Syntax Extensions#

MyST-Parser is highly configurable, utilising the inherent “plugability” of the markdown-it-py parser. The following syntaxes are optional (disabled by default) and can be enabled via the sphinx conf.py configuration file (see also Configuration). Their goal is generally to add more Markdown friendly syntaxes; often enabling and rendering markdown-it-py plugins that extend the CommonMark specification.

To enable all the syntaxes explained below:

myst_enable_extensions = [
"amsmath",
"attrs_inline",
"colon_fence",
"deflist",
"dollarmath",
"fieldlist",
"html_image",
"replacements",
"smartquotes",
"strikethrough",
"substitution",
]


Changed in version 0.13.0: myst_enable_extensions replaces previous configuration options: admonition_enable, figure_enable, dmath_enable, amsmath_enable, deflist_enable, html_img_enable

1.  Typography#

Adding "smartquotes" to myst_enable_extensions (in the sphinx conf.py configuration file) will automatically convert standard quotations to their opening/closing variants:

• 'single quotes': ‘single quotes’

• "double quotes": “double quotes”

Adding "replacements" to myst_enable_extensions (in the sphinx conf.py configuration file) will automatically convert some common typographic texts:

text

converted

(c), (C)

(tm), (TM)

(r), (R)

®

(p), (P)

§

+-

±

...

?....

?..

!....

!..

????????

???

!!!!!

!!!

,,,

,

--

---

2.  Strikethrough#

New in version 0.17.0.

The strikethrough extension allows text within ~~ delimiters to have a strikethrough (horizontal line) placed over it. For example, ~~strikethrough with *emphasis*~~ renders as: strikethrough with emphasis.

Warning

This extension is currently only supported for HTML output, and you will need to suppress the myst.strikethrough warning (see Build Warnings)

3.  Math shortcuts#

Math is parsed by adding to the myst_enable_extensions list option, in the sphinx conf.py configuration file one or both of:

• "dollarmath" for parsing of dollar $ and $$ encapsulated math. • "amsmath" for direct parsing of amsmath LaTeX environments. These options enable their respective Markdown parser plugins, as detailed in the markdown-it plugin guide. Changed in version 0.13.0: myst_dmath_enable=True and myst_amsmath_enable=True are deprecated, and replaced by myst_enable_extensions = ["dollarmath", "amsmath"] 3.1. Dollar delimited math# Enabling dollarmath will parse the following syntax: • Inline math: ... • Display (block) math: $$...$$ Additionally if myst_dmath_allow_labels=True is set (the default): • Display (block) math with equation label: $$...$$(1) For example, x_{hey}=it+is^{math} renders as $$x_{hey}=it+is^{math}$$. This is equivalent to writing: {math}x_{hey}=it+is^{math}  Block-level math can be specified with $$ signs that wrap the math block you’d like to parse. For example: $$y & = ax^2 + bx + c \\ f(x) & = x^2 + 2xy + y^2$$  $\begin{split} y & = ax^2 + bx + c \\ f(x) & = x^2 + 2xy + y^2 \end{split}$ This is equivalent to the following directive: {math} y & = ax^2 + bx + c \\ f(x) & = x^2 + 2xy + y^2   $\begin{split} y & = ax^2 + bx + c \\ f(x) & = x^2 + 2xy + y^2\end{split}$ You can also add labels to block equations: $$e = mc^2$$ (eqn:best) This is the best equation {eq}eqn:best  (1)#$e = mc^2$ This is the best equation (1) There are a few other options available to control dollar math parsing: myst_dmath_allow_space=False, will cause inline math to only be parsed if there are no initial / final spaces, e.g. $a$ but not $ a$ or $a $. myst_dmath_allow_digits=False, will cause inline math to only be parsed if there are no initial / final digits, e.g. $a$ but not 1$a$ or $a$2. These options can both be useful if you also wish to use $ as a unit of currency.

