added data to app breadcrumb

[kla7]

Fixes #155 .

[keighrim]

Some general suggestions;

• why keeping those first two commits that are simply merging main into the working branch, instead of moving the beginning of the working branch to the latest commit on the main ?
• the role of app_page_data.py is kind of duplicate of this liquid-based loop, and I wonder if there's an efficient way to de-duplicate those efforts
  

  apps/docs/index.md

  Lines 17 to 31 in d53ad4e

  	{% for app in site.data.app-index %}
  	{%- assign check = app[0] | split:'http://apps.clams.ai/' -%}
  	{% if check.size == 2 %}
  	### {{ check[1] }}
  	{{ app[1]["description"] }}
  	{% for version in app[1]["versions"] %}
  	* [{{ version[0] }}]({{ check[1] }}/{{ version[0] }}) ([`@{{ version[1] }}`](https://github.com/{{ version[1] }}))
  	{% endfor %}
  	{% else %}
  	### {{ app[0] }}
  	{% for version in app[1] %}
  	[{{ version }}]({{ app[0] }}/{{ version }})
  	{% endfor %}
  	{% endif %}
  	{% endfor %}

• the vXXX part of the breadcrumbs sill doesn't work. I look at the breadcrumb implementation in the theme source, only to find that it's not easy to fix the issue without changing the theme source code (and for that, I forked the theme source to clamsproject org)
  • 99e2ac4 changes the theme to our custom MM theme repo (https://github.com/clamsproject/website-theme-minimal-mistakes) that includes fix for the middle items in the breancrumb bar.
• lastly, I'm thinking of integrating the script in this PR to the website building GH actions script, so that the individual index.md files under app subdirs will be automatically re-generated every time the website is built by the GH. To that end, I don't think we need to include those files in this PR at the end of the day.

[keighrim]

A few issues with this line;

1. it "hard-codes" '/' as the path separator, and this line will probably break the program on windows.
2. this assumes the "user" of this script runs the script from REPO_ROOT/scripts directory (i.e., their working directory is REPO_ROOT/scripts) but this isn't always guaranteed or forced. You can make directory paths more "portable" (i.e., resilient to the users' working directory at the runtime) by using __file__ special variable automatically set for every .py file by python interpreter
3. the variable name directory is too generic. I'd prefer verbose but understandable variable names to concise but almost meaningless names.

So, taking all above in, this line can be, instead, something like this;

apps_dir = os.path.join(os.path.dirname(__file__), '..', 'docs', '_apps')

(as a practice, think about why we need two "go-up" operations (once with dirname, again with '..')

[keighrim]

This file seems to use CRLF. Use linux style line breaks. (https://en.wikipedia.org/wiki/Newline)

[keighrim]

this will not work, example;

>>> sorted('1.0.1 1.0.2 1.0.10'.split())
['1.0.1', '1.0.10', '1.0.2']

Probably the easiest and safest way to use submission time metadata pulled from each app-version subdir.

[keighrim]

Instead of hard-coding the public website address here (http://apps.clams.ai/), replace with jekyll {{ url }} variable and defer the substitution to jekyll engine. This way the root URL will be automatically adjusted when developing site locally.