Обучение
Модул
Communicate effectively on GitHub using Markdown - Training
Learn to use Markdown to communicate with brevity, clarity, and expression.
Този браузър вече не се поддържа.
Надстройте до Microsoft Edge, за да се възползвате от най-новите функции, актуализации на защитата и техническа поддръжка.
This article provides an alphabetical reference for writing Markdown for Microsoft Learn.
Markdown is a lightweight markup language with plain text formatting syntax. The Microsoft Learn platform supports CommonMark compliant Markdown parsed through the Markdig parsing engine. Microsoft Learn also supports custom Markdown extensions that provide richer content on the Microsoft Learn site.
You can use any text editor to write Markdown, but we recommend Visual Studio Code with the Learn Authoring Pack. The Learn Authoring Pack provides editing tools and preview functionality that lets you see what your articles will look like when rendered on Microsoft Learn.
Alerts are a Markdown extension to create block quotes that render on Microsoft Learn with colors and icons that indicate the significance of the content.
Avoid notes, tips, and important boxes. Readers tend to skip over them. It's better to put that info directly into the article text.
If you need to use alerts, limit them to one or two per article. Multiple notes should never be next to each other in an article.
The following alert types are supported:
> [!NOTE]
> Information the user should notice even if skimming.
> [!TIP]
> Optional information to help a user be more successful.
> [!IMPORTANT]
> Essential information required for user success.
> [!CAUTION]
> Negative potential consequences of an action.
> [!WARNING]
> Dangerous certain consequences of an action.
These alerts look like this on Microsoft Learn:
Бележка
Information the user should notice even if skimming.
Съвет
Optional information to help a user be more successful.
Важно
Essential information required for user success.
Внимание
Negative potential consequences of an action.
Предупреждение
Dangerous certain consequences of an action.
If you use angle brackets in text in your file (for example, to denote a placeholder), you need to manually encode the angle brackets. Otherwise, Markdown thinks that they're intended to be an HTML tag.
For example, encode <script name>
as <script name>
or \<script name>
.
Angle brackets don't have to be escaped in text formatted as inline code or in code blocks.
If you copy from Word into a Markdown editor, the text might contain "smart" (curly) apostrophes or quotation marks. These need to be encoded or changed to basic apostrophes or quotation marks. Otherwise, you end up with things like this when the file is published: It’s
Here are the encodings for the "smart" versions of these punctuation marks:
“
”
’
‘
Съвет
To avoid "smart" characters in your Markdown files, rely on the Learn Authoring Pack's smart quote replacement feature. For more information, see smart quote replacement.
Blockquotes are created using the >
character:
> This is a blockquote. It is usually rendered indented and with a different background color.
The preceding example renders as follows:
This is a blockquote. It is usually rendered indented and with a different background color.
To format text as bold, enclose it in two asterisks:
This text is **bold**.
To format text as italic, enclose it in a single asterisk:
This text is *italic*.
To format text as both bold and italic, enclose it in three asterisks:
This text is both ***bold and italic***.
For guidance on when to use bold and italic text, see text formatting guidelines.
Learn Markdown supports the placement of code snippets both inline in a sentence and as a separate "fenced" block between sentences. For more information, see How to add code to docs.
The columns Markdown extension gives authors the ability to add column-based content layouts that are more flexible and powerful than basic Markdown tables, which are only suited for true tabular data. You can add up to four columns, and use the optional span
attribute to merge two or more columns.
While the columns extension still works, we no longer recommend it for creating custom layouts. We've found that many custom column layouts have accessibility issues or otherwise violate the style guidelines. Don't create custom layouts. Use standard Microsoft Learn features.
The syntax for columns is as follows:
:::row:::
:::column span="":::
Content...
:::column-end:::
:::column span="":::
More content...
:::column-end:::
:::row-end:::
Columns should only contain basic Markdown, including images. Headings, tables, tabs, and other complex structures shouldn't be included. A row can't have any content outside of column.
