Detect intraword emphasis in Markdown

[not-my-profile]

It would be neat if typos could understand intraword emphasis in Markdown.

$ echo '`su` comes from **s**uper **u**ser.' > foo.md
$ typos foo.md
error: `uper` should be `super`, `upper`
  --> foo.md:1:22
  |
1 | `su` comes from **s**uper **u**ser.
  |                      ^^^^
  |
error: `ser` should be `set`
  --> foo.md:1:32
  |
1 | `su` comes from **s**uper **u**ser.
  |                                ^^^
  |

[epage]

This would require full knowledge of the languages involved to fix

• This could be in a doc comment in a rust file
• We would need to process code fences, mapping their infostring to a file type, and process that section accordingly

Currently, our word extraction has no context like what the file type is. We do handle some common cases of syntax, but only in broad terms. For example, we ignore the word right after a \ because we don't know whether \tword is \t and word or \ and tword. Applying the same concept here, I'm not sure of a good rule of thumb that wouldn't cause us to miss checking a good swath of code.