It seems to me there is a fairly big difference in how the Python port handles MarkdownIt.__init__ parameters compared to the JS implementation, that isn't documented in port.yaml. Two major differences that I see:

1. A setting mapping in JS init is options. In Python implementation the mapping is a similar structure that JS passes to .configure(), that is it is the "preset" structure that contains options and components.
2. When passed in options, Javascript always merges the options with a set of preset values. In Python, passing in a settings dict will skip using any preset values at all.

As a result, in Javascript I can do the following to get "default" presets and make a change in one setting:

var md = markdownit({ html: false });

In Python the same would be more like:

from markdown_it import presets conf = presets.default.make() conf["options"]["html"] = False md = MarkdownIt(conf)

Is this difference in the API intended? If yes, should we document it in port.yaml? If no, maybe we should redesign the init function before v1.0.0?