For example, the following Markdown creates one column that spans two column widths, and one standard (no span
) column:
:::row:::
:::column span="2":::
**This is a 2-span column with lots of text.**
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Donec vestibulum mollis nunc
ornare commodo. Nullam ac metus imperdiet, rutrum justo vel, vulputate leo. Donec
rutrum non eros eget consectetur.
:::column-end:::
:::column span="":::
**This is a single-span column with an image in it.**

:::column-end:::
:::row-end:::
This renders as follows:
This is a 2-span column with lots of text.
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Donec vestibulum mollis nunc ornare commodo. Nullam ac metus imperdiet, rutrum justo vel, vulputate leo. Donec rutrum non eros eget consectetur.
This is a single-span column with an image in it.
Microsoft Learn supports HTML comments if you must comment out sections of your article:
<!--- Here's my comment --->
Предупреждение
Do not put private or sensitive information in HTML comments. Microsoft Learn carries HTML comments through to the published HTML that goes public. While HTML comments are invisible to the reader's eye, they are exposed in the HTML underneath.
Microsoft Learn supports six levels of Markdown headings:
# This is a first level heading (H1)
## This is a second level heading (H2)
...
###### This is a sixth level heading (H6)
#
and heading text.<h1>
, aren't recommended, and in some cases will cause build warnings.Although Markdown supports inline HTML, HTML isn't recommended for publishing to Microsoft Learn, and except for a limited list of values will cause build errors or warnings.
The following file types are supported by default for images:
To support other image types, such as .gif, you must add them as resources in docfx.json:
"resource": [
{
"files" : [
"**/*.png",
"**/*.jpg,
"**/*.gif"
],
The basic Markdown syntax to embed an image is:

Example:

Where <alt text>
is a brief description of the image and <folder path>
is a relative path to the image. Alternate text is required for screen readers for the visually impaired. It's also useful if there's a site bug where the image can't render.
Underscores in alt text aren't rendered properly unless you escape them by prefixing them with a backslash (\_
). However, don't copy file names for use as alt text. For example, instead of this:

Write this:

The custom :::image:::
extension on Microsoft Learn supports standard images, complex images, and icons.
For standard images, the older Markdown syntax will still work, but the new extension is recommended because it supports more powerful functionality, such as specifying a localization scope that's different from the parent topic. Other advanced functionality, such as selecting from the shared image gallery instead of specifying a local image, will be available in the future. The new syntax is as follows:
:::image type="content" source="<folderPath>" alt-text="<alt text>":::
If type="content"
(the default), both source
and alt-text
are required.
You can also use this extension to add an image with a long description that is read by screen readers but not rendered visually on the published page. Long descriptions are an accessibility requirement for complex images, such as graphs. The syntax is the following:
:::image type="complex" source="<folderPath>" alt-text="<alt text>":::
<long description here>
:::image-end:::
If type="complex"
, source
, alt-text
, a long description, and the :::image-end:::
tag are all required.
When your changes are in preview or published, you can check whether the long description exists by right-clicking on the image and selecting Inspect (when using Microsoft Edge browser, although other browsers have similar features). This action brings you to the image source in the HTML code, underneath which you'll find a visually-hidden class. Expand the dropdown on this class, and you'll find your long description:
The :::image:::
extension also supports the border
property, which automatically adds a 1-pixel gray border around your image. The border
property is true
by default for content
and complex
images, so you'll get the border automatically unless you explicitly add the property with a value of false
. The border
property is false
by default for icon
images.
The border
property is the recommended way to add a border. Don't create your own borders manually.
Sometimes the localization scope for an image is different from that of the article or module that contains it. This can cause a bad global experience: for example, if a screenshot of a product is accidentally localized into a language the product isn't available in. To prevent this, you can specify the optional loc-scope
attribute in images of types content
and complex
, and is required for screenshots that show a product with a different localization scope than the article or module that contains it.
