CalcifyArchive

Tool for creating a single page static site for hosting articles and blog posts, built on Typescript and Python, mimicking a desktop file browser.

Example

jtpond.github.io

Requirements

• Python (> 3.6)
• commonmark module
• Typescript compiler

How To

1. Clone the repository
2. In the static subdirectory, run either build.sh or build.bat depending on your system.
3. In the root directory, run python(3) indexer.py <path/to/dir/to/be/archived>. Get help: python(3) indexer.py -h.
4. Modify ./dist/index.json for your settings: 1. Set any pinned articles pinned field to true. 2. Update the f_order and a_order field of any folders that have specific orders for how they should appear in the client.
5. Deploy: 1. Point a webserver at the ./dist/ directory, or 2. Commit ./dist/ to GitHub pages, or 3. Sync ./dist/ to another static CDS like S3+CloudFront

Archive format

• Each article is an HTML, or markdown file.
• For HTML there's no need to add a head, or the body tags.
  • You have two options:
    • parent_dir/article_name.html, or parent_dir/article_name/article.html with a companion parent_dir/article_name/assets/ which can contain anything your article references.
      • If you use the second option, always use src="./assets/blah.blah", the indexer will replace all of these patterns with src="./archive/path/to/assets/blah.blah".
• For markdown you can only use the second option and parent_dir/article_name/article.html will be created from parent_dir/article_name/article.md using commonmark. The assets directory is the same.

Re-indexing

Re-running the indexer will update any articles' html form and their modified time and update dist/.

If you've modified any article.html files that were produced by the indexer such that their modified time is greater than their markdown version, the indexer will not overwrite them. If you modify the markdown file, though, the html will get overwritten.

Other Content

For content other than basic md/html content you can create an article_name/article.html where the article is:

  <object height="100%" width="100%" type="content/type" data="./assets/other.content">
    Not Supported
  </object>

and the indexer will update the data field and the object will load naturally in the client. This works well with Jupyter Notebooks exported as html.