. Generate a 3x3 album collage for the past 7 days
image = generator.generate(
    entity="album",
    username="your_username",
    cols=3,
    rows=3,
    period="7day",
)

# 3. Save the resulting PIL Image
image.save("weekly_collage.png")

💡 Usage & Customization Examples

1. Artists and Tracks

# 4x4 top artists collage over the last 3 months
artist_collage = generator.generate(
    entity="artist",
    username="your_username",
    cols=4,
    rows=4,
    period="3month",
)

# Top tracks of all time using convenience method
track_collage = generator.generate_top_tracks_collage(
    username="your_username",
    cols=3,
    rows=3,
    period="overall",
)

2. Social Media Presets & Wallpapers

Presets automatically configure grid dimensions, aspect ratios, and letterbox backdrops (Gaussian-blurred from your top artwork):

# Instagram Story (1080x1920, 9:16)
story = generator.generate(
    entity="album",
    username="your_username",
    cols=3,
    rows=5,
    preset="instagram-story",
)
story.save("story.png")

Available presets: instagram-story, instagram-post, twitter-header, desktop-wallpaper, desktop-wallpaper-4k.

3. Tile Geometry (Borders, Rounded Corners & Spacing)

# Rounded tiles with custom spacing and borders
collage = generator.generate(
    entity="album",
    username="your_username",
    cols=3,
    rows=3,
    period="1month",
    corner_radius=18,
    spacing=12,
    border_width=3,
    border_color="#FF5A5F",
)

4. Built-in Themes & Overlay Styles

# Sunset theme with compact pill badges
sunset_collage = generator.generate(
    entity="album",
    username="your_username",
    cols=3,
    rows=3,
    theme="sunset",
    overlay_style="pill",
)

# Neon theme with full tint overlay
neon_collage = generator.generate(
    entity="artist",
    username="your_username",
    cols=3,
    rows=3,
    theme="neon",
    overlay_style="full_tint",
)

# Clean mode (pure album covers without text or overlay banners)
clean_collage = generator.generate(
    entity="album",
    username="your_username",
    cols=4,
    rows=4,
    show_text=False,
)

Sunset + Pill	Neon + Full Tint	Clean Mode

5. Custom Themes

Define your own color palette with the Theme class:

from lastfmcollagegenerator import CollageGenerator, Theme

forest_theme = Theme(
    name="forest",
    overlay_bg=(20, 50, 30, 190),      # RGBA banner background
    text_color=(235, 255, 235),        # RGB text color
    accent_color=(100, 200, 120, 220), # Accent chip/pill color
)

custom_collage = generator.generate(
    entity="album",
    username="your_username",
    cols=3,
    rows=3,
    theme=forest_theme,
    overlay_style="banner",
)

6. Visual Filters & Duotone Effects

Apply post-processing filters across the generated collage:

from lastfmcollagegenerator import CollageGenerator, DuotoneFilter

# Map image luminance to a two-color duotone palette
duotone = DuotoneFilter(
    black_color="#0F0C29",
    white_color="#FF007F",
)

duotone_collage = generator.generate(
    entity="album",
    username="your_username",
    cols=3,
    rows=3,
    filters=duotone,
)

7. Async Generation (Discord Bots / Web APIs)

import asyncio
from lastfmcollagegenerator import CollageGenerator

async def main():
    generator = CollageGenerator("YOUR_API_KEY", "YOUR_API_SECRET")
    image = await generator.generate_async(
        entity="album",
        username="your_username",
        cols=3,
        rows=3,
        period="7day",
    )
    image.save("async_collage.png")

asyncio.run(main())

8. Image Export & In-Memory Streaming

Use export_image() for file export or stream directly to in-memory buffers for web frameworks:

import io
from lastfmcollagegenerator import export_image

# 1. Export to modern WebP (optimized quality)
export_image(image, "output/collage.webp", quality=85)

# 2. Export to JPEG (automatically flattens alpha onto black background)
export_image(image, "output/collage.jpg", quality=90)

# 3. Stream in-memory for FastAPI / Flask / Discord bot response
buffer = io.BytesIO()
image.save(buffer, format="PNG")
buffer.seek(0)
raw_png_bytes = buffer.getvalue()

📖 Parameters Reference (generate)

🤖 GitHub Actions

The repository includes an action.yml to generate collages automatically within GitHub Actions workflows (e.g. to update a profile README):

- uses: paurieraf/lastfm-collage-generator@v1.3.2
  with:
    username: ${{ secrets.LASTFM_USERNAME }}
    api-key: ${{ secrets.LASTFM_API_KEY }}
    api-secret: ${{ secrets.LASTFM_API_SECRET }}
    output-path: weekly-recap.png

🤝 Contributing

Contributions are welcome! See CONTRIBUTING.md for local development setup, testing workflows, and guidelines.

📄 License

This project is licensed under the MIT License.