Thanks for creating this extension. At Pydantic, we have a similar custom implementation to create tabs, but it has been a source of bugs and I wanted to move away from this approach.
Unfortunately, it seems like the markdown implementation has to be done via regex, and as you probably experienced the Sphinx/docutils approach is easier to reason about.
I want to switch from our current implementation to this extension. However, we make use of mkdocs-material annotations. Looking at the implementation of annotations, it seems to be a mkdocs-material specific feature implemented in JS, and what it does is it replaces lists with annotations:
```python
a = 1 # (1)!
```
1. I will be replaced when the page is rendered
This is imo a poor design choice, but there isn't much we can do.
Do you think the regex logic of the markdown_ext.py module could be updated to support annotations? Here are some valid examples, for reference:
```python
a = 1 # (1)!
b = 2 # (2)!
```
1. An arbitrary amount of new lines can be used between the code block and the first element of the list.
2. Same for any following list element.
I can also span multiple lines.
=== "Tab 1"
```python
a: int # (1)!
```
1. Annotation in tab. Watch out for the indentation.
=== "Tab 2"
```python
a: int
```Note that the new block API could be used, as this avoids headaches with indentation.
Thanks for creating this extension. At Pydantic, we have a similar custom implementation to create tabs, but it has been a source of bugs and I wanted to move away from this approach.
Unfortunately, it seems like the markdown implementation has to be done via regex, and as you probably experienced the Sphinx/docutils approach is easier to reason about.
I want to switch from our current implementation to this extension. However, we make use of
mkdocs-materialannotations. Looking at the implementation of annotations, it seems to be amkdocs-materialspecific feature implemented in JS, and what it does is it replaces lists with annotations:This is imo a poor design choice, but there isn't much we can do.
Do you think the regex logic of the
markdown_ext.pymodule could be updated to support annotations? Here are some valid examples, for reference:Note that the new block API could be used, as this avoids headaches with indentation.