PSSpeech.psm1
|
function Get-SpeechToken { <# .SYNOPSIS Get OAuth token for authorization to Azure Cognitive Services. .DESCRIPTION This function uses Invoke-RestMethod to get a bearer token that can be used in the Authorization header when calling Azure Cognitive Services. This requires access to an Azure subscription and API key for the speech service. .EXAMPLE PS C:\> Get-SpeechToken -Region <region> -Key <apikey> This example gets a token using the provided key and region. .INPUTS None. .OUTPUTS [psobject] .NOTES Key should probably be a secure string, update once secrets management module is released. The token is stored in $script:SpeechToken and can be retrieved with Get-SpeechTokenResult #> [CmdletBinding(HelpUri = 'https://ntsystems.it/PowerShell/Get-SpeechToken/')] param ( [Parameter(mandatory=$true)] [ValidateSet("centralus", "eastus", "eastus2", "northcentralus", "southcentralus", "westcentralus", "westus", "westus2", "canadacentral", "brazilsouth", "eastasia", "southeastasia", "australiaeast", "centralindia", "japaneast", "japanwest", "koreacentral", "northeurope", "westeurope", "francecentral", "uksouth")] [string] $Region, [Parameter(Mandatory=$true)] [ValidateNotNullOrEmpty()] [string] $Key ) $FetchTokenHeader = @{ 'Content-type'='application/x-www-form-urlencoded'; 'Content-Length'= '0'; 'Ocp-Apim-Subscription-Key' = $Key } $script:SpeechToken = New-Object -TypeName psobject -Property (@{ TimeStamp = Get-Date Token = Invoke-RestMethod -Method POST -Uri "https://$region.api.cognitive.microsoft.com/sts/v1.0/issueToken" -Headers $FetchTokenHeader Region = $Region }) } function Get-SpeechTokenResult { return $script:SpeechToken } function Get-SpeechVoicesList { <# .SYNOPSIS Get a list of available voices from the speech service. .DESCRIPTION This function uses Invoke-RestMethod to get a list of available voices from the Azure Cognitive Services Speech Service. Use the Token parameter to specify a token created with Get-SpeechToken and use the Region parameter to specify a region other than the default westeurope. If the Token parameter is not specified, the global variable created by Save-SpeechToken is used. .EXAMPLE PS C:\> Get-SpeechVoicesList This example gets a list of available voices. .INPUTS None. .OUTPUTS [psobject] .NOTES General notes #> [CmdletBinding(HelpUri = 'https://ntsystems.it/PowerShell/Get-SpeechVoicesList/')] param ( $token = $script:SpeechToken.Token ) $AuthHeader = @{ 'Content-type' = 'application/ssml+xml'; 'Authorization' = "Bearer $Token"; 'Content-Length'= '0'; } $region = $script:SpeechToken.region $voices = Invoke-RestMethod -uri "https://$region.tts.speech.microsoft.com/cognitiveservices/voices/list" -Headers $AuthHeader -Method Get $voices } function Convert-TextToSpeech { <# .SYNOPSIS Convert a string to audio using Azure Cognitive Services. .DESCRIPTION This function uses Invoke-RestMethod to call the Azure Cognitive Service Speech Service API, convert a string to speech, and save the resulting audio to a file. .EXAMPLE PS C:\> Convert-TextToSpeech -Text "Hi, this is a test." -Path test.mp3 This example converts the string "Hi, this is a test." to speech and saves the audio to the test.mp3 file. .INPUTS None. .OUTPUTS None. #> [CmdletBinding(HelpUri = 'https://ntsystems.it/PowerShell/Convert-TextToSpeech/')] param ( [Parameter(Mandatory)] [ValidateNotNullOrEmpty()] [String] $Text, [Parameter(Mandatory)] [System.IO.FileInfo] $Path, [Parameter()] [ValidateScript({Get-SpeechVoicesList | Select-Object ShortName})] [string] $Voice = 'en-GB-LibbyNeural', [Parameter()] [ValidateSet("raw-16khz-16bit-mono-pcm","audio-16khz-128kbitrate-mono-mp3","audio-16khz-32kbitrate-mono-mp3","audio-24khz-96kbitrate-mono-mp3","audio-24khz-48kbitrate-mono-mp3","audio-24khz-160kbitrate-mono-mp3","audio-16khz-64kbitrate-mono-mp3")] [string] $OutputFormat = "audio-16khz-32kbitrate-mono-mp3" ) $SpokenVoice = Get-SpeechVoicesList | ? {$_.ShortName -eq $Voice} $token = $script:SpeechToken.Token $AuthHeader = @{ 'Content-type' = 'application/ssml+xml' 'Authorization' = "Bearer $token" 'X-Microsoft-OutputFormat' = $OutputFormat 'User-Agent' = "powershell" } $region = $script:SpeechToken.region # build the ssml xml #<voice xml:lang='en-US' xml:gender='Female' name='en-US-AriaRUS'> #will assume language and voice are the same [xml]$xml = "<speak version='1.0' xml:lang='en-GB'><voice xml:lang='en-GB' xml:gender='Female' name='$Voice'></voice></speak>" $xml.speak.voice.InnerText=$text $xml.speak.lang = $spokenvoice.Locale $xml.speak.voice.lang = $spokenvoice.Locale $xml.speak.voice.gender = $spokenvoice.Gender $xml.speak.voice.name = $spokenvoice.ShortName # send to speech service and save output in file Invoke-RestMethod -Uri "https://$region.tts.speech.microsoft.com/cognitiveservices/v1" -Headers $AuthHeader -Method Post -Body $xml -OutFile $Path -Verbose } |