maxitg
English linter
The problem
In addition to the Markdown linter which would check the formatting/spacing/empty lines in Markdown, we can have an English linter, which is going to spot issues with the language itself.
Possible solution
It could be based on something like Grammarly. We need to keep it reasonably conservative so that there are almost no false positives. But we can probably check for misspellings (while keeping a special terms dictionary) and maybe some grammar without too much risk.
@maxitg I'm going to use mistune for parsing markdown and language_tool_python for linting. It looks like markdown AST will have to be modified a bit in order to store line number metadata. Spelling looks like it won't be an issue, but grammar will be quite difficult, especially when code blocks and links are in the middle of sentences. However, it may be possible to piece together these tokens to form coherent, lintable sentences. For now, I'll limit it to just spelling and in the future grammar could be supported.
I'll make a repo for this subproject (and mistune fork) on my account and invite you as a collaborator, since linting english in markdown is not really specific to this project. Does this sound reasonable to you?
