dotnet lambda deploy-function AspNetCoreWebAPI does not infer runtime from project file

[bjhogan]

Describe the bug

Create a new serverless Lambda function -

dotnet new serverless.AspNetCoreWebAPI -n AspNetCoreWebAPI

Change to the AspNetCoreWebAPI/src/AspNetCoreWebAPI

Attempt to deploy the function -

dotnet lambda deploy-function AspNetCoreWebAPI       
Amazon Lambda Tools for .NET Core applications (5.4.2)
Project Home: https://github.com/aws/aws-extensions-for-dotnet-cli, https://github.com/aws/aws-lambda-dotnet

Enter Runtime: (The runtime environment for the Lambda function)

Runtime should not be requested.

Expected Behavior

I expect the tooling to infer the runtime from the csproj file, or that the runtime in included in the aws-lambda-tools-defaults.json file.

Current Behavior

User is asked to enter the runtime.

Reproduction Steps

See description.

Possible Solution

No response

Additional Information/Context

No response

AWS .NET SDK and/or Package version used

Amazon Lambda Tools for .NET Core applications (5.4.2)

Targeted .NET Platform

.NET 6

Operating System and version

Windows 10

[ashishdhingra]

@bjhogan Thanks for reporting the issue. I'm curious that you are using serverless.AspNetCoreWebAPI template, but using the dotnet lambda deploy-function command. Shouldn't you be using dotnet lambda deploy-serverless command? Please check.

[bjhogan]

Hi @ashishdhingra, you are right! Been working on too many lambda.* templates lately and forgot the deploy-serverless.

Maybe the ticket should be - "Don't allow dotnet lambda deploy-function... to run if there is serverless.template present?
Would be a better user experience?

[ashishdhingra]

Hi @ashishdhingra, you are right! Been working on too many lambda.* templates lately and forgot the deploy-serverless.

Maybe the ticket should be - "Don't allow dotnet lambda deploy-function... to run if there is serverless.template present? Would be a better user experience?

@bjhogan It might not a good idea to add proposed validation. The respective subcommands are in place to be used with respective project types. What if we have the project for simple lambda function, but somehow user also includes serverless.template in the project accidentally (rare case, but possible)? This would add unnecessary complexity.

[bjhogan]

@ashishdhingra How about having only the serverless.template file in serverless project templates, do not include the aws-lambda-tools-defaults.json.