The image extension supports icons, which are decorative images and should not have alt text. The syntax for icons is:
:::image type="icon" source="<folderPath>":::
If type="icon"
, source
should be specified but alt-text
shouldn't be.
The border
property is false
by default for icons. If your decorative image requires the standard image border, explicitly add border="true"
to the :::image:::
tag.
Where markdown files need to be repeated in multiple articles, you can use an include file. The includes feature instructs Microsoft Learn to replace the reference with the contents of the include file at build time. You can use includes in the following ways:
An inline or block include file is a Markdown (.md) file. It can contain any valid Markdown. Include files are typically located in a common includes subdirectory, in the root of the repository. When the article is published, the included file is seamlessly integrated into it.
Block include is on its own line:
[!INCLUDE [<title>](<filepath>)]
Inline include is within a line:
Text before [!INCLUDE [<title>](<filepath>)] and after.
Where <title>
is the name of the file and <filepath>
is the relative path to the file. INCLUDE
must be capitalized and there must be a space before the <title>
.
Here are requirements and considerations for include files:
/Includes
folders are excluded from build. Therefore, images stored in /includes
folders and referenced in included files won't be displayed in published content. Store images in a /media
folder outside the /includes
folder.In Markdown, spaces before the first character of a line determine the line's alignment relative to the preceding lines. Indentation especially influences numbered and bulleted lists to render multiple levels of nesting in a hierarchical or outline format.
To indent text to align with a preceding paragraph or an item in a numbered or bulleted list, use spaces.
The following two examples show how indented paragraphs render based on their relationship to other paragraphs.
1. This is a numbered list example (one space after the period before the letter T).
This sentence is indented three spaces.
This code block is indented three spaces.
- This is a bulleted list example (one space after the bullet before the letter T).
This sentence is indented two spaces.
> [!TIP]
> This tip is indented two spaces.
- This is a second-level bullet (indented two spaces, with one space after the bullet before the letter T).
This sentence is indented four spaces.
> This quote block is indented four spaces.
The example above renders as:
This is a numbered list example (one space after the period before the letter T).
This sentence is indented three spaces.
This code block is indented three spaces.
This is a bulleted list example (one space after the bullet before the letter T).
This sentence is indented two spaces.
Съвет
This tip is indented two spaces.
This is a second-level bullet (indented two spaces, with one space after the bullet before the letter T).
This sentence is indented four spaces.
This quote block is indented four spaces.
For information on syntax for links, see Use links in documentation.
To create a numbered list, you can use all 1s. The numbers are rendered in ascending order as a sequential list when published. For increased source readability, you can increment your lists manually.
Don't use letters in lists, including nested lists. They don't render correctly when published to Microsoft Learn. Nested lists using numbers will render as lowercase letters when published. For example:
1. This is
1. a parent numbered list
1. and this is
1. a nested numbered list
1. (fin)
This renders as follows:
To create a bulleted list, use -
or *
followed by a space at the beginning of each line:
- This is
- a parent bulleted list
- and this is
- a nested bulleted list
- All done!
This renders as follows:
Whichever syntax you use, -
or *
, use it consistently within an article.
Checklists are available for use on Microsoft Learn via a custom Markdown extension:
> [!div class="checklist"]
> * List item 1
> * List item 2
> * List item 3
This example renders on Microsoft Learn like this:
Use checklists at the beginning or end of an article to summarize "What will you learn" or "What have you learned" content. Do not add random checklists throughout your articles.
You can use a custom extension to add a next step action button to Microsoft Learn pages.
The syntax is as follows:
> [!div class="nextstepaction"]
> [button text](link to topic)
For example:
> [!div class="nextstepaction"]
> [Learn about adding code to articles](code-in-docs.md)
This renders as follows:
You can use any supported link in a next step action, including a Markdown link to another web page. In most cases, the next action link will be a relative link to another file in the same docset.
