From 1c1d0e7692f7e2e8ceda9dd9524803dd7fad2762 Mon Sep 17 00:00:00 2001
From: Pete Lumbis <pete@upbound.io>
Date: Tue, 18 Jul 2023 12:47:17 -0400
Subject: [PATCH] Weekly scheduled external link checking (#488)

---
 .github/workflows/weekly-link-checker.yml | 34 +++++++++++++++++++++++
 utils/htmltest/.htmltest.yml              |  5 +++-
 2 files changed, 38 insertions(+), 1 deletion(-)
 create mode 100644 .github/workflows/weekly-link-checker.yml

diff --git a/.github/workflows/weekly-link-checker.yml b/.github/workflows/weekly-link-checker.yml
new file mode 100644
index 00000000..75baff01
--- /dev/null
+++ b/.github/workflows/weekly-link-checker.yml
@@ -0,0 +1,34 @@
+name: Validate external links
+
+on:
+  schedule:
+    # “At 04:05 on Monday”
+    - cron:  '5 4 * * 1'
+
+  # Allows manual workflow run (must in default branch to work)
+  workflow_dispatch:
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout the docs repo
+        uses: actions/checkout@v3
+
+      - name: Setup Hugo
+        uses: peaceiris/actions-hugo@v2
+        with:
+          hugo-version: '0.109.0'
+          extended: true
+
+      - name: Build
+        # Set to env=dev so SCSS isn't required.
+        run: hugo --environment development
+
+      - name: Enable external link checking
+        run: "sed -i 's/CheckExternal: false/CheckExternal: true/' utils/htmltest/.htmltest.yml"
+
+      - name: Run htmltest
+        uses: wjdp/htmltest-action@master
+        with:
+          config: ./utils/htmltest/.htmltest.yml
diff --git a/utils/htmltest/.htmltest.yml b/utils/htmltest/.htmltest.yml
index 0732e868..99997423 100644
--- a/utils/htmltest/.htmltest.yml
+++ b/utils/htmltest/.htmltest.yml
@@ -4,4 +4,7 @@ DirectoryPath: "public"
 IgnoreInternalEmptyHash: true
 CheckExternal: false
 IgnoreURLs:
-  - "github.com/crossplane/docs/tree/master/content/(.*).md" # Ignore the links to "view this source"
\ No newline at end of file
+  - "github.com/crossplane/docs/tree/master/content/(.*).md" # Ignore the links to "view this source"
+  - "www.googletagmanager.com/*" # Ignore google tag manager
+  - "twitter.com/*" # Ignore twitter links since they send to login page
+