Fix CI for large PRs (#4461)

cli/cli#10712 resulted in our Runtime PR
failing.

I debated using Git to get the information, but didn't because our
`fetch-depth` is 1 and we would need to fetch more, which isn't great
with our repository's size. I also don't want to use
`tj-actions/changed-files` due to supply chain security and because it
also under-the-hood Git fetches more commits.

So, we now use the API directly. The REST API has overly verbose
information with the actual patch for each file, so we use GraphQL.

Author: Eric-Arellano

.github/workflows/dev-docs-sync.yml

@@ -25,7 +25,7 @@ jobs:
       - name: Set up Node.js
         uses: actions/setup-node@v4
         with:
-          node-version: 18
+          node-version: 24
       - name: Install dependencies
         run: |
           sudo apt-get update

.github/workflows/main-branch-preview.yml

@@ -36,7 +36,7 @@ jobs:
       - uses: actions/checkout@v4
       - uses: actions/setup-node@v4
         with:
-          node-version: 22
+          node-version: 24
       - uses: actions/setup-python@v5
         with:
           python-version: "3.11"

.github/workflows/main.yml

@@ -16,16 +16,41 @@ jobs:
   decide-jobs:
     name: Decide which jobs to run
     runs-on: ubuntu-latest
-    # We don't run these tests in the merge queue
-    if: github.repository_owner == 'Qiskit' && github.event_name == 'pull_request'
+    if: github.repository_owner == 'Qiskit'
     outputs:
+      all-changed-files: ${{ steps.changed-files.outputs.ALL_CHANGED_FILES }}
       run-notebook-tester: ${{ steps.decide.outputs.RUN_NOTEBOOK_TESTER }}
       run-api-checks: ${{ steps.decide.outputs.RUN_API_CHECKS }}
     steps:
+      - uses: actions/checkout@v4
+      - name: Set up Node.js
+        uses: actions/setup-node@v4
+        with:
+          node-version: 24
+      - name: Determine all changed files
+        id: changed-files
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+        run: |
+          # GitHub API doesn't give PR information in the merge_group event, but
+          # we can get the PR number from the branch name
+          if [[ "${{ github.event_name }}" == 'merge_group' ]]; then
+            BRANCH_NAME="$(git rev-parse --abbrev-ref HEAD)"
+            PR_NUMBER="$(echo $BRANCH_NAME | grep -oP '(?<=gh-readonly-queue/main/pr-)\d+')"
+          else
+            PR_NUMBER="${{ github.event.number }}"
+          fi
+
+          CHANGED_FILES=$(node scripts/ci/determine-changed-files.ts "${PR_NUMBER}")
+          {
+            echo "ALL_CHANGED_FILES<<__GH_OUTPUT__"
+            printf '%s\n' "$CHANGED_FILES"
+            echo "__GH_OUTPUT__"
+          } >> "$GITHUB_OUTPUT"
       - name: Decide which jobs to run
         id: decide
         env:
-          GH_TOKEN: ${{ github.token }}
+          CHANGED_FILES: ${{ steps.changed-files.outputs.ALL_CHANGED_FILES }}
           # If changing these regexes, test them on https://regex101.com/ with the
           # following paths (you can paste the block in and uncomment it):
           # Add new tests if necessary
@@ -56,13 +81,10 @@ jobs:
           #  public/learning/image.svg
           #  .github/workflows/notebook-test-cron.yml
         run: |
-          CHANGED_FILES=$(gh pr diff -R Qiskit/documentation ${{ github.event.number }} --name-only)
-          if [ "$(echo "${CHANGED_FILES[@]}" | grep -P $NOTEBOOK_TEST_REGEX || true)" != '' ]
-          then
+          if [[ "${{ github.event_name }}" == 'pull_request' && -n "$(printf '%s\n' "${CHANGED_FILES}" | grep -P "$NOTEBOOK_TEST_REGEX")" ]]; then
             echo "RUN_NOTEBOOK_TESTER=true" >> "$GITHUB_OUTPUT"
           fi
-          if [ "$(echo "${CHANGED_FILES[@]}" | grep -P $API_CHECKS_REGEX || true)" != '' ]
-          then
+          if [[ "${{ github.event_name }}" == 'pull_request' && -n "$(printf '%s\n' "${CHANGED_FILES}" | grep -P "$API_CHECKS_REGEX")" ]]; then
             echo "RUN_API_CHECKS=true" >> "$GITHUB_OUTPUT"
           fi
 
@@ -75,7 +97,7 @@ jobs:
       - name: Set up Node.js
         uses: actions/setup-node@v4
         with:
-          node-version: 22
+          node-version: 24
       - name: Install Node.js dependencies
         run: npm ci
       - name: Install ImageMagick
@@ -110,7 +132,7 @@ jobs:
 
