Closed Matt-Scheetz closed 1 month ago
There is not much we can do. Their Markdown output is invalid CommonMark, as 4 spaces will always result in a code block as per the spec here.
You can try it here on GitHub and it will be the same:
- *Description*: too bad
- Hello world
Will give you this, similar to markdig:
- *Description*: too bad
- Hello world
Rendering LLM output (gpt-4o) and I've observed the contents after the header for a numbered list item is getting placed in a code block (See content after 10):![image](https://github.com/xoofx/markdig/assets/120581026/922a237d-136c-4da9-944c-0740276bfe1b)
Looking at the raw text and I see that there is an extra space between the Numbered header and the content![image](https://github.com/xoofx/markdig/assets/120581026/8acf838c-6fb9-40b4-bd3a-7cce8b741c1d)
As the images states there are 4 spaces after 10 and 3 spaces after all other numbered headers.
When I pasted the raw text to have it rendered and remove the extra space, the text renders without the code block:![image](https://github.com/xoofx/markdig/assets/120581026/b52beae0-2980-46ff-9fd9-0ebdb2508eb0)
Using Markdig v 0.36.2