Overview of the Markdown Indexer architecture
The Markdown Indexer is a tool that parses Markdown source files directly to build Algolia search indices, bypassing the need for HTML crawlers. This approach ensures that index updates are synchronized with deployments (via GitHub Actions), remains unaffected by UI/theme changes, and provides full control over the index structure.
The core engine follows a five-step process:
- Scans the document directory for all Markdown files.
- Parses each file to extract frontmatter and body content.
- Invokes processors to generate index data.
- Groups data by index name.
- Performs a batch upload to Algolia.