       - name: Get all changed content files
         env:
-          GH_TOKEN: ${{ github.token }}
+          CHANGED_FILES: ${{needs.decide-jobs.outputs.all-changed-files}}
           # If changing these regexes, test them on https://regex101.com/ with the
           # following paths (you can paste the block in and uncomment it):
           # Add new tests if necessary
@@ -126,18 +148,7 @@ jobs:
           #  scripts/nb-tester/example-notebook.ipynb
         id: changed-content-files
         run: |
-          # GitHub API doesn't give PR information in the merge_group event, but
-          # we can get the PR number from the branch name
-          if [ "${{ github.event_name }}" == 'merge_group' ]
-          then
-            BRANCH_NAME="$(git rev-parse --abbrev-ref HEAD)"
-            PR_NUMBER="$(echo $BRANCH_NAME | grep -oP '(?<=gh-readonly-queue/main/pr-)\d+')"
-          else
-            PR_NUMBER="${{ github.event.number }}"
-          fi
-
-          CHANGED_FILES=$(gh pr diff -R Qiskit/documentation $PR_NUMBER --name-only)
-          CHANGED_CONTENT_FILES=$(echo "${CHANGED_FILES[@]}" | grep -P $CONTENT_FILE_REGEX || true)
+          CHANGED_CONTENT_FILES=$(echo "${CHANGED_FILES}" | grep -P $CONTENT_FILE_REGEX || true)
           if [ "$CHANGED_CONTENT_FILES" != '' ]
           then
             echo "ANY_CHANGED=true" >> "$GITHUB_OUTPUT"
@@ -194,13 +205,12 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
-
       - name: Get changed files
         env:
-          GH_TOKEN: ${{ github.token }}
+          CHANGED_FILES: ${{needs.decide-jobs.outputs.all-changed-files}}
         run: |
           mkdir -p .github/outputs
-          gh pr diff -R Qiskit/documentation ${{ github.event.number }} --name-only > .github/outputs/changed-files.txt
+          echo "${CHANGED_FILES}" > .github/outputs/changed-files.txt
 
       - name: Check if extra linux deps needed
         id: check-deps
@@ -265,7 +275,7 @@ jobs:
       - name: Set up Node.js
         uses: actions/setup-node@v4
         with:
-          node-version: 22
+          node-version: 24
       - name: Install Node.js dependencies
         run: npm ci
       - name: Check internal links

.github/workflows/observability-sync.yml

@@ -25,17 +25,15 @@ jobs:
       - name: Set up Node.js
         uses: actions/setup-node@v4
         with:
-          node-version: 18
+          node-version: 24
       - name: Install dependencies
-        run: |
-          run: npm ci
+        run: npm ci
 
       - name: Run observability script
         run: npx tsx scripts/js/commands/generateObservabilityPage.ts
 
       - name: Get all changed docs files
         id: changed-docs-files
-
         run: |
           echo "CHANGED_FILES<<EOF" >> $GITHUB_OUTPUT
           git diff --name-only docs >> $GITHUB_OUTPUT

.github/workflows/pr-preview.yml

@@ -41,7 +41,7 @@ jobs:
       - uses: actions/checkout@v4
       - uses: actions/setup-node@v4
         with:
-          node-version: 18
+          node-version: 24
       - uses: actions/setup-python@v5
         with:
           python-version: "3.11"

.github/workflows/weekly-checks.yml

@@ -25,7 +25,7 @@ jobs:
       - name: Set up Node.js
         uses: actions/setup-node@v4
         with:
-          node-version: 18
+          node-version: 24
       - name: Install Node.js dependencies
         run: npm ci
       - name: Start local Docker preview
@@ -51,7 +51,7 @@ jobs:
       - name: Set up Node.js
         uses: actions/setup-node@v4
         with:
-          node-version: 18
+          node-version: 24
       - name: Install Node.js dependencies
         run: npm ci
       - name: Check external links
@@ -67,7 +67,7 @@ jobs:
       - name: Set up Node.js
         uses: actions/setup-node@v4
         with:
-          node-version: 18
+          node-version: 24
       - name: Install Node.js dependencies
         run: npm ci
       - name: Check outdated new pills

scripts/ci/determine-changed-files.ts

@@ -0,0 +1,68 @@
+async function fetchAllFiles(prNumber: number): Promise<string[]> {
+  const results: string[] = [];
+  let cursor: string | null = null;
+
+  while (true) {
+    const response = await fetch(`https://api.github.com/graphql`, {
+      method: "POST",
+      headers: {
+        "Content-Type": "application/json",
+        Authorization: `Bearer ${process.env.GITHUB_TOKEN}`,
+      },
+      signal: AbortSignal.timeout(10_000),
+      body: JSON.stringify({
+        query: `
+          query($pr: Int!, $cursor: String) {
+            repository(owner: "Qiskit", name: "documentation") {
+              pullRequest(number: $pr) {
+                files(first: 100, after: $cursor) {
+                  nodes {
+                    path
+                  }
+                  pageInfo {
+                    endCursor
+                    hasNextPage
+                  }
+                }
+              }
+            }
+          }
+        `,
+        variables: {
+          pr: prNumber,
+          cursor,
+        },
+      }),
+    });
+    if (!response.ok) {
+      const error = await response.text();
+      throw new Error(`Failed to fetch changed files: ${error}`);
+    }
+
+    const { data } = await response.json();
+    const files = data.repository.pullRequest.files;
+    results.push(...files.nodes.map((n: { path: string }) => n.path));
+
+    if (!files.pageInfo.hasNextPage) {
+      break;
+    }
+    cursor = files.pageInfo.endCursor;
+  }
+
+  return results;
+}
+
+async function main(): Promise<void> {
+  const prNumber = process.argv[2];
+  if (!prNumber) {
+    throw new Error("Must set the PR number as the first argument.");
+  }
+
+  const changed = await fetchAllFiles(Number(prNumber));
+  console.log(changed.join("\n"));
+}
+
+main().catch((e) => {
+  console.error(e);
+  process.exit(1);
+});