Helpers/DslMethods/Format-RPDslItComment.ps1
|
<#
.SYNOPSIS Format a comment before the Pester It block. .DESCRIPTION The comment must be wrapped into a block comment and start right before the It block definition. #> function Format-RPDslItComment { [CmdletBinding()] [OutputType([System.String])] param () try { # Use reflection on the call stack of PowerShell to get the comment # block above the function itself. This is how the call stack locks # like if we have not test cases: # [0]: Format-RPDslItComment This function itself. # [1]: Step ReportPortal test step function. # [2]: <ScriptBlock> The actual test script block. # And this is how the call stack locks with test cases: # [0]: Format-RPDslItComment This function itself. # [1]: Step ReportPortal test step function. # [2]: Step ReportPortal test step function. # [3]: <ScriptBlock> The actual test script block. # ... [System.Management.Automation.CallStackFrame[]] $callStack = Get-PSCallStack # Select the element with id 2, so the element containing the actual # test function implementation. It will also point to the script full # path and line number of the implementation. $scriptBlockFrame = $callStack | Where-Object { $_.Command -eq '<ScriptBlock>' } | Select-Object -First 1 #-Skip 2 if ($null -eq $scriptBlockFrame) { throw "Call stack frame number 3 (index 2) not found!" } $scriptFullName = $scriptBlockFrame.ScriptName $scriptLineNumber = $scriptBlockFrame.ScriptLineNumber # Read the content of the script file up to the line before the script # block, this is where the comment should end. Then reverse the array # for parsing. [System.String[]] $scriptContent = Get-Content -Path $scriptFullName -TotalCount ($scriptLineNumber - 1) -Encoding 'UTF8' [System.Array]::Reverse($scriptContent) # Check for a single line comment block. if ($scriptContent[0].Trim().StartsWith('<#') -and $scriptContent[0].Trim().EndsWith('#>')) { $comment = $scriptContent[0].Trim() $comment = $comment.Substring(2, $comment.Length -4).Trim() return $comment } # Check for a multi line comment block. if ($scriptContent[0].Trim() -eq '#>') { $comment = '' for ($i = 1; $i -lt $scriptContent.Count; $i++) { # If we have reached the end of the script block, exit the # loop. Else append the new text to the comment. $scriptContentLine = $scriptContent[$i].Trim() if ($scriptContentLine -eq '<#') { break } if ($scriptContentLine -in '.SYNOPSIS', '.DESCRIPTION') { $scriptContentLine = "**TEST DESCRIPTION**`n" } if ($scriptContentLine -in '.ROLE', '.COMPONENT') { $scriptContentLine = "**RESPONSIBLE TEAM**`n" } if ($scriptContentLine -in '.EXAMPLE', '.NOTES') { $scriptContentLine = "**POSSIBLE SOLUTION**`n" } $comment = "{0}`n{1}" -f $scriptContentLine, $comment } return $comment } # IMPORTANT # This option is actually disabled because the content can be markdown # and this is problematic with single line comments. # # Check if we have a comment line. # if ($scriptContent[0].Trim().StartsWith('# ')) # { # $comment = '' # for ($i = 0; $i -lt $scriptContent.Count; $i++) # { # # As long as the line starts with a comment, add it to the # # output. Else stop the loop. # $scriptContentLine = $scriptContent[$i].Trim() # if (-not $scriptContentLine.StartsWith('# ')) { break } # $comment = '{0} {1}' -f $scriptContentLine.Substring(2), $comment # } # return $comment # } return 'Test step description not found.' } catch { return "Error generating the test step description.`n`n```````n$_`n``````" } } |