Markdown magic cell transformation

This document describes how the Snowpark Migration Accelerator (SMA) handles the transformation of Markdown magic cells during notebook migration.

Magic Markdown cell transformation

When the SMA processes a notebook and detects a magic cell that begins with %md, it automatically transforms the cell into a standard Jupyter notebook (.ipynb) markdown cell.

How it works

In Databricks notebooks, Markdown content is commonly written using magic commands:

%md
# My Documentation
This is a **markdown** cell with formatted text.
Copy

During migration, the SMA recognizes this pattern and converts it into a native notebook cell with the cell metadata set to "markdown". This ensures that:

  • The content is properly recognized as documentation/markdown in the target environment.

  • Markdown rendering is correctly applied.

  • The notebook maintains its intended documentation structure.

Before migration (Databricks)

A cell with the %md magic command in the notebook JSON structure:

{
  "cell_type": "code",
  "source": [
    "%md\n",
    "# Customer Analysis\n",
    "This notebook performs **customer segmentation** analysis."
  ],
  "metadata": {},
  "outputs": []
}
Copy

After migration (Snowflake)

The same content is converted to a notebook cell with the language metadata set to markdown:

{
  "cell_type": "code",
  "source": [
    "# Customer Analysis\n",
    "This notebook performs **customer segmentation** analysis."
  ],
  "metadata": {
    "language": "markdown"
  },
  "outputs": []
}
Copy

Note that the %md magic command is removed from the source, and the cell metadata now includes "language": "markdown" to indicate the cell contains documentation content.

Benefits

  • Native markdown support: The migrated notebook uses native markdown cell types instead of magic commands.

  • Better rendering: Markdown cells are properly rendered in notebook environments without requiring code execution.

  • Cleaner structure: Removal of magic command prefixes results in cleaner, more portable documentation cells.

Language: English