New in version 0.14.0: myst_dmath_double_inline option

To allow display math (i.e. $$) within an inline context, set myst_dmath_double_inline = True (False by default). This allows for example: Hence, for \alpha \in (0, 1),$$
\mathbb P (\alpha \bar{X} \ge \mu) \le \alpha;
$$i.e., [\alpha \bar{X}, \infty) is a lower 1-sided 1-\alpha confidence bound for \mu.  Hence, for $$\alpha \in (0, 1)$$, $\mathbb P (\alpha \bar{X} \ge \mu) \le \alpha;$ i.e., $$[\alpha \bar{X}, \infty)$$ is a lower 1-sided $$1-\alpha$$ confidence bound for $$\mu$$. 3.2. Math in other block elements# Math will also work when nested in other block elements, like lists or quotes: - A list -$$ a = 1 $$> A block quote >$$ a = 1 

• A list

• $a = 1$

A block quote

$a = 1$

3.3.  Direct LaTeX Math#

By adding "amsmath" to myst_enable_extensions (in the sphinx conf.py configuration file), you can enable direct parsing of amsmath LaTeX equations. These top-level math environments will then be directly parsed:

equation, multline, gather, align, alignat, flalign, matrix, pmatrix, bmatrix, Bmatrix, vmatrix, Vmatrix, eqnarray.

As expected, environments ending in * will not be numbered, for example:

\begin{gather*}
a_1=b_1+c_1\\
a_2=b_2+c_2-d_2+e_2
\end{gather*}

\begin{align}
a_{11}& =b_{11}&
a_{12}& =b_{12}\\
a_{21}& =b_{21}&
a_{22}& =b_{22}+c_{22}
\end{align}

$\begin{gather*} a_1=b_1+c_1\\ a_2=b_2+c_2-d_2+e_2 \end{gather*}$
(2)#\begin{align} a_{11}& =b_{11}& a_{12}& =b_{12}\\ a_{21}& =b_{21}& a_{22}& =b_{22}+c_{22} \end{align}

Note

\labels inside the environment are not currently identified, and so cannot be referenced. We hope to implement this in a future update (see executablebooks/MyST-Parser#202)!

This syntax will also work when nested in other block elements, like lists or quotes:

- A list
- \begin{gather*}
a_1=b_1+c_1\\a_2=b_2+c_2-d_2+e_2
\end{gather*}

> A block quote
> \begin{gather*}
a_1=b_1+c_1\\a_2=b_2+c_2-d_2+e_2
\end{gather*}

• A list

• $\begin{gather*} a_1=b_1+c_1\\a_2=b_2+c_2-d_2+e_2 \end{gather*}$

A block quote

$\begin{gather*} a_1=b_1+c_1\\a_2=b_2+c_2-d_2+e_2 \end{gather*}$

3.4.  Mathjax and math parsing#

When building HTML using the sphinx.ext.mathjax extension (enabled by default), If dollarmath is enabled, Myst-Parser injects the tex2jax_ignore (MathJax v2) and mathjax_ignore (MathJax v3) classes in to the top-level section of each MyST document, and adds the following default MathJax configuration:

MathJax version 2 (see the tex2jax preprocessor:

MathJax.Hub.Config({"tex2jax": {"processClass": "tex2jax_process|mathjax_process|math|output_area"}})


MathJax version 3 (see the document options):

window.MathJax = {"options": {"processHtmlClass": "tex2jax_process|mathjax_process|math|output_area"}}


This ensurea that MathJax processes only math, identified by the dollarmath and amsmath extensions, or specified in math directives.

To change this behaviour, set a custom regex, for identifying HTML classes to process, like myst_mathjax_classes="math|myclass", or set myst_update_mathjax=False to inhibit this override and process all HTML elements.

Adding "linkify" to myst_enable_extensions (in the sphinx conf.py configuration file) will automatically identify “bare” web URLs and add hyperlinks:

www.example.com -> www.example.com

To only match URLs that start with schema, such as http://example.com, set myst_linkify_fuzzy_links=False.

Important

This extension requires that linkify-it-py is installed. Either directly; pip install linkify-it-py or via pip install myst-parser[linkify].

