The existing documentation has used multiple styles, over time, and the formatting rules have changed multiple times. We have adopted a consistent style for PowerShell code blocks and output in our documentation.
Markdown supports two different code styles:
- Code spans (inline) - marked by a single backtick (
`
) character. Used inline in a paragraph rather than as a standalone block. Most commonly used around cmdlet names. See Formatting command syntax elements. - Code blocks - a multi-line block surrounded by triple-backtick (
```
) strings.
Markdown allows for indentation to signify a code block, but this pattern can be problematic and
should be avoided. All code blocks should be contained in a code fence. A code fence is a block of
code surrounded by triple-backtick (```
) strings. The code fence markers must be on their own
line before and after the code sample. The marker at the start of the code block may have an
optional language label. Microsoft's Open Publishing System (OPS) uses the language label to support
the syntax highlighting feature.
OPS also adds a Copy button that copies the contents of the code block to the clipboard. This allows you to quickly paste the code into a script for testing the code example. However, not all examples in our documentation are intended to be run. Some code blocks are simple illustrations of a PowerShell concept.
There are two types code blocks used in our documentation:
- Illustrative examples
- Executable examples
Illustrative examples are used to explain a PowerShell concept. They are not meant to be copied to the clipboard for execution. These are most commonly used for simple examples that are easy to type. They are also used for syntax examples where you are illustrating the syntax of a command.
Illustrative examples use a "naked" code fence to mark the beginning and end of the code block. The code block may contain example output from the command being illustrated.
This example illustrates all the possible parameters of the Get-Command
cmdlet.
```
Get-Command [-Verb <String[]>] [-Noun <String[]>] [-Module <String[]>]
[-FullyQualifiedModule <ModuleSpecification[]>] [-TotalCount <Int32>] [-Syntax] [-ShowCommandInfo]
[[-ArgumentList] <Object[]>] [-All] [-ListImported] [-ParameterName <String[]>]
[-ParameterType <PSTypeName[]>] [<CommonParameters>]
```
This example describes the for
statement in generalized terms:
```
for (<init>; <condition>; <repeat>)
{<statement list>}
```
Here is an example illustrating PowerShell comparison operators:
```
PS> 2 -eq 2
True
PS> 2 -eq 3
False
PS> 1,2,3 -eq 2
2
PS> "abc" -eq "abc"
True
PS> "abc" -eq "abc", "def"
False
PS> "abc", "def" -eq "abc"
abc
```
Note that this example has the simplified PowerShell prompt and shows the resulting output. In this case, we don't intend the reader to copy and run this example.
More complex examples or examples that are intended to be useful to copy and execute should use the following block-style markup:
```powershell
<PowerShell code goes here>
```
The output displayed by PowerShell commands should be enclosed in an Output code block to prevent syntax highlighting. For example:
```powershell
Get-Command -Module Microsoft.PowerShell.Security
```
```Output
CommandType Name Version Source
----------- ---- ------- ------
Cmdlet ConvertFrom-SecureString 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet ConvertTo-SecureString 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet Get-Acl 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet Get-AuthenticodeSignature 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet Get-CmsMessage 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet Get-Credential 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet Get-ExecutionPolicy 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet Get-PfxCertificate 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet New-FileCatalog 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet Protect-CmsMessage 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet Set-Acl 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet Set-AuthenticodeSignature 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet Set-ExecutionPolicy 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet Test-FileCatalog 3.0.0.0 Microsoft.PowerShell.Security
Cmdlet Unprotect-CmsMessage 3.0.0.0 Microsoft.PowerShell.Security
```
The Output code label is not an official "language" supported by the syntax highlighting system. However, this label is useful because OPS adds the "Output" label to the code box on the web page. And this "Output" code box has no syntax highlighting.
Avoid using line continuation characters (`
) in PowerShell code examples. These are hard to
see and can cause problems if there are extra spaces at the end of the line.
- Use PowerShell splatting to reduce line length for cmdlets that have a lot of parameters.
- Take advantage of PowerShell's natural line break opportunities, like after pipe (
|
) characters, opening braces, parentheses, and brackets.
Use of the prompt string is discouraged and should be limited to scenarios that are meant to illustrate command line usage. Prompts are required for examples that illustrate commands that alter the prompt or when the path displayed is significant to the scenario being illustrated.
- PowerShell prompts should only be used in illustrative examples. For most of these examples, the
prompt string should be "
PS>
". This prompt is independent of OS-specific indicators. - Prompts should NOT be used in executable examples.
The following example illustrates how the prompt changes when using the Registry provider.
PS C:\> cd HKCU:\System\
PS HKCU:\System\> dir
Hive: HKEY_CURRENT_USER\System
Name Property
---- --------
CurrentControlSet
GameConfigStore GameDVR_Enabled : 1
GameDVR_FSEBehaviorMode : 2
Win32_AutoGameModeDefaultProfile : {2, 0, 1, 0...}
Win32_GameModeRelatedProcesses : {1, 0, 1, 0...}
GameDVR_HonorUserFSEBehaviorMode : 0
GameDVR_DXGIHonorFSEWindowsCompatible : 0
You should always use the full name of all cmdlets and parameters unless you are specifically talking about the alias. Cmdlet and parameter names must use the proper Pascal-case spelling defined in the code.
Avoid using positional parameters. In general, you should use always include the parameter name in an example, even if the parameter positional. This reduces the chance of confusion in your examples.
See the Updating Reference article.