Class: PartyResearch::Orchestrator
- Inherits:
-
Object
- Object
- PartyResearch::Orchestrator
- Defined in:
- app/services/party_research/orchestrator.rb
Class Method Summary collapse
Instance Method Summary collapse
- #call ⇒ Object
-
#initialize(run, on_progress: nil) ⇒ Orchestrator
constructor
A new instance of Orchestrator.
Constructor Details
#initialize(run, on_progress: nil) ⇒ Orchestrator
Returns a new instance of Orchestrator.
28 29 30 31 |
# File 'app/services/party_research/orchestrator.rb', line 28 def initialize(run, on_progress: nil) @run = run @on_progress = on_progress end |
Class Method Details
.call(run:, on_progress: nil) ⇒ Object
24 25 26 |
# File 'app/services/party_research/orchestrator.rb', line 24 def self.call(run:, on_progress: nil) new(run, on_progress: on_progress).call end |
Instance Method Details
#call ⇒ Object
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 |
# File 'app/services/party_research/orchestrator.rb', line 33 def call # Sidekiq may retry the job after a partial failure. If the run # already reached a terminal state, don't redo the work — the # previous attempt's findings are persisted and the reviewer can # rerun explicitly if needed. If it died mid-run (state == # 'running'), pick up where we left off without re-flipping state # or re-superseding prior findings. return if @run.state.in?(%w[completed failed]) if @run.state == 'queued' @run.start! supersede_prior_pending_findings seed_adapter_progress! end adapter_names = @run.adapters_list total = adapter_names.size adapter_errors = [] adapter_names.each_with_index do |name, idx| # Idempotent resume guard: if the run is already partway through # (Sidekiq retry, replayed worker), skip adapters that already # reached a terminal state in the previous attempt. Re-running # `completed` / `skipped` ones would duplicate pending findings # and burn paid API credits twice. `failed` and `running` still # rerun — the former because retry is the whole point, the # latter because we can't tell if findings were persisted before # the crash (Sidekiq's job-uniqueness keeps two live workers off # the same run, so this isn't a concurrent-claim hazard). prior_state = @run.adapter_progress&.dig(name, 'state') if prior_state.in?(%w[completed skipped]) @on_progress&.call(idx, total, name) next end @on_progress&.call(idx, total, name) mark_adapter(name, state: 'running', started_at: Time.current.iso8601) begin findings_count = run_adapter(name) mark_adapter(name, state: findings_count == :skipped ? 'skipped' : 'completed', findings: findings_count == :skipped ? 0 : findings_count, finished_at: Time.current.iso8601) rescue StandardError => e Rails.logger.warn("[PartyResearch] adapter=#{name} failed: #{e.class}: #{e.}") adapter_errors << "#{name}: #{e.}" mark_adapter(name, state: 'failed', error: e..truncate(400), finished_at: Time.current.iso8601) end end if adapter_errors.size == adapter_names.size && adapter_errors.any? @run.fail!(adapter_errors.join("\n")) else @run.update!(error_message: adapter_errors.join("\n")) if adapter_errors.any? @run.complete! end rescue StandardError => e @run.fail!(e.) raise end |