|
My console app does not use async, and I use What is the correct pattern to use for detecting execution errors when calling Do I always wrap in a Do I test for Or do I test for If I use a cancellation token, will it trigger IsFaulted or throw an exception? E.g. a variant of the wrapper being used: public bool Execute(Command command, out BufferedCommandResult result)
{
// TODO: do we need to wrap in try/catch?
// Start execution to get the process id or error
result = null;
CommandTask<BufferedCommandResult> task = command.ExecuteBufferedAsync();
if (task.Task.IsFaulted)
{
Log.Error(
"Error executing {ToolType} : {Parameters}",
GetToolType(),
command.Arguments
);
// TODO: Will IsFaulted always have an exception?
Log.Error("{Error}", task.Task.Exception);
return false;
}
Log.Information(
"Executing {ToolType} : ProcessId: {ProcessId}, {Parameters}",
GetToolType(),
task.ProcessId,
command.Arguments
);
// Wait for the process to complete
result = task.Task.Result;
// TODO: Is IsFaulted always true if the process fails or is cancelled?
return !task.Task.IsFaulted;
} |
Replies: 1 comment 2 replies
|
If you call |
|
[Edit] Unfortunately it is a slippery slope to make things async, will require large refactor to have the async make it from main to where this code is being executed, so going async is not an option, I have to make do with just "properly" handle CliWrap's async calls. |
|
|
If you call
GetAwaiter().GetResult()instead of.Result, it should propagate exceptions correctly. But I would recommend just making everything async instead. Console app's main method can be async as well.