metadoc / metadoc.js Goto Github PK

Metadoc is great for producing API metadata, but there is currently nothing for generating guides. Guides are almost always static HTML sites (not just a page). It's possible to identify the structure of an individual document by it's headings (like the <h1> tag), but there isn't much for producing a master index/table of contents.

Metadoc should be capable of producing indexes that can be used to generate a table of contents for a guide.

This may require some inspection of directory/folder structures to determine things like chapters, subchapters, or other types of volumes. Metadoc may also need to parse the markdown/HTML content for comments to assist with this. Finally, it should be capable of verifying links between documents.

The following post-processors should be created as add-on modules:

• NGN Code Generator (w/ JET View Registries)
• Markdown processing
  • GFM
  • Table Support
  • Mermaid Support
  • MathJAX
• MDN link creation for common JS elements.
• Lunr search indexing (Content Only)
• Github issue references (autolink).
• StackOverflow references (autolink)
• Code Embedding
  • Fenix
  • Codepen
  • JSFiddle
  • RunKit

• Sanity Checks

• Classes

  • Inheritance
  • Namespacing

• Methods

• Properties

• Configurations

• Callbacks

• Events

• Event Bus (NGN)

• Namespaces

• Tags

  • @author
  • @cfg
  • @cfgproperty
  • @Class
  • @constructor
  • @exception
  • @extends
  • @fires
  • @hidden
  • @ignore
  • @info
  • @method
  • @namespace
  • @param
  • @Private
  • @Property
  • @readonly
  • @return
  • @todo
  • @writeonly
  • @Protected
  • @deprecated
  • @experimental
  • @warning
  • @hidden
  • @singleton
  • @interface
  • @static
  • @SInCE
  • @typedef

MD processing should read description attributes, converting any markdown into static HTML.

• GFM
• Table Support
• Mermaid Support
• MathJAX

Currently, cherow (AST parser) will silently fail if there is a syntax error in a file. It ultimately creates a memory leak and throws a heap error if it cannot parse a file. Files with incorrect syntax should not fail silently.

To resolve this, metadoc needs to lint the code before running it through the parser. This should guarantee the code is valid JavaScript before cherow parses the AST.

is the project metadoc still running on the RnD phase or has it's been shut down already?
The website is offline too.

To support fast searching, the primary content needs to be indexed by Lunr. The result should be one or more index files that can be used to seed a search system on a web page.