EnhancedHTML2.psm1
$script = '' function ConvertTo-EnhancedHTML { <# .SYNOPSIS Provides an enhanced version of the ConvertTo-HTML command that includes inserting an embedded CSS style sheet, JQuery, and JQuery Data Tables for interactivity. Intended to be used with HTML fragments that are produced by ConvertTo-EnhancedHTMLFragment. This command does not accept pipeline input. .PARAMETER jQueryURI A Uniform Resource Indicator (URI) pointing to the location of the jQuery script file. You can download jQuery from www.jquery.com; you should host the script file on a local intranet Web server and provide a URI that starts with http:// or https://. Alternately, you can also provide a file system path to the script file, although this may create security issues for the Web browser in some configurations. Tested with v1.8.2. Defaults to http://ajax.aspnetcdn.com/ajax/jQuery/jquery-1.8.2.min.js, which will pull the file from Microsoft's ASP.NET Content Delivery Network. .PARAMETER jQueryDataTableURI A Uniform Resource Indicator (URI) pointing to the location of the jQuery Data Table script file. You can download this from www.datatables.net; you should host the script file on a local intranet Web server and provide a URI that starts with http:// or https://. Alternately, you can also provide a file system path to the script file, although this may create security issues for the Web browser in some configurations. Tested with jQuery DataTable v1.9.4 Defaults to http://ajax.aspnetcdn.com/ajax/jquery.dataTables/1.9.3/jquery.dataTables.min.js, which will pull the file from Microsoft's ASP.NET Content Delivery Network. .PARAMETER CssStyleSheet The CSS style sheet content - not a file name. If you have a CSS file, you can load it into this parameter as follows: -CSSStyleSheet (Get-Content MyCSSFile.css) Alternately, you may link to a Web server-hosted CSS file by using the -CssUri parameter. .PARAMETER CssUri A Uniform Resource Indicator (URI) to a Web server-hosted CSS file. Must start with either http:// or https://. If you omit this, you can still provide an embedded style sheet, which makes the resulting HTML page more standalone. To provide an embedded style sheet, use the -CSSStyleSheet parameter. .PARAMETER Title A plain-text title that will be displayed in the Web browser's window title bar. Note that not all browsers will display this. .PARAMETER PreContent Raw HTML to insert before all HTML fragments. Use this to specify a main title for the report: -PreContent "<H1>My HTML Report</H1>" .PARAMETER PostContent Raw HTML to insert after all HTML fragments. Use this to specify a report footer: -PostContent "Created on $(Get-Date)" .PARAMETER HTMLFragments One or more HTML fragments, as produced by ConvertTo-EnhancedHTMLFragment. -HTMLFragments $part1,$part2,$part3 .EXAMPLE For examples, please see the free ebooks, "Creating HTML Reports in PowerShell," available at http://powershell.org/ebooks. #> [CmdletBinding()] param( [string]$jQueryURI = 'http://ajax.aspnetcdn.com/ajax/jQuery/jquery-1.8.2.min.js', [string]$jQueryDataTableURI = 'http://ajax.aspnetcdn.com/ajax/jquery.dataTables/1.9.3/jquery.dataTables.min.js', [Parameter(ParameterSetName='CSSContent')][string[]]$CssStyleSheet, [Parameter(ParameterSetName='CSSURI')][string[]]$CssUri, [string]$Title = 'Report', [string]$PreContent, [string]$PostContent, [Parameter(Mandatory=$True)][string[]]$HTMLFragments ) <# Add CSS style sheet. If provided in -CssUri, add a <link> element. If provided in -CssStyleSheet, embed in the <head> section. Note that BOTH may be supplied - this is legitimate in HTML. #> Write-Verbose "Making CSS style sheet" $stylesheet = "" if ($PSBoundParameters.ContainsKey('CssUri')) { $stylesheet = "<link rel=`"stylesheet`" href=`"$CssUri`" type=`"text/css`" />" } if ($PSBoundParameters.ContainsKey('CssStyleSheet')) { $stylesheet = "<style>$CssStyleSheet</style>" | Out-String } <# Create the HTML tags for the page title, and for our main javascripts. #> Write-Verbose "Creating <TITLE> and <SCRIPT> tags" $titletag = "" if ($PSBoundParameters.ContainsKey('title')) { $titletag = "<title>$title</title>" } $script += "<script type=`"text/javascript`" src=`"$jQueryURI`"></script>`n<script type=`"text/javascript`" src=`"$jQueryDataTableURI`"></script>" <# Render supplied HTML fragments as one giant string #> Write-Verbose "Combining HTML fragments" $body = $HTMLFragments | Out-String <# If supplied, add pre- and post-content strings #> Write-Verbose "Adding Pre and Post content" if ($PSBoundParameters.ContainsKey('precontent')) { $body = "$PreContent`n$body" } if ($PSBoundParameters.ContainsKey('postcontent')) { $body = "$body`n$PostContent" } <# Add a final script that calls the datatable code We dynamic-ize all tables with the .enhancedhtml-dynamic-table class, which is added by ConvertTo-EnhancedHTMLFragment. #> Write-Verbose "Adding interactivity calls" $datatable = "" $datatable = "<script type=`"text/javascript`">" $datatable += '$(document).ready(function () {' $datatable += "`$('.enhancedhtml-dynamic-table').dataTable();" $datatable += '} );' $datatable += "</script>" <# Datatables expect a <thead> section containing the table header row; ConvertTo-HTML doesn't produce that so we have to fix it. #> Write-Verbose "Fixing table HTML" $body = $body -replace '<tr><th>','<thead><tr><th>' $body = $body -replace '</th></tr>','</th></tr></thead>' <# Produce the final HTML. We've more or less hand-made the <head> amd <body> sections, but we let ConvertTo-HTML produce the other bits of the page. #> Write-Verbose "Producing final HTML" ConvertTo-HTML -Head "$stylesheet`n$titletag`n$script`n$datatable" -Body $body Write-Debug "Finished producing final HTML" } function ConvertTo-EnhancedHTMLFragment { <# .SYNOPSIS Creates an HTML fragment (much like ConvertTo-HTML with the -Fragment switch that includes CSS class names for table rows, CSS class and ID names for the table, and wraps the table in a <DIV> tag that has a CSS class and ID name. .PARAMETER InputObject The object to be converted to HTML. You cannot select properties using this command; precede this command with Select-Object if you need a subset of the objects' properties. .PARAMETER EvenRowCssClass The CSS class name applied to even-numbered <TR> tags. Optional, but if you use it you must also include -OddRowCssClass. .PARAMETER OddRowCssClass The CSS class name applied to odd-numbered <TR> tags. Optional, but if you use it you must also include -EvenRowCssClass. .PARAMETER TableCssID Optional. The CSS ID name applied to the <TABLE> tag. .PARAMETER DivCssID Optional. The CSS ID name applied to the <DIV> tag which is wrapped around the table. .PARAMETER TableCssClass Optional. The CSS class name to apply to the <TABLE> tag. .PARAMETER DivCssClass Optional. The CSS class name to apply to the wrapping <DIV> tag. .PARAMETER As Must be 'List' or 'Table.' Defaults to Table. Actually produces an HTML table either way; with Table the output is a grid-like display. With List the output is a two-column table with properties in the left column and values in the right column. .PARAMETER Properties A comma-separated list of properties to include in the HTML fragment. This can be * (which is the default) to include all properties of the piped-in object(s). In addition to property names, you can also use a hashtable similar to that used with Select-Object. For example: Get-Process | ConvertTo-EnhancedHTMLFragment -As Table ` -Properties Name,ID,@{n='VM'; e={$_.VM}; css={if ($_.VM -gt 100) { 'red' } else { 'green' }}} This will create table cell rows with the calculated CSS class names. E.g., for a process with a VM greater than 100, you'd get: <TD class="red">475858</TD> You can use this feature to specify a CSS class for each table cell based upon the contents of that cell. Valid keys in the hashtable are: n, name, l, or label: The table column header e or expression: The table cell contents css or csslcass: The CSS class name to apply to the <TD> tag Another example: @{n='Free(MB)'; e={$_.FreeSpace / 1MB -as [int]}; css={ if ($_.FreeSpace -lt 100) { 'red' } else { 'blue' }} This example creates a column titled "Free(MB)". It will contain the input object's FreeSpace property, divided by 1MB and cast as a whole number (integer). If the value is less than 100, the table cell will be given the CSS class "red." If not, the table cell will be given the CSS class "blue." The supplied cascading style sheet must define ".red" and ".blue" for those to have any effect. .PARAMETER PreContent Raw HTML content to be placed before the wrapping <DIV> tag. For example: -PreContent "<h2>Section A</h2>" .PARAMETER PostContent Raw HTML content to be placed after the wrapping <DIV> tag. For example: -PostContent "<hr />" .PARAMETER MakeHiddenSection Used in conjunction with -PreContent. Adding