Bound link-check issue summaries and test offline reporting

.github/workflows/link-check.yml

@@ -40,10 +40,11 @@ jobs:
           GH_TOKEN: ${{ github.token }}
           ISSUE_TITLE: "Link check: dead or moved links found"
         run: |
+          python scripts/link_check_issue.py
           existing=$(gh issue list --state open --json number,title \
             --jq "[.[] | select(.title == \"$ISSUE_TITLE\") | .number] | first // empty")
           if [ -n "$existing" ]; then
-            gh issue comment "$existing" --body-file link-check-report.md
+            gh issue comment "$existing" --body-file link-check-issue.md
           else
-            gh issue create --title "$ISSUE_TITLE" --body-file link-check-report.md
+            gh issue create --title "$ISSUE_TITLE" --body-file link-check-issue.md
           fi

.github/workflows/validate.yml

@@ -11,4 +11,5 @@ jobs:
         with:
           python-version: '3.x'
       - run: python scripts/validate.py
+      - run: python -m unittest discover -s tests -v
       - run: python scripts/search.py SQLite --json

scripts/link_check_issue.py

@@ -0,0 +1,25 @@
+"""Prepare a bounded issue preview without modifying the full reports."""
+import os
+from pathlib import Path
+
+
+def issue_body(report, run_url):
+    limit = 60000
+    encoded = report.encode("utf-8")
+    if len(encoded) <= limit:
+        return report
+    notice = (
+        "\n\nReport preview truncated. Full Markdown and JSON reports: "
+        f"{run_url}#artifacts\n"
+    )
+    preview = encoded[:limit - len(notice.encode("utf-8"))].decode("utf-8", errors="ignore")
+    return preview + notice
+
+
+if __name__ == "__main__":
+    run_url = (
+        f"{os.environ['GITHUB_SERVER_URL']}/{os.environ['GITHUB_REPOSITORY']}"
+        f"/actions/runs/{os.environ['GITHUB_RUN_ID']}"
+    )
+    report = Path("link-check-report.md").read_bytes().decode("utf-8")
+    Path("link-check-issue.md").write_bytes(issue_body(report, run_url).encode("utf-8"))

tests/test_link_check_issue.py

@@ -0,0 +1,38 @@
+import os
+import subprocess
+import sys
+import tempfile
+import unittest
+from pathlib import Path
+
+SCRIPTS = Path(__file__).resolve().parents[1] / "scripts"
+sys.path.insert(0, str(SCRIPTS))
+from link_check_issue import issue_body
+
+
+class LinkCheckIssueTests(unittest.TestCase):
+    def test_short_report_unchanged(self):
+        report = "# Link report\nAll findings: café 🌍\n"
+        self.assertEqual(issue_body(report, "https://example.com/run/1"), report)
+
+    def test_large_reports_preserve_sources_and_trailing_notice(self):
+        for content in ("a" * 200000, "é🌍" * 100000):
+            with self.subTest(multibyte=content.startswith("é")), tempfile.TemporaryDirectory() as directory:
+                root = Path(directory)
+                report = root / "link-check-report.md"
+                results = root / "link-check-results.json"
+                report.write_bytes(content.encode("utf-8"))
+                results.write_bytes(b'{"findings": "unchanged"}\n')
+                originals = (report.read_bytes(), results.read_bytes())
+                subprocess.run(
+                    [sys.executable, str(SCRIPTS / "link_check_issue.py")],
+                    cwd=root, check=True,
+                    env={**os.environ, "GITHUB_SERVER_URL": "https://github.com",
+                         "GITHUB_REPOSITORY": "owner/repo", "GITHUB_RUN_ID": "123"},
+                )
+                body = (root / "link-check-issue.md").read_bytes()
+                self.assertLessEqual(len(body), 60000)
+                preview, notice = body.decode("utf-8").split("\n\nReport preview truncated.")
+                self.assertTrue(content.startswith(preview))
+                self.assertEqual(notice, " Full Markdown and JSON reports: https://github.com/owner/repo/actions/runs/123#artifacts\n")
+                self.assertEqual((report.read_bytes(), results.read_bytes()), originals)