workflows/eval: skip on ready_for_review

We don't need to run full eval when undrafting a PR. We already have an
eval result, so we can use that to do the maintainer pings.

We need to wait for eval to finish first, but we're already half there
because of how we're waiting for the artifact to appear. Since the
ready_for_review case is triggered in a different workflow run, we'll
need to fetch the ID of the relevant Eval workflow first, though.

.github/workflows/eval.yml

@@ -4,8 +4,8 @@ on:
   pull_request:
     paths:
       - .github/workflows/eval.yml
+      - .github/workflows/reviews.yml # needs eval results from the same event type
   pull_request_target:
-    types: [opened, ready_for_review, synchronize, reopened]
   push:
     # Keep this synced with ci/request-reviews/dev-branches.txt
     branches:

.github/workflows/reviewers.yml

@@ -4,6 +4,11 @@
 name: Reviewers
 
 on:
+  pull_request:
+    paths:
+      - .github/workflows/reviewers.yml
+  pull_request_target:
+    types: [ready_for_review]
   workflow_call:
 
 permissions: {}
@@ -39,16 +44,29 @@ jobs:
           permission-members: read
           permission-pull-requests: write
 
+
+      # In the regular case, this workflow is called via workflow_call from the eval workflow directly.
+      # In the more special case, when a PR is undrafted an eval run will have started already.
       - name: Wait for comparison to be done
         uses: actions/github-script@60a0d83039c74a4aee543508d2ffcb1c3799cdea # v7.0.1
         with:
           script: |
-            // Waiting 24 * 5 sec = 2 min. max.
-            for (let i = 0; i < 24; i++) {
+            const run_id = (await github.rest.actions.listWorkflowRuns({
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              workflow_id: 'eval.yml',
+              event: context.eventName,
+              head_sha: context.payload.pull_request.head.sha
+            })).data.workflow_runs[0].id
+
+            // Waiting 120 * 5 sec = 10 min. max.
+            // The extreme case is an Eval run that just started when the PR is undrafted.
+            // Eval takes max 5-6 minutes, normally.
+            for (let i = 0; i < 120; i++) {
               const result = await github.rest.actions.listWorkflowRunArtifacts({
                 owner: context.repo.owner,
                 repo: context.repo.repo,
-                run_id: context.runId,
+                run_id,
                 name: 'comparison'
               })
               if (result.data.total_count > 0) return