this switch, which needs no value, turns your -PreContent into clickable report section header. The section will be hidden by default, and clicking the header will toggle its visibility. When using this parameter, consider adding a symbol to your -PreContent that helps indicate this is an expandable section. For example: -PreContent '<h2>♦ My Section</h2>' If you use -MakeHiddenSection, you MUST provide -PreContent also, or the hidden section will not have a section header and will not be visible. .PARAMETER MakeTableDynamic When using "-As Table", makes the table dynamic. Will be ignored if you use "-As List". Dynamic tables are sortable, searchable, and are paginated. You should not use even/odd styling with tables that are made dynamic. Dynamic tables automatically have their own even/odd styling. You can apply CSS classes named ".odd" and ".even" in your CSS to style the even/odd in a dynamic table. .EXAMPLE $fragment = Get-WmiObject -Class Win32_LogicalDisk | Select-Object -Property PSComputerName,DeviceID,FreeSpace,Size | ConvertTo-HTMLFragment -EvenRowClass 'even' ` -OddRowClass 'odd' ` -PreContent '<h2>Disk Report</h2>' ` -MakeHiddenSection ` -MakeTableDynamic You will usually save fragments to a variable, so that multiple fragments (each in its own variable) can be passed to ConvertTo-EnhancedHTML. .NOTES Consider adding the following to your CSS when using dynamic tables (replace the * with .): *paginate_enabled_next, .paginate_enabled_previous { cursor:pointer; border:1px solid #222222; background-color:#dddddd; padding:2px; margin:4px; border-radius:2px; } *paginate_disabled_previous, .paginate_disabled_next { color:#666666; cursor:pointer; background-color:#dddddd; padding:2px; margin:4px; border-radius:2px; } *dataTables_info { margin-bottom:4px; } This applies appropriate coloring to the next/previous buttons, and applies a small amount of space after the dynamic table. If you choose to make sections hidden (meaning they can be shown and hidden by clicking on the section header), consider adding the following to your CSS (replace the * with .): *sectionheader { cursor:pointer; } *sectionheader:hover { color:red; } This will apply a hover-over color, and change the cursor icon, to help visually indicate that the section can be toggled. #> [CmdletBinding()] param( [Parameter(Mandatory=$True,ValueFromPipeline=$True)] [object[]]$InputObject, [string]$EvenRowCssClass, [string]$OddRowCssClass, [string]$TableCssID, [string]$DivCssID, [string]$DivCssClass, [string]$TableCssClass, [ValidateSet('List','Table')] [string]$As = 'Table', [object[]]$Properties = '*', [string]$PreContent, [switch]$MakeHiddenSection, [switch]$MakeTableDynamic, [string]$PostContent ) BEGIN { <# Accumulate output in a variable so that we don't produce an array of strings to the pipeline, but instead produce a single string. #> $out = '' <# Add the section header (pre-content). If asked to make this section of the report hidden, set the appropriate code on the section header to toggle the underlying table. Note that we generate a GUID to use as an additional ID on the <div>, so that we can uniquely refer to it without relying on the user supplying us with a unique ID. #> Write-Verbose "Precontent" if ($PSBoundParameters.ContainsKey('PreContent')) { if ($PSBoundParameters.ContainsKey('MakeHiddenSection')) { [string]$tempid = [System.Guid]::NewGuid() $out += "<span class=`"sectionheader`" onclick=`"`$('#$tempid').toggle(500);`">$PreContent</span>`n" } else { $out += $PreContent $tempid = '' } } <# The table will be wrapped in a <div> tag for styling purposes. Note that THIS, not the table per se, is what we hide for -MakeHiddenSection. So we will hide the section if asked to do so. #> Write-Verbose "DIV" if ($PSBoundParameters.ContainsKey('DivCSSClass')) { $temp = " class=`"$DivCSSClass`"" } else { $temp = "" } if ($PSBoundParameters.ContainsKey('MakeHiddenSection')) { $temp += " id=`"$tempid`" style=`"display:none;`"" } else { $tempid = '' } if ($PSBoundParameters.ContainsKey('DivCSSID')) { $temp += " id=`"$DivCSSID`"" } $out += "<div $temp>" <# Create the table header. If asked to make the table dynamic, we add the CSS style that ConvertTo-EnhancedHTML will look for to dynamic-ize tables. #> Write-Verbose "TABLE" $_TableCssClass = '' if ($PSBoundParameters.ContainsKey('MakeTableDynamic') -and $As -eq 'Table') { $_TableCssClass += 'enhancedhtml-dynamic-table ' } if ($PSBoundParameters.ContainsKey('TableCssClass')) { $_TableCssClass += $TableCssClass } if ($_TableCssClass -ne '') { $css = "class=`"$_TableCSSClass`"" } else { $css = "" } if ($PSBoundParameters.ContainsKey('TableCSSID')) { $css += "id=`"$TableCSSID`"" } else { if ($tempid -ne '') { $css += "id=`"$tempid`"" } } $out += "<table $css>" <# We're now setting up to run through our input objects and create the table rows #> $fragment = '' $wrote_first_line = $false $even_row = $false if ($properties -eq '*') { $all_properties = $true } else { $all_properties = $false } } PROCESS { foreach ($object in $inputobject) { Write-Verbose "Processing object" $datarow = '' $headerrow = '' <# Apply even/odd row class. Note that this will mess up the output if the table is made dynamic. That's noted in the help. #> if ($PSBoundParameters.ContainsKey('EvenRowCSSClass') -and $PSBoundParameters.ContainsKey('OddRowCssClass')) { if ($even_row) { $row_css = $OddRowCSSClass $even_row = $false Write-Verbose "Even row" } else { $row_css = $EvenRowCSSClass $even_row = $true Write-Verbose "Odd row" } } else { $row_css = '' Write-Verbose "No row CSS class" } <# If asked to include all object properties, get them. #> if ($all_properties) { $properties = $object | Get-Member -MemberType Properties | Select -ExpandProperty Name } <# We either have a list of all properties, or a hashtable of properties to play with. Process the list. #> foreach ($prop in $properties) { Write-Verbose "Processing property" $name = $null $value = $null $cell_css = '' <# $prop is a simple string if we are doing "all properties," otherwise it is a hashtable. If it's a string, then we can easily get the name (it's the string) and the value. #> if ($prop -is [string]) { Write-Verbose "Property $prop" $name = $Prop $value = $object.($prop) } elseif ($prop -is [hashtable]) { Write-Verbose "Property hashtable" <# For key "css" or "cssclass," execute the supplied script block. It's expected to output a class name; we embed that in the "class" attribute later. #> if ($prop.ContainsKey('cssclass')) { $cell_css = $Object | ForEach $prop['cssclass'] } if ($prop.ContainsKey('css')) { $cell_css = $Object | ForEach $prop['css'] } <# Get the current property name. #> if ($prop.ContainsKey('n')) { $name = $prop['n'] } if ($prop.ContainsKey('name')) { $name = $prop['name'] } if ($prop.ContainsKey('label')) { $name = $prop['label'] } if ($prop.ContainsKey('l')) { $name = $prop['l'] } <# Execute the "expression" or "e" key to get the value of the property. #> if ($prop.ContainsKey('e')) { $value = $Object | ForEach $prop['e'] } if ($prop.ContainsKey('expression')) { $value = $Object | ForEach $prop['expression'] } <# Make sure we have a name and a value at this point. #> if ($name -eq $null -or $value -eq $null) { Write-Error "Hashtable missing Name and/or Expression key" } } else { <# We got a property list that wasn't strings and wasn't hashtables. Bad input. #> Write-Warning "Unhandled property $prop" } <# When constructing a table, we have to remember the property names so that we can build the table header. In a list, it's easier - we output the property name and the value at the same time, since they both live on the same row of the output. #> if ($As -eq 'table') { Write-Verbose "Adding $name to header and $value to row" $headerrow += "<th>$name</th>" $datarow += "<td$(if ($cell_css -ne '') { ' class="'+$cell_css+'"' })>$value</td>" } else { $wrote_first_line = $true $headerrow = "" $datarow = "<td$(if ($cell_css -ne '') { ' class="'+$cell_css+'"' })>$name :</td><td$(if ($cell_css -ne '') { ' class="'+$cell_css+'"' })>$value</td>" $out += "<tr$(if ($row_css -ne '') { ' class="'+$row_css+'"' })>$datarow</tr>" } } <# Write the table header, if we're doing a table. #> if (-not $wrote_first_line -and $as -eq 'Table') { Write-Verbose "Writing header row" $out += "<tr>$headerrow</tr><tbody>" $wrote_first_line = $true } <# In table mode, write the data row. #> if ($as -eq 'table') { Write-Verbose "Writing data row" $out += "<tr$(if ($row_css -ne '') { ' class="'+$row_css+'"' })>$datarow</tr>" } } } END { <# Finally, post-content code, the end of the table, the end of the <div>, and write the final string. #> Write-Verbose "PostContent" if ($PSBoundParameters.ContainsKey('PostContent')) { $out += "`n$PostContent" } Write-Verbose "Done" $out += "</tbody></table></div>" Write-Output $out } } |