5.  Substitutions (with Jinja2)#

Adding "substitution" to myst_enable_extensions (in the sphinx conf.py configuration file) will allow you to add substitutions, added in either the conf.py using myst_substitutions:

myst_substitutions = {
"key1": "I'm a **substitution**"
}


or at the top of the file, in the front-matter section (see this section):

---
myst:
substitutions:
key1: "I'm a **substitution**"
key2: |
{note}
{{ key1 }}

key3: |
{image} img/fun-fish.png
:alt: fishy
:width: 200px

key4: example
---


Important

Keys in the front-matter will override ones in the conf.py.

You can use these substitutions inline or as blocks, and you can even nest substitutions in other substitutions (but circular references are prohibited):

Inline: {{ key1 }}

Block level:

{{ key2 }}

| col1     | col2     |
| -------- | -------- |
| {{key2}} | {{key3}} |


Inline: I’m a substitution

Block level:

Note

I’m a substitution

col1

col2

Note

I’m a substitution

Important

Substitutions will only be assessed where you would normally use Markdown, e.g. not in code blocks:


{{ key1 }}


{{ key1 }}


One should also be wary of using unsuitable directives for inline substitutions. This may lead to unexpected outcomes.

Substitution references are assessed as Jinja2 expressions which can use filters, and also contains the Sphinx Environment in the context (as env). Therefore you can do things like:

- version: {{ env.config.version }}
- docname: {{ env.docname | upper }}
- {{ "a" + "b" }}

• version: 1.0.0

• docname: SYNTAX/OPTIONAL

• ab

You can also change the delimiter if necessary, for example setting in the conf.py:

myst_sub_delimiters = ["|", "|"]


Will parse: || "a" + "b" ||. This should be changed with care though, so as not to affect other syntaxes.

The exact logic for handling substitutions is:

1. Combine global substitutions (specified in conf.py) with front-matter substitutions, to create a variable context (front-matter takes priority)

2. Add the sphinx env to the variable context

3. Create the string content to render using Jinja2 (passing it the variable context)

4. If the substitution is inline and not a directive, render ignoring block syntaxes (like lists or block-quotes), otherwise render with all syntax rules.

5.1.  Substitutions and URLs#

Substitutions cannot be directly used in URLs, such as [a link](https://{{key4}}.com) or <https://{{key4}}.com>. However, since Jinja2 substitutions allow for Python methods to be used, you can use string formatting or replacements:

{{ '[a link](https://{}.com)'.format(key4) }}



6.  Code fences using colons#

By adding "colon_fence" to myst_enable_extensions (in the sphinx conf.py configuration file), you can also use ::: delimiters to denote directives, instead of .

Using colons instead of back-ticks has the benefit of allowing the content to be rendered correctly, when you are working in any standard Markdown editor. It is ideal for admonition type directives (as documented in Directives) or tables with titles, for example:

:::{note}
This text is **standard** _Markdown_
:::

:::{table} This is a **standard** _Markdown_ title
:align: center
:widths: grid

abc | mnp | xyz
--- | --- | ---
123 | 456 | 789
:::


Note

This text is standard Markdown

This is a standard Markdown title#

abc

mnp

xyz

123

456

789

Similar to normal directives, these directives can also be nested:

::::{important}
:::{note}
This text is **standard** _Markdown_
:::
::::


Important

Note

This text is standard Markdown

and also parameter options can be used:

:::{admonition} This *is* also **Markdown**
:class: warning

This text is **standard** _Markdown_
:::


This is also Markdown

This text is standard Markdown

Changed in version 0.13.0: myst_admonition_enable is deprecated and replaced by myst_enable_extensions = ["colon_fence"] (see above). Also, classes should now be set with the :class: myclass option.

The MyST Parser can automatically generate label “slugs” for header anchors so that you can reference them from markdown links. For example, you can use header bookmark links, locally; [](#header-anchor), or cross-file [](path/to/file.md#header-anchor). To achieve this, use the myst_heading_anchors = DEPTH configuration option, where DEPTH is the depth of header levels for which you wish to generate links.