You can use the custom no-loc
Markdown extension to identify strings of content that you would like the localization process to ignore.
All strings called out will be case-sensitive; that is, the string must match exactly to be ignored for localization.
To mark an individual string as non-localizable, use this syntax:
:::no-loc text="String":::
For example, in the following, only the single instance of Document
will be ignored during the localization process:
# Heading 1 of the Document
Markdown content within the :::no-loc text="Document":::. The are multiple instances of Document, document, and documents.
Бележка
Use \
to escape special characters:
Lorem :::no-loc text="Find a \"Quotation\""::: Ipsum.
You can also use metadata in the YAML header to mark all instances of a string within the current Markdown file as non-localizable:
author: cillroy
no-loc: [Global, Strings, to be, Ignored]
Бележка
The no-loc metadata is not supported as global metadata in docfx.json file. The localization pipeline doesn't read the docfx.json file, so the no-loc metadata must be added into each individual source file.
In the following example, both in the metadata title
and the Markdown header the word Document
will be ignored during the localization process.
In the metadata description
and the Markdown main content the word document
is localized, because it does not start with a capital D
.
---
title: Title of the Document
author: author-name
description: Description for the document
no-loc: [Title, Document]
---
# Heading 1 of the Document
Markdown content within the document.
Selectors are UI elements that let the user switch between multiple flavors of the same article. They are used in some docsets to address differences in implementation across technologies or platforms. Selectors are typically most applicable to our mobile platform content for developers.
Because the same selector Markdown goes in each article file that uses the selector, we recommend placing the selector for your article in an include file. Then you can reference that include file in all your article files that use the same selector.
There are two types of selectors: a single selector and a multi-selector.
> [!div class="op_single_selector"]
> - [Universal Windows](../articles/notification-hubs-windows-store-dotnet-get-started/)
> - [Windows Phone](../articles/notification-hubs-windows-phone-get-started/)
> - [iOS](../articles/notification-hubs-ios-get-started/)
> - [Android](../articles/notification-hubs-android-get-started/)
> - [Kindle](../articles/notification-hubs-kindle-get-started/)
> - [Baidu](../articles/notification-hubs-baidu-get-started/)
> - [Xamarin.iOS](../articles/partner-xamarin-notification-hubs-ios-get-started/)
> - [Xamarin.Android](../articles/partner-xamarin-notification-hubs-android-get-started/)
... will be rendered like this:
> [!div class="op_multi_selector" title1="Platform" title2="Backend"]
> - [(iOS | .NET)](./mobile-services-dotnet-backend-ios-get-started-push.md)
> - [(iOS | JavaScript)](./mobile-services-javascript-backend-ios-get-started-push.md)
> - [(Windows universal C# | .NET)](./mobile-services-dotnet-backend-windows-universal-dotnet-get-started-push.md)
> - [(Windows universal C# | Javascript)](./mobile-services-javascript-backend-windows-universal-dotnet-get-started-push.md)
> - [(Windows Phone | .NET)](./mobile-services-dotnet-backend-windows-phone-get-started-push.md)
> - [(Windows Phone | Javascript)](./mobile-services-javascript-backend-windows-phone-get-started-push.md)
> - [(Android | .NET)](./mobile-services-dotnet-backend-android-get-started-push.md)
> - [(Android | Javascript)](./mobile-services-javascript-backend-android-get-started-push.md)
> - [(Xamarin iOS | Javascript)](./partner-xamarin-mobile-services-ios-get-started-push.md)
> - [(Xamarin Android | Javascript)](./partner-xamarin-mobile-services-android-get-started-push.md)
... will be rendered like this:
You should only use subscript or superscript when necessary for technical accuracy, such as when writing about mathematical formulas. Don't use them for non-standard styles, such as footnotes.
For both subscript and superscript, use HTML:
Hello <sub>This is subscript!</sub>
This renders as follows:
Hello This is subscript!
