fix: parse markdown header more carefully

[dandhlee]

There are some files that don't have a header for markdown files and some headers that were wrongly extracted. Also adding support for alternate header type in markdown, see https://www.markdownguide.org/basic-syntax/#headings.

• For the traditional markdown header using # for h1 header, check that:

  • There's only one # for an h1 header
  • There's exactly one space after one #
  • Is not just a short string consisting only of #

• For the alternate markdown header using = characters below the header, check that:

  • The header above the = divider isn't empty
  • The divider line only consists of = and whitespace characters if necessary

Otherwise, I've defaulted to returning the file name if no header can be found. I'll leave it to the library to better format the markdown files rather than dealing with individual formatting issues on the plugin.

Updated unit tests.

Fixes #110.

• Tests pass

[parthea]

Added minor observations, otherwise LGTM.

[parthea]

This line can be removed if mdfile = open('tests/markdown_example_h2.md', 'r') is changed to a with statement. For example, with open('tests/markdown_example_h2.md', 'r') as mdfile:

[dandhlee]

Thank you for the tip! Updating the files.

[parthea]

If we're looking for both header and a space, we could change the check to if "# " in header_line: to look for a header and a space

[dandhlee]

There is an additional overhead to ensure that it doesn't parse h2 header if given ## but it does reduce needing to check for a space followed by #. The updated code looks easier to follow, I think.

[dandhlee]

Updated function name from is_markdown_header to parse_markdown_header to be more in line with what it returns, which isn't a boolean.