Start/Stop trigger: catch non-terminating Az errors so retries and warnings work #435 #508

Part of #435.

Problem

Start-AzDataFactoryV2Trigger and Stop-AzDataFactoryV2Trigger report failures as non-terminating errors. With the default $ErrorActionPreference = 'Continue', the try/catch in private/Start-Trigger.ps1 and private/Stop-Trigger.ps1 never sees them:

• no retry, no "Failed starting/stopping trigger after N attempts" message, no warning (-WarningVariable stays empty)
• Stop-Trigger returns $true even when the trigger was not stopped, so the Adds TryCatch to StopTrigger, continuing if fails #404 retry (v1.19.0) only works when the caller uses -ErrorAction Stop (e.g. Azure DevOps AzurePowerShell@5 default)

Verified against Azure: the error is caught only with -ErrorAction Stop.

Expected

• Add -ErrorAction Stop to the Az calls inside try, so retries and warnings work whatever the caller's preference is.
• Do not retry permanent errors (e.g. Missing or invalid pipeline references), only temporary ones (provisioning, concurrent operations).
• Start-Trigger returns $true/$false like Stop-Trigger; Start-Triggers reports all triggers that failed to start.
• The deployment carries on: a failed start or stop is a warning, not an error. Details come from the Azure API message; the precise validation belongs to Test-AdfCode (see the sibling sub-issue).

Tests

Mock the Az cmdlets to write a non-terminating error (not throw), which is how the real cmdlets behave. Existing mocks that throw hid this bug.