[FEATURE] - no mechanism to handle panic in jobs for v2

[trungdlp-wolffun]

Is your feature request related to a problem? Please describe

package main

import (
	"fmt"
	"time"

	"github.com/go-co-op/gocron/v2"
)

func main() {
	// create a scheduler
	s, err := gocron.NewScheduler()
	if err != nil {
		// handle error
	}

	// add a job to the scheduler
	j, err := s.NewJob(
		gocron.DurationJob(
			10*time.Second,
		),
		gocron.NewTask(
			func(a string, b int) {
				panic("panic occurred")
			},
			"hello",
			1,
		),
	)
	if err != nil {
		// handle error
	}
	// each job has a unique id
	fmt.Println(j.ID())

	// start the scheduler
	s.Start()

	// block until you are ready to shut down
	select {
	case <-time.After(time.Minute):
	}

	// when you're done, shut it down
	err = s.Shutdown()
	if err != nil {
		// handle error
	}
}

Describe the solution you'd like

Describe alternatives you've considered

Additional context

Related issue: #332

[khall-ms]

Have you looked at https://pkg.go.dev/github.com/go-co-op/gocron/v2#example-AfterJobRunsWithError

[khall-ms]

An example would look like:

package main

import (
	"fmt"
	"time"

        "github.com/google/uuid"
	"github.com/go-co-op/gocron/v2"
)

func main() {
	// create a scheduler
	s, err := gocron.NewScheduler()
	if err != nil {
		// handle error
	}

	// add a job to the scheduler
	j, err := s.NewJob(
		gocron.DurationJob(
			10*time.Second,
		),
		gocron.NewTask(
			func(a string, b int) {
				panic("panic occurred")
			},
			"hello",
			1,
		),
                gocron.WithEventListeners(
		    gocron.AfterJobRunsWithError(
			    func(jobID uuid.UUID, jobName string, err error) {
				// do something when the job returns an error
                    fmt.Printf("%c+", err)
			    },
		    ),
              ),
	)
	if err != nil {
		// handle error
	}
	// each job has a unique id
	fmt.Println(j.ID())

	// start the scheduler
	s.Start()

	// block until you are ready to shut down
	select {
	case <-time.After(time.Minute):
	}

	// when you're done, shut it down
	err = s.Shutdown()
	if err != nil {
		// handle error
	}
}

[trungdlp-wolffun]

@khall-ms, I think it's not working. I tested with your code, and when a panic occurred, the main program also exited.

f895a01f-27dd-40c0-be06-03e430b116d3
panic: panic occurred

... stack trace

Process finished with the exit code 2

[trungdlp-wolffun]

Currently I must recover manual

gocron.NewTask(
  func(a string, b int) {
    defer func() {
	    if r := recover(); r != nil {
	       fmt.Println("recovered from panic")
	    }
    }()
    
    
    panic("panic occurred")
  },
  "hello",
  1,
),
		

[trungdlp-wolffun]

@JohnRoesler please take a look 🥹 and can you guide me on how to fix that, like a senior developer helping a junior developer?

[khall-ms]

Hey mate,

I think its been a while since I used this library. I'm getting the same error, so either somethings changed or I could have been confusing it with go pond.

I don't have time to go down the rabbit hole atm so you could use this:

package main

import (
    "fmt"
    "time"

    "github.com/google/uuid"
    "github.com/go-co-op/gocron/v2"
)

func main() {
    // create a scheduler
    s, err := gocron.NewScheduler()
    if err != nil {
        // handle error
    }

    // add a job to the scheduler
    j, err := s.NewJob(
        gocron.DurationJob(
            5*time.Second,
        ),
        gocron.NewTask(
            func(a string, b int) (err error) {
                defer func() {
                    if r := recover(); r != nil {
                        err = fmt.Errorf("Recovered. Error:%v", r)
                    }
                }()
                if err != nil {
                    return err
                }

                // Your code here
                panic("panic occurred")

            },
            "hello",
            1,
        ),
        gocron.WithEventListeners(
            gocron.AfterJobRunsWithError(
                func(jobID uuid.UUID, jobName string, err error) {
                // do something when the job returns an error
                    fmt.Printf("Job %s had error %s\n", jobID, err)
                },
            ),
        ),
    )
    if err != nil {
        // handle error
    }
    // each job has a unique id
    fmt.Println(j.ID())

    // start the scheduler
    s.Start()

    // block until you are ready to shut down
    select {
    case <-time.After(time.Minute):
    }

    // when you're done, shut it down
    err = s.Shutdown()                                                                                                                                                                       if err != nil {
        // handle error
    }
}

[trungdlp-wolffun]

I think we should add a panic handler like this: #335 @khall-ms

[khall-ms]

That would be ideal 💯

[JohnRoesler]

Thanks for the collaborative discussion on this. 🚀 I requested changes on the PR to fit the design with how other "event listener" things are handled.

I'm curious, what type of logic are you running in your jobs that is resulting in panics? I haven't added panic handling support because my general philosophy is that your code should only ever panic during development.

[trungdlp-wolffun]

@JohnRoesler, sometimes I deal with nil pointers or division by zero errors