Just an example of simple JSON parsing with/without JMESPath

webhook.js

// JSON webhook
[
  {
    "msys" => {
      "message_event" => {
        "type" => "spam_complaint",
        "rcpt_to" => "user@example.com"
      }
    }
  },
  { ... }
]

with_jmespath.rb

# getting all email addresses for events with type "spam_complaint"

# with JMESPath
jmespath_query = "[].msys.message_event | [?type=='spam_complaint'].rcpt_to"
spam_complaint_emails = JMESPath.search(jmespath_query, payload)

without_jmespath.rb

# getting all email addresses for events with type "spam_complaint"

# without JMESPath
spam_complaint_emails = payload
  .map { |el|
    el['msys']
  }.select { |el|
    el.fetch('message_event', {}).fetch('type', nil) == 'spam_complaint'
  }.map { |el|
    el['message_event']['rcpt_to']
  }

Is this the JavaScript equivalent?

spam_complaint_emails = arr.filter(o => o.msys.message_event.type=='spam_complaint')

Well, given

var arr = [
  {
    "msys" : {
      "message_event" : {
        "type" : "spam_complaint",
        "rcpt_to" : "user@example.com"
      }
    }
  }
]

I think the equivalent in JavaScript would be

spam_complaint_emails = arr.map(item => item.msys.message_event)
  .filter(me => me.type=='spam_complaint')
  .map(me => me.rcpt_to);

or

spam_complaint_emails = arr.filter(item => item
  .msys.message_event.type=='spam_complaint')
  .map(item=>item.msys.message_event.rcpt_to);

Which path you choose as better is likely to depend on where additional attributes from the might be desired later.

Given I would probably do the same thing in Perl with maps and gresp, and I understand Ruby has all those same mechanisms, I'm not sure why the overly complex Ruby code, unless Ruby throws an error on invalid hash access?. In Perl I would just turn autovivification off and get eh undef (nil) as expected which evals to false for the grep.