SVG to Raster Conversion

Udility Diffuser leverages CairoSVG to transform generated vector graphics (SVG) into high-quality raster images (PNG). This conversion process ensures that the illustrative diagrams created by the model are easily viewable, shareable, and compatible with standard image processing tools.

While the conversion happens automatically within the generate_image_from_text workflow, the library provides a standalone utility for users who wish to convert raw SVG strings manually.

The `svg_to_png` Function

The svg_to_png function handles the byte-level conversion of SVG code into a portable PNG file.

Usage

from Udility.diffuser import svg_to_png

# Example SVG string
svg_data = '<svg height="100" width="100"><circle cx="50" cy="50" r="40" fill="red" /></svg>'

# Convert to PNG
svg_to_png(svg_data, output_filename='my_illustration.png')

API Reference

| Parameter | Type | Description | | :--- | :--- | :--- | | svg_code | str | The raw SVG XML string to be converted. | | output_filename | str | (Optional) The target file path for the generated PNG. Defaults to 'output.png'. |

Returns:

This function writes the file to the local disk and does not return a value.

Backend Dependencies

The conversion process relies on the following engine components, which are bundled with the Udility installation:

CairoSVG: The primary converter used to parse SVG and render it via Cairo.
cairocffi: A CFFI-based drop-in replacement for Pycairo, used for graphics rendering.
Pillow (PIL): Used by the diffuser for post-conversion image handling and display.

Error Handling

If the provided svg_code is malformed or if there is a system-level issue with the Cairo backend, the function will raise a RuntimeError. Ensure that your environment has the necessary system dependencies for Cairo if you are running on a custom Linux distribution (though standard environments like Google Colab and Ubuntu typically work out-of-the-box).

SVG to Raster Conversion