• All pages are exported as Markdown to a temporary location.
• An index is generated for the pages that will be published and have a menu entry.
• The links are fixed in the generated markdown pages.
• Templates from the selected layout are preprocessed and variables for translation are identified.
• Each page that is selected for publishing is processed with an appropriate processor. Processors generate new attributes and add them to the page extra attributes.
• The index is added to the extra attributes of each page.
• The generated extra attributes are added to the beginning of the generated Markdown files as YAML blocks.
• The Markdown files are processed by Pandoc using the template appropriate for each page.
• Links are extracted from the generated HTML files to find the resources required by the generated site.
• The HTML files and the discovered resources are copied to the final directory.