Goodbye <sup>This is superscript!</sup>
This renders as follows:
Goodbye This is superscript!
The simplest way to create a table in Markdown is to use pipes and lines. To create a standard table with a header, follow the first line with dashed line:
|This is |a simple |table header|
|----------|-----------|------------|
|table |data |here |
|it doesn't|actually |have to line up nicely!|
This renders as follows:
This is | a simple | table header |
---|---|---|
table | data | here |
it doesn't | actually | have to line up nicely! |
You can align the columns by using colons:
| Fun | With | Tables |
| :------------------- | -------------------: |:---------------:|
| left-aligned column | right-aligned column | centered column |
| $100 | $100 | $100 |
| $10 | $10 | $10 |
| $1 | $1 | $1 |
Renders as follows:
Fun | With | Tables |
---|---|---|
left-aligned column | right-aligned column | centered column |
$100 | $100 | $100 |
$10 | $10 | $10 |
$1 | $1 | $1 |
Съвет
The Learn Authoring Extension for VS Code makes it easy to add basic Markdown tables!
You can also use an online table generator.
Long words in a Markdown table might make the table expand to the right navigation and become unreadable. You can solve that by allowing rendering to automatically insert line breaks within words when needed. Just wrap up the table with the custom class [!div class="mx-tdBreakAll"]
.
Here is a Markdown sample of a table with three rows that will be wrapped by a div
with the class name mx-tdBreakAll
.
> [!div class="mx-tdBreakAll"]
> |Name|Syntax|Mandatory for silent installation?|Description|
> |-------------|----------|---------|---------|
> |Quiet|/quiet|Yes|Runs the installer, displaying no UI and no prompts.|
> |NoRestart|/norestart|No|Suppresses any attempts to restart. By default, the UI will prompt before restart.|
> |Help|/help|No|Provides help and quick reference. Displays the correct use of the setup command, including a list of all options and behaviors.|
It will be rendered like this:
Name | Syntax | Mandatory for silent installation? | Description |
---|---|---|---|
Quiet | /quiet | Yes | Runs the installer, displaying no UI and no prompts. |
NoRestart | /norestart | No | Suppresses any attempts to restart. By default, the UI will prompt before restart. |
Help | /help | No | Provides help and quick reference. Displays the correct use of the setup command, including a list of all options and behaviors. |
You might want line breaks to be automatically inserted within words only in the second column of a table. To limit the breaks to the second column, apply the class mx-tdCol2BreakAll
by using the div
wrapper syntax as shown earlier.
You may notice that the column widths of the tables in your articles look odd or inconsistent. This behavior occurs because the length of text within the cells determines the layout of the table. Unfortunately, there's no way to control how the tables render. This is a limitation of Markdown. Even though it would look nicer to have the width of table columns be consistent, this would have some disadvantages too:
A data matrix table has both a header and a weighted first column, creating a matrix with an empty cell in the top left. Microsoft Learn has custom Markdown for data matrix tables:
| |Header 1 |Header 2|
|------------------|---------|--------|
|**First column A**|Cell 1A |Cell 2A |
|**First column B**|Cell 1B |Cell 2B |
The example renders as:
Header 1 | Header 2 | |
---|---|---|
First column A | Cell 1A | Cell 2A |
First column B | Cell 1B | Cell 2B |
Every entry in the first column must be styled as bold (**bold**
); otherwise the tables won't be accessible for screen readers or valid for Microsoft Learn.
Съвет
The Learn Authoring Pack for VS Code includes a function to convert a regular Markdown table into a data matrix table. Just select the table, right-click, and select Convert to data matrix table.
HTML tables aren't recommended for Microsoft Learn. They aren't human readable in the source - which is a key principle of Markdown.
Обучение
Модул
Communicate effectively on GitHub using Markdown - Training
Learn to use Markdown to communicate with brevity, clarity, and expression.