For example, the following configuration in conf.py tells the myst_parser to generate labels for heading anchors for h1, h2, and h3 level headings (corresponding to #, ##, and ### in markdown).

myst_heading_anchors = 3


[](#auto-generated-header-anchors)


The paths to other files should be relative to the current file:

[**link text**](./typography.md#headings)


8.1.  Anchor slug structure#

The anchor “slugs” created aim to follow the GitHub implementation:

• lower-case text

• remove punctuation

• replace spaces with -

• enforce uniqueness via suffix enumeration -1

To change the slug generation function, set myst_heading_slug_func in your conf.py to a function that accepts a string and returns a string.

New in version 0.19.0: myst_heading_slug_func can now also be set to a string, which will be interpreted as an import path to a function, e.g. myst_heading_slug_func = "mypackage.mymodule.slugify".

9.  Definition Lists#

By adding "deflist" to myst_enable_extensions (in the sphinx conf.py configuration file), you will be able to utilise definition lists. Definition lists utilise the markdown-it-py deflist plugin, which itself is based on the Pandoc definition list specification.

This syntax can be useful, for example, as an alternative to nested bullet-lists:

- Term 1
- Definition
- Term 2
- Definition

• Term 1

• Definition

• Term 2

• Definition

Term 1
: Definition

Term 2
: Definition

Term 1

Definition

Term 2

Definition

From the Pandoc documentation:

Each term must fit on one line, which may optionally be followed by a blank line, and must be followed by one or more definitions. A definition begins with a colon or tilde, which may be indented one or two spaces.

A term may have multiple definitions, and each definition may consist of one or more block elements (paragraph, code block, list, etc.)

Here is a more complex example, demonstrating some of these features:

Term *with Markdown*
: Definition [with reference](syntax/definition-lists)

A second paragraph
: A second definition

Term 2
~ Definition 2a
~ Definition 2b

Term 3
:     A code block
: > A quote
: A final definition, that can even include images:

<img src="img/fun-fish.png" alt="fishy" width="200px">

Term with Markdown

Definition with reference

A second paragraph

A second definition

Term 2

Definition 2a

Definition 2b

Term 3
A code block


A quote

A final definition, that can even include images:

By adding "tasklist" to myst_enable_extensions (in the sphinx conf.py configuration file), you will be able to utilise task lists. Task lists utilise the markdown-it-py tasklists plugin, and are applied to markdown list items starting with [ ] or [x]:

- [ ] An item that needs doing
- [x] An item that is complete

• An item that needs doing

• An item that is complete

11.  Field Lists#

New in version 0.16.0.

By adding "fieldlist" to myst_enable_extensions (in the sphinx conf.py configuration file), you will be able to utilise field lists. Field lists are mappings from field names to field bodies, based on the reStructureText syntax.

:name only:
:name: body
:*Nested syntax*: Both name and body may contain **nested syntax**.
:Paragraphs: Since the field marker may be quite long, the second
and subsequent lines of a paragraph do not have to line up
with the first line.
:Blocks:

As well as paragraphs, any block syntaxes may be used in a field body:

- Me
- Myself
- I

python
print("Hello, world!")


name only:

name:

body

Nested syntax:

Both name and body may contain nested syntax.

Paragraphs:

Since the field marker may be quite long, the second and subsequent lines of a paragraph do not have to line up with the first line.

Blocks:

As well as paragraphs, any block syntaxes may be used in a field body:

• Me

• Myself

• I

print("Hello, world!")


A prominent use case of field lists is for use in API docstrings, as used in Sphinx’s docstring renderers:

{py:function} send_message(sender, priority)

Send a message to a recipient

:param str sender: The person sending the message
:param priority: The priority of the message, can be a number 1-5
:type priority: int
:return: the message id
:rtype: int
:raises ValueError: if the message_body exceeds 160 characters


send_message(sender, priority)#

Send a message to a recipient

Parameters:
• sender (str) – The person sending the message

• priority (int) – The priority of the message, can be a number 1-5

Returns:

the message id

Return type:

int

Raises:

ValueError – if the message_body exceeds 160 characters

Note

Currently sphinx.ext.autodoc does not support MyST, see Use sphinx.ext.autodoc in Markdown files.

