Markdown Blog Post Import / Sync - Serve the blog post content from a markdown directory

Inizia leggendo l’entità BlogPost e l’entry point NCronJob.IJob, quindi esamina appsettings.json e l’area delle migrazioni di UpgradeAssistent. L’issue richiede inoltre di documentare l’interfaccia markdown. Il lavoro è considerato completato quando l’importazione proposta da una directory piatta, la gestione dei metadati, il tracciamento di ExternalId, la configurazione, la migrazione e la documentazione sono tutti definiti e implementati.

The current state of the blog is: Users write Markdown inside the blog post and the blog renders this to the enduser to HTML.

An obvious automation for that is that we introduce a background job that monitors (via NCronJob.IJob) a directory/repository/... and automatically imports them into the blog. That allows the blog post as a vehicle, but authoring/versioning happens via GitHub for example.

The question is, how should we get certain metadata:

• Via folder structure
• One markdown with a header section
• A Mix of both?

My current way of thinking is the second option, as we also need some way of having a unique ID in the markdown file in case users change the title and we can't associated markdown file to the blog post in the blog itself. It might look something like:

----------
id: some-string-that-has-to-be-unique
title: Here comes the title
tags: tag 1, tag 2
image: https://some-image.url/image.webp
fallbackimage: https://some-image.url/fallback-image.jpg
published: true
----------
The first block is the Short Description that allows arbitrary Markdown
----------
The second block is the Content that allows arbitrary Markdown

The "interface" has to be well documented in our docs. Our BlogPost entity needs the new ExternalId as property to keep track of the source.

The configuration of the target will happen in the appsettings.json file and we need a new migration for the UpgradeAssistent. We may want a new section like:

{
  "markdownImport": {
    "sourceType": "GitHub",
    "url": "https://github.com/...."
  }
}

For starters we can begin with a flat directory as this doesn't need any additional packages and knowledge about the source. So for example url would be https://github.com/linkdotnet/blog-source/main/md/ and in there would be all files, which we would parse from top to bottom in alphatical order.