Check only new links in pull requests

Author: Soumoditya

.github/workflows/test.yml

@@ -25,18 +25,11 @@ jobs:
     if: ${{ github.repository_owner == 'pypa' || github.event_name != 'schedule' }}
     runs-on: ubuntu-latest
     timeout-minutes: 20
-    continue-on-error: >-
-      ${{ fromJSON(matrix.continue-on-error) }}
     strategy:
       matrix:
         noxenv:
         - build
-        continue-on-error:
-        - false
-        include:
-        - noxenv: linkcheck
-          continue-on-error: >-  # Don't block PRs on linkcheck unrelated failures
-            ${{ toJSON(github.event_name == 'pull_request') }}
+        - linkcheck
 
     steps:
       - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
@@ -50,17 +43,44 @@ jobs:
           cache: 'pip'
           cache-dependency-path: 'requirements.txt'
 
+      - name: Restore linkcheck report
+        if: matrix.noxenv == 'linkcheck' && github.event_name == 'pull_request'
+        id: linkcheck-cache
+        uses: actions/cache/restore@5a3ec84eff668545956fd18022155c47e93e2684 # v4.2.3
+        with:
+          path: build/output.json
+          key: linkcheck-${{ github.event.pull_request.base.sha }}
+          restore-keys: |
+            linkcheck-
+
       - name: Install dependencies
         run: |
           python -m pip install --upgrade nox virtualenv
 
       - name: Nox ${{ matrix.noxenv }}
+        id: nox
+        continue-on-error: ${{ matrix.noxenv == 'linkcheck' && github.event_name == 'pull_request' }}
         env:
           # Authenticate github.com requests during linkcheck to avoid rate limits.
           GITHUB_TOKEN: ${{ matrix.noxenv == 'linkcheck' && github.token || '' }}
         run: |
           python -m nox -s ${{ matrix.noxenv }}
 
+      - name: Fail on newly added broken links
+        if: >-
+          matrix.noxenv == 'linkcheck' &&
+          github.event_name == 'pull_request' &&
+          steps.linkcheck-cache.outputs.cache-hit != '' &&
+          steps.nox.outcome == 'failure'
+        run: exit 1
+
+      - name: Save linkcheck report
+        if: matrix.noxenv == 'linkcheck' && github.event_name != 'pull_request' && success()
+        uses: actions/cache/save@5a3ec84eff668545956fd18022155c47e93e2684 # v4.2.3
+        with:
+          path: build/output.json
+          key: linkcheck-${{ github.sha }}
+
 
   check:
     # This job does nothing and is only used for the branch protection

source/conf.py

@@ -1,8 +1,10 @@
 # -- Project information ---------------------------------------------------------------
 # https://www.sphinx-doc.org/en/master/usage/configuration.html#project-information
 
+import json
 import os
 import pathlib
+import re
 import sys
 
 _ROOT = pathlib.Path(__file__).resolve().parent.parent
@@ -167,6 +169,21 @@
 ]
 linkcheck_retries = 2
 linkcheck_timeout = 30
+
+# On pull requests, ignore links that were already checked on the default
+# branch. The cached JSONL report is refreshed after successful default-branch
+# linkcheck runs, so newly introduced links remain blocking.
+if os.getenv("GITHUB_EVENT_NAME") == "pull_request":
+    previous_linkcheck_report = pathlib.Path("build/output.json")
+    if previous_linkcheck_report.exists():
+        for line in previous_linkcheck_report.read_text(encoding="utf-8").splitlines():
+            try:
+                uri = json.loads(line)["uri"]
+            except (json.JSONDecodeError, KeyError, TypeError):
+                continue
+            if isinstance(uri, str):
+                linkcheck_ignore.append(rf"^{re.escape(uri)}$")
+
 # Ignore anchors for common targets when we know they likely won't be found
 linkcheck_anchors_ignore_for_url = [
     # GitHub synthesises anchors in JavaScript, so Sphinx can't find them in the HTML