12.  Attributes#

New in version 0.19: This feature is in beta, and feedback is welcome.

attrs_inline also replace the previous attrs_image extension, which is now deprecated.

Attributes are a way of enriching standard CommonMark syntax, by adding additional information to elements.

Attributes are specified inside curly braces {}, for example {#my-id .my-class key="value"}, and come before a block element or after an inline element.

Inside the curly braces, the following syntax is recognised:

• .foo specifies foo as a class. Multiple classes may be given in this way; they will be combined.

• #foo specifies foo as an identifier. An element may have only one identifier; if multiple identifiers are given, the last one is used.

• key="value" or key=value specifies a key-value attribute. Quotes are not needed when the value consists entirely of ASCII alphanumeric characters or _ or : or -. Backslash escapes may be used inside quoted values. Note only certain keys are supported, see below.

• % begins a comment, which ends with the next % or the end of the attribute (}).

Attributes are cumulative, so that if multiple attributes follow each other, the inner attributes override the outer ones. One exception is that if multiple classes are given, they are combined. For example:

{#id1 .class1 key1="value1"}
{#id2 .class2 key2="value2"}
block

[inline]{#id2 .class2 key2="value2"}{#id1 .class1 key1="value1"}


is equivalent to:

{#id2 .class1 .class2 key1="value1" key2="value2"}
block

[inline]{#id2 .class1 .class2 key1="value1" key2="value2"}


This is adapted from djot inline/block attributes, and also related to pandoc bracketed spans.

12.1.  Block attributes#

By adding "attrs_block" to myst_enable_extensions (in the sphinx conf.py configuration file), you can enable parsing of block attributes before certain block syntaxes.

For example, the following Markdown:

{#mypara .bg-warning}
Here is a paragraph with attributes.

{ref}A reference to my paragraph <mypara>


Here is a paragraph with attributes.

A reference to my paragraph

id and class are supported for most block syntaxes, but only certain key-value attributes are supported for each syntax.

For ordered lists, the style key is supported, and can be one of decimal, lower-alpha, upper-alpha, lower-roman, upper-roman:

{style=lower-alpha}
1. a
2. b

{style=upper-alpha}
1. a
2. b

{style=lower-roman}
1. a
2. b

{style=upper-roman}
1. a
2. b

1. a

2. b

1. a

2. b

1. a

2. b

1. a

2. b

For code fences, the lineno-start and emphasize-lines keys are supported:

{lineno-start=1 emphasize-lines="2,3"}
python
a = 1
b = 2
c = 3


1a = 1
2b = 2
3c = 3


For block quotes, the attribution key is supported:

{attribution="Chris Sewell"}
> Hallo


Hallo

—Chris Sewell

For definition lists, the adding a glossary key turns the definition list into a glossary (similar to using the glossary directive in Sphinx):

{.glossary}
term name
: Definition of the term

{term}term name

term name

Definition of the term

term name

12.2.  Inline attributes#

By adding "attrs_inline" to myst_enable_extensions (in the sphinx conf.py configuration file), you can enable parsing of inline attributes after certain inline syntaxes.

For example, the following Markdown:

- [A span of text with attributes]{#spanid .bg-warning},
{ref}a reference to the span <spanid>

- A literal with attributes{#literalid .bg-warning},
{ref}a reference to the literal <literalid>

- An autolink with attributes: <https://example.com>{.bg-warning title="a title"}

{ref}a reference to the link <linkid>

- ![An image with attribute](img/fun-fish.png){#imgid .bg-warning w=100px align=center}
{ref}a reference to the image <imgid>


id and class are supported for most inline syntaxes, but only certain key-value attributes are supported for each syntax.

For literals, the following attributes are supported:

• language/lexer/l defines the syntax lexer, e.g. a = "b"{l=python} is displayed as a = "b". Note, this is only supported in sphinx >= 5.

