MDXML

Markdown Extensible Markup Language

Version: 1.1

Tags

Tags are done with markdown H1 headers. The syntax is a # with no free spaces preceding it.

# TagName

Attributes

Attributes are done with markdown H2 headers. The syntax is ## with no free spaces preceding it. Attributes MUST come after tags. Attributes MAY come before any content.

# TagName
## Attribute
### Value

or

# TagName
> ##Attribute
> ###Value

Attribute Values

Attribute Values are done with markdown H3 headers and must come immediately after an attribute. The syntax is ### with no free spaces preceding it. Attribute values MUST come after attributes. Attribute values MAY come before any content.

# TagName
## Attribute
### Value

or

# TagName
> ##Attribute
> ###Value

or

# TagName
##Attribute
> ###Value

Namespaces

Namespaces are done with markdown H4 and H5 headers. H4 headers specify tag namespace, H5 headers specify attribute namespace. The syntax is #### with no free spaces preceding it and ##### with no free spaces preceding it, respectively. Namespaces MUST come after attributes or tags. Namespaces MAY come before any content.

# Tag
#### TagNamespace
## Attribute
### AttributeValue
##### AttributeNamespace

or

# Tag
> #### TagNamespace
> ## Attribute
> ### AttributeValue
> ##### AttributeNamespace

or

# Tag
## Attribute
#### TagNamespace
> ### AttributeValue
> ##### AttributeNamespace

Extension mechanism

Markdown H6 headers may be used as an extension mechanism. There are no restrictions on the behaviour of ######.

Raw Data/CDATA

Raw Data are done with markdown code blocks

    this is raw data!

You need an empty line at the same depth to trigger a raw data block.

depth 0
    this is not raw data

depth 0

    but this is

Note that when using a raw data block with inner data you need to use 5 spaces instead of just 4

>    not a raw data block
>
>     is a raw data block

Note that, as with inline data below, to get out of a raw data block you need an empty line.

    this is raw data!

this is not raw data! putting it on the above line would keep it as raw data

Inner Data

Inner Data are done with markdown quote blocks

> this is inner data

The syntax is a > with no free spaces preceding it, optionally followed by a single ASCII space or tab character.

Going a depth in is as simple as adding more >. Going one or more depths out requires at least 1 new depth indicator line.

depth 0
> depth 1
> > depth 2
> > > depth 3
>
> depth 1. putting it on the above line would keep it a depth 3

depth 0. putting it on the above line would keep it at depth 1

This also means that:

# root tag
> # inner tag
contents

Is valid and "contents" is the contents of inner tag.

Comments

Comments are done with markdown inline code

`this is a comment`

Multiline comments [Not Markdown Compatible]

Multiline comments are done with Github Flavored Markdown's fenced code blocks.

```
this is a multiline comment
```

Version

To add a version attribute, add it at the absolute start of the document

## version
### 1.0
[ rest of the document here ]

Encoding

As with version, encoding also goes at the absolute start of the document

## version
### 1.0
## encoding
### utf-8
[ rest of the document here]

Case-Sensitivity

MDXML is case-sensitive.

Spaces

As with markdown, leading and trailing spaces are ignored (stripped) except in a few specific situations: (Note: the following examples are for clarification only. Their rules are described in the rest of this document.)

#something

is the same as

#     something

but is NOT the same as

   # something

And for mixing raw data and inner data blocks:

>>>>something

is the same as

> > > > something

which is the same as

>>>>    something

but is NOT the same as

>>>>     something

(the latter is a code block with "something")

A line starting with 4 spaces is a code block. An inner data line starting with 5 spaces ("> ") is also a code block.

Escaping

As with markdown, escaping can be done with \.

\    this counts as 4 spaces in a row
\# this is not a tag
\## this is not an attribute
\`this isn't a comment\`

Unlike markdown, MDXML allows you to use \n to insert virtual newlines.

#Tag\n##Attribute\n###Value\nContent

is equivalent to:

#Tag
##Attribute
###Value
Content

All other escapes are disallowed and should be implicitly unescaped. E.g. \x is converted to x. Note that escapes don't work inside raw data blocks or comments (single-line or multi-line). Note that strict parsers are allowed to error on invalid escapes.

Compatibility with XML

It may be possible to convert a MDXML document into an XML document if at least the following conditions are met:

• The document contains only properly encoded legal Unicode characters
• The tag names don't contain any of the characters !"#$%&'()*+,/;<=>?@[\]^{|}~ , nor a backtick, and don't start with -, ., or a numeric digit.
• A single "root" element contains all the other elements.

An example of an XML-compatible MDXML document is available here.

Reference implementation

At https://bitbucket.org/SoniEx2/mdxml