Allow jobs to declare theirs schedule #715
Conversation
|
@jeremy sorry for pinging you directly, do you have any though about this? Thanks! Btw as we are here 😄 I work for a company (Nozomi Networks - OT & IoT Security and Visibility) which would like to contribute and maintain this project (and moving on to resque itselt) as we heavily use ruby and rely on resque. Do you have any suggestion for us on how we could help? Thanks! |
|
@and9000 Would love your help maintaining this! The first thing we'll need to do is get CI working again. Could you submit a separate PR to move CI to GH actions? Here's what we did for the core resque gem: https://github.com/resque/resque/blob/master/.github/workflows/ci.yml |
|
Sure, I'll start working on CI 👍 |
lib/resque/scheduler/job.rb
Outdated
| def queue(value = nil) | ||
| return @queue ||= nil if value.nil? | ||
| @queue = value | ||
| end |
There was a problem hiding this comment.
The standard Resque convention is to use @queue without a class method. I think we should continue to use that convention if we can.
There was a problem hiding this comment.
You are right. Fixed.
lib/resque/scheduler.rb
Outdated
| rescue NameError | ||
| log! "Can't load file #{file}" | ||
| end | ||
| load_schedule_job( |
There was a problem hiding this comment.
Is it possible to run load_schedule_job when the schedule is set up on the class? If we used a resque_schedule method on the class-level as suggested below, do you think we could eliminate the auto-load logic from the PR and instead run load_schedule_job as part of resque_schedule.
Autoloading and scheduling on the class-level feel like two different concerns to me, and it's not clear why we should couple them in this case.
Additionally, autoloading feels like something that makes sense on the core resque level as opposed to the scheduler plugin.
There was a problem hiding this comment.
I've moved schedule logic into class method resque_schedule.
I'm still stuck thinking on where the autoloading feature should be: in principle yes it shouldn't be here but on the application level (not even resque) as the app is responsible to load stuff. But this mean we should add a big warning "in order to self-schedule-jobs to work please remember to require your jobs". But this is exactly what auto_load is doing so my doubt is: why don't make that requirement explicit?
lib/resque/scheduler/job.rb
Outdated
| end | ||
|
|
||
| module ClassMethods | ||
| def cron(value = nil) |
There was a problem hiding this comment.
If feels like it could be cleaner to have a single class method resque_schedule which has a bunch of optional args (cron:, every:, etc). I don't love the idea of adding a bunch of methods that aren't clearly ties to resque scheduler (i.e. description could easily be a method users of this gem have defined on their class).
What do you think?
There was a problem hiding this comment.
Yes code looks cleaner this way 😄
* removed auto_load fields in favour of resque_schedule method
Hello,
I've implemented the possibility for the job to declare it's schedule so it's self contained.