For images, the following attributes are supported (equivalent to the image directive):

• width/w defines the width of the image (in %, px, em, cm, etc)

• height/h defines the height of the image (in px, em, cm, etc)

• align/a defines the scale of the image (left, center, or right)

13.  HTML Images#

MyST provides a few different syntaxes for including images in your documentation, as explained below.

The first is the standard Markdown syntax:

![fishy](img/fun-fish.png)


This will correctly copy the image to the build folder and will render it in all output formats (HTML, TeX, etc). However, it is limited in the configuration that can be applied, for example setting a width.

As discussed above, MyST allow for directives to be used such as image and figure (see the sphinx documentation):

{image} img/fun-fish.png
:alt: fishy
:class: bg-primary
:width: 200px
:align: center



Additional options can now be set, however, in contrast to the Markdown syntax, this syntax will not show the image in common Markdown viewers (for example when the files are viewed on GitHub).

The final option is directly using HTML, which is also parsed by MyST. This is usually a bad option, because the HTML is treated as raw text during the build process and so sphinx will not recognise that the image file is to be copied, and will not output the HTML into non-HTML output formats.

HTML parsing to the rescue!

By adding "html_image" to myst_enable_extensions (in the sphinx conf.py configuration file), MySt-Parser will attempt to convert any isolated img tags (i.e. not wrapped in any other HTML) to the internal representation used in sphinx.

<img src="img/fun-fish.png" alt="fishy" width="200px">
<img src="img/fun-fish.png" alt="fishy" width="200px" class="bg-primary">


Allowed attributes are equivalent to the image directive: src, alt, class, width, height and name. Any other attributes will be dropped.

HTML image can also be used inline!

I’m an inline image:

14.  Markdown Figures#

By adding "colon_fence" to myst_enable_extensions (in the sphinx conf.py configuration file), we can combine the above two extended syntaxes, to create a fully Markdown compliant version of the figure directive named figure-md.

Changed in version 0.13.0: myst_figure_enable with the figure directive is deprecated and replaced by myst_enable_extensions = ["colon_fence"] and figure-md.

The figure block must contain only two components; an image, in either Markdown or HTML syntax, and a single paragraph for the caption.

The first line argument is optional and taken as the reference target of the figure:

:::{figure-md}
![fishy](img/fun-fish.png){.bg-primary .mb-1 width=200px}

This is a caption in **Markdown**
:::

:::{figure-md} fig-target
:class: myclass

<img src="img/fun-fish.png" alt="fishy" class="bg-primary mb-1" width="200px">

This is a caption in **Markdown**
:::


As we see here, the target we set can be referenced:

[Go to the fish!](#fig-target)


By adding "html_admonition" to myst_enable_extensions (in the sphinx conf.py configuration file), you can enable parsing of <div class="admonition"> HTML blocks. These blocks will be converted internally to Sphinx admonition directives, and so will work correctly for all output formats. This is helpful when you care about viewing the “source” Markdown, such as in Jupyter Notebooks.

If the first element within the div is <div class="title"> or <p class="title">, then this will be set as the admonition title. All internal text (and the title) will be parsed as MyST-Markdown and all classes and an optional name will be passed to the admonition:

<div class="admonition note" name="html-admonition" style="background: lightgreen; padding: 10px">
<p class="title">This is the **title**</p>
This is the *content*
</div>


This is the title

This is the content

During the Sphinx render, both the class and name attributes will be used by Sphinx, but any other attributes like style will be discarded.

Warning

There can be no empty lines in the block, otherwise they will be read as two separate blocks. If you want to use multiple paragraphs then they can be enclosed in <p>:

<div class="admonition note">
<p>Paragraph 1</p>
<p>Paragraph 2</p>
</div>


Note

Paragraph 1

Paragraph 2

You can also nest HTML admonitions:

<div class="admonition">
<p>Some **content**</p>
<div class="title">A *title*</div>
<p>Paragraph 1</p>
<p>Paragraph 2</p>
</div>
</div>


Note

Some content

A title

Paragraph 1

Paragraph 2