Markdown API

1. About
2. How it works
3. Installation
4. Configuration
5. Usage

About

Markdown API is a JavaScript command line interface to build and manage (JSON) API endpoints using JavaScript and Markdown files to describe resource contents.

It provides a developer experience in content management using an advanced editor and/or the command line interface, as an alternative to the CMS experience using the browser and a WYSIWYG editor, and bound to multiple dependencies like PHP or Ruby, MySQL or MongoDB, etc…

It also provides:

• endpoints for resources lists (indexes)
• a low level JavaScript API to create, read, update, and delete endpoints (headless CMS)
• a small build time, by writing or removing only what needs to be created, updated, or removed
• (opt-in) long term caching of endpoints via a hash in endpoints filenames
• (opt-in) defining fields, relations, and files contents using schemas for each resource type
• (opt-in) batching CRUD operations, to avoid building after each individual operation

How it works?

Each resource entry should be contained in its own directory (its name will act as its identifier), contained itself in a directory named after the resource type, eg. posts, and each resource entry has 3 mandatory files:

src
  └─ posts
    └─ my-post-identifier-1
      └─ content.md
      └─ excerpt.md
      └─ index.js
      └─ static/
          └─ …
    └─ my-post-identifier-2
      └─ …
    └─ …
  └─ …

• content.md: the main content
• excerpt.md: the content used in indexes
• index.js: the meta data (also used in indexes)

The optional static directory may contain static files referenced in content.md or excerpt.md. It will be copied into the dist folder. jpg and png files will be processed by imagemin and imagemin-webp.

Note: later (see TODO) custom Markdown (default to markedjs) and files processors can be defined via a renderers property in the optional configuration file.

index.js:

module.exports = {
  date: Number,      // Note: format should be YYYYMMDD
  draft: Boolean,    // Note: if `true`, the corresponding endpoint will not be built
  categories: Array, // Note: values will be slugified in indexes endpoints paths
  slug: String,      // Note: not used yet (cf. TODO)
  title: String,
}

The resource will be indexed in resources lists named after the specified categories.

Note: later, loading schema-s (see TODO and lib/schema.js) via the CLI --schema option or the schemas property in the optional configuration file, will provide a way to define relations for each resource type with other resource type, as well as its fields and their association to specific file names.

With the following source files tree…

src
  └─ posts
    └─ entry-1
      └─ content.md
      └─ excerpt.md
      └─ index.js
      └─ static
        └─ img.jpg
    └─ entry-2
      └─ …

… the api build command will create the following distribution files tree:

dist
  └─ api
    └─ categories
      └─ posts
        └─ all
          └─ 1/index.json   // List of n first posts
          └─ 2/index.json   // List of 2n first posts
        └─ category-1
          └─ 1/index.json   // List of n first posts from "category-1"
          └─ 2/index.json   // List of 2n first posts from "category-1"
        └─ index.json       // List of categories
    └─ posts
      └─ entry-1/index.json
      └─ entry-2/index.json
  └─ static
    └─ posts
      └─ entry-1
        └─ img.jpg
        └─ img.webp

Installation

  npm i @cdoublev/markdown-api

Markdown API is tested with current NodeJS LTS and latest versions.

Configuration

A configuration file named api.config.js can be used to define default values for all command arguments except the --name, -n and --type, -t, which are required to update or remove a resource entry.

Its path will be resolved to the directory where the command is run. Each configuration value might be a function returning the expected value.

module.exports = {
  categories: [String],
  content: String,
  date: Number, // YYYYMMDD
  dist: String,
  entitiesPerPage: Number,
  excerpt: String,
  hash: Boolean,
  src: String,
  subVersion: Boolean,
  title: String,
}

Usage

  node_modules/.bin/api -h

• api add
• api set
• api remove
• api get
• api build
• api stats

api add

api add scaffolds a new resource entry.

Parameters:

  -t, --type <type>              type for the new entry (required)
  -s, --src <src>                path to sources directory (required) (default: "./src/api")
  -T, --title <title>            title for the new entry (required) (default: "New post")
  -c, --content [content]        content for the new entry (default: "Introduction\n\n## Chapter 1\n\n## Chapter 2\n\n## Chapter 3\n\n## Conclusion")
  -e, --excerpt [excerpt]        excerpt for the new entry (default: "TODO: excerpt.")
  -C, --categories [categories]  categories for the new entry (default: [])
  -D, --date [date]              date for the new entry (default: today with format yyyddmm)

  api add -t posts -T "My new post title" -C announcements,general -D 20190101

api set

api set sets the title, excerpt, content, categories, and/or the date of a resource entry.

Parameters:

  -t, --type <type>              type of the entry to set (required)
  -s, --src <src>                path to sources directory (required) (default: "./src/api")
  -n, --name <name>              name of the directory containing the sources files (required)
  -T, --title [title]            new title for the entry
  -c, --content [content]        new content for the entry
  -e, --excerpt [excerpt]        new excerpt for the entry
  -C, --categories [categories]  new categories for the entry
  -D, --date [date]              date for the new entry

  api set -t posts -n my-new-post-title -T "My new blog post"

api remove

api remove remove a resource entry.

Parameters:

  -t, --type <type>  type of the entry to remove (required)
  -s, --src <src>    path to sources directory (required) (default: "./src/api")
  -d, --dist <dist>  path to distribution directory (required) (default: "./dist/api")
  -n, --name <name>  name of the directory containing the sources files (required)

  api remove -t posts -n my-new-blog-post

api get

(TODO) api get outputs either raw or processed contents of a single or multiple resource entries, optionally using filters.

It's a kind of search interface on top of a low level JavaScript API made to handle client requests to GET (read) specific resources.

api build

api build creates/removes/updates endpoints of new/removed/updated resources.

Parameters:

  -s, --src <src>          path to sources directory (required) (default: "./src/api")
  -d, --dist <dist>        path to distribution directory (required) (default: "./dist/api")
  -f, --force              build without checking if sources have been updated
  -p, --entities-per-page  entities per (index) page (default: 10)
  -h, --hash               create endpoints using hashes for long term cache (default: false)
  -S, --sub-version        keep previous generated (JSON) endpoints (default: false)
  -w, --watch              automatically build on change

Check doc/long-term-cache to learn more about config.hash, long term caching of endpoints, config.subVersion, and preserving endpoints versions between build updates.

--no-hash is ignored with --sub-version.

api stats

(TODO) api stats outputs some resources statistics.

TODO

• Feature: api get (define and implement behavior)
• Feature: api stats (define and implement behavior)
• Feature: -t, --type option for api build (to only build endpoints of the corresponding resources type(s))
• Feature: --no-index option for api-build to opt-out from building indexes (and just let entities be consumed by an aggregator like GraphQL)
• Feature: define resource endpoints paths using its slug instead of its name
• Feature: schema data type and command line option to define resources types, files, fields, and relations
• Feature: hook to process content with any Markdown flavor (default to marked)
• Feature: hook to process static files (default to imagemin and imagemin-webp for jpg|png)
• Feature: batch operations