Microdown Parser and Elements

I'm a parser and object model for Microdown originally implemented by S. Ducasse, L. Dargaud and G. Polito. The implementation is based on the work on markdown of K. Osterbye. Further developments by S. Ducasse and K. Osterbye.

Microdown is a smaller markdown but it is more extensible. It contains a nice builder and some visitors. Microdown is now the default markup for the Pillar document compilation chain.

Why should you use Microdown?

Microdown is a smaller markdown but it is more extensible. It is used to produce books, slides, websites, doc. It can be read on GitHub but also on Pharo itself using the DocumentationBrowser

It supports

• Header
• Lists
• Paragraph
• Code with arguments: key=val&key2=val2 ...

But also

• Environment with arguments
• Table
• Anchor
• Annotated paragraph
• Math with arguments
• and more intra-block elements such as extensible markups, raw, math, and references.

Core Syntax in 2 seconds

   	# Header
        @anchor
	% This is a line comment

	```language=Pharo&caption=Beautiful&label=Fig1
   	code
	```
   
   	![Pharo is cool .%anchor=fig:pharo&width=80](http://pharo.org)
	
   	- list
   	1. ordered list 

  	`in text` and for Pharo hyperlinks to class, method and package: 
  	`Point`, `Point class`, `Point>>#setX:setY:`, `#’Microdown-Tests’ (for packages)

  	References: *@ref@*
	

Full Syntax

Headers

Similar to markdown headers are composed of # space text on one line. The headers can be from 1 to 6.

### Header Level3

There is no support for other forms of declaration.

Anchors

In Microdown we can define anchors and reference to them (see References). There are three ways to create anchors:

• @anchorlabel will create an anchor to the preceding element.
• Figures, mathematical environments, and environment can specify labels as arguments (anchor)
• Code block can specify label as argument (argument named anchor)

Various

• % comments
• *** horizontal line
• File metadata is plain JSON

{
"date" : "12 december 2025"
}

• Raw text

{{ raw text }}

Math support

• $$ mathematical environment with a label for easy referencing.

$$ %label=refToTheGreatEquation
V_i = C_0 - C_3 
$$

• Math in text

'abc$	V_i = C_0 - C_3	$def'.

will generate a LaTeX equivalent and can be referenced using *@refToTheGreatEquation@*

Codeblock

Microdown offers the same code block that markdown but arguments can be specified and the annotation should be named. The first line after the ``` can be language=pharo&label=code1&caption=this is my great piece of code

The following code is not able to display it because markdown quote blocks are strange and interpret nested blocks. So we cannot use quoteblock for quoting!

   ```language=pharo&anchor=code1&caption=this is my great piece of code
    codeBlockMarkupString
    ^ '```'
    ```

```language=pharo&anchor=code1&caption=this is my great piece of code
codeBlockMarkupString
   ^ '\`\`\`'
```

Extensions

• {! aTag | parameters!} is the way to use an extension with parameters
• Environments are defined using <!tag | parameters !>

<!agenda|title=International Workshop on Smalltalk Technologies

<!day|start=2023 August 29th&title=Monday

<!segment|start=10:30

<!talk|subject=Pharo DataFrame: Past, Present, and Future&length=30&author=Safina, Zaitsev, Ferlicot-Delbecque and Sow&room=Room B!>
<!talk|subject=Improving Performance Through Object Lifetime Profiling: the DataFrame Case&length=30&author=Jordan-Montaño, Palumbo, Polito, Ducasse and Tesone&room=Room B!> <!talk|subject=Garbage Collector Tuning in Pathological Allocation Pattern Applications&length=30&author=Palumbo, Jordan-Montaño, Polito, Tesone and Ducasse&room=Room B!>
!>
!>
!>

• Citations {!citation|ref=Blac09a!} -- note that the bib file should be defined in the pillar.conf file

Known limits

Math should be tested

Quote block

When a line starts with '> ' it delimits a quoteblock. The markup is not interpreted.

Codeblock

Codeblock does not support more than four backticks.

Development in Pharo 12!

Loading specific version

To load the latest stable version load the master. If you have trouble loading in the latest Pharo just execute the preloading.st script in the .github folder. This script will remove the existing Microdown package and clear the system.

Metacello new
  baseline: 'Microdown';
  repository: 'github://pillar-markup/Microdown:master/src';
  load.

The process is the following:

• Development in dev
• When stable dev -> in master
• When we can build books master is tagged.
• Then there is the Pharo integration in dedicated branches.

Loading the latest development version

The following script loads all groups in the Baseline:

#( 'Microdown' 'BeautifulComments' 'DocumentBrowser' ) do: [ :name |
        (IceRepository repositoryNamed: name)
            ifNil: [ self inform: 'Project not found: ' , name ]
            ifNotNil: [ :found |
                found
                    unload;
                    forget ] ].

Metacello new
	baseline: 'Microdown';
	repository: 'github://pillar-markup/Microdown:dev/src';
	onConflict: [ :ex | ex useIncoming ];
	onUpgrade: [ :ex | ex useIncoming ];
	load: #('All').

History

We have two sources: Pharo in one hand and Pillar and both are not totally synchronized.

Using Pharo 12: v2.5.x

• v2.5.5 - add support for top-level header as slide definition

• v2.5.4 - add backward compatible anchor in caption + tonel V3 format

• v2.5.1 - add LaTeX math with reference support for Pharo 12 and Pillar development up to v10.0.0

• v2.4.2 for Pillar 9.0.1

Watch out v2.6.0 is older than v.2.5.4

Pillar History

For Pharo 12

• v10.0.0 but with some links problems due to new inline parser using MD v2.5.0

For Pharo 11

• v9.0.1 Fixing link problems.
• v9.0.0 loading in Pharo 11. The development will now happen in P11.

For Pharo 10 -v8.3.2 fixed baseline and updated readme

Implementation

The parser follows the design mentioned in https://github.github.com/gfm, in particular the parsing strategy in Appendix A.

In short, the strategy is that at any point in time, we might have several children of the root which are ""open"". The deepest in open in the tree is called ""current"". All the parents of the current are open.

When a new line is read we do the following:

1. Check if the new line can be consumed by current.
   • as part of this a child of current can be made which can consume the new line
2. If current cannot consume the new line, we close current, move current to its parent, and repeat 1.
3. The root node can consume anything, for instance by making new nodes for storing the new line.
4. The root node is not closed until input is exhausted

The other packages in this repository are the extensions made to produce Pillar model. Such packages should be moved in the future to other locations (probably pillar itself).