Today I was studying the notebooks that published along with the book AI Powered Search and realized that author was using the “fence code block” quite a lot. We all know the jupyter notebook comes with the markdown which is a content focused language. To be content focused, it means the users use the precanned markdown syntax to explicitly set the relationship between contexts, like what is a header, what is a paragraph. In that way, you don’t have to worry about the formatting yourself, as there is a fixed way and consistent way of doing it. If you ever failed to add in a new line in your content, you will understand that this “inability of adding a new line” is a feature and benefit that comes with Markdown (just like latex) instead of a bug (“this sucks, I cannot even add in a new line”).

Attached are several markdowns of using the fenced code blocks.

Can you guess what will be the output after it is rendered?

Even if there seems like two new lines between L1 and L4. Due to the very reason mentioned above, Markdown will treat as two paragraphs and only insert a new line. And the fenced code block will be treated as inline, if the opening triple backticks are on its own line, like L11, it will be formatted as a standalone code block with a different formatting. Inline code blocks comes with greyed background while standalone comes with a white background but indented differently. In addition, when you have fenced code block with a programming language next to the open backticks, the text within the code block will be highlighted per the syntax of that programming language.