From 2db730d837f4cdafc98eac1a3e26b52fdad38026 Mon Sep 17 00:00:00 2001
From: Marlon Saglia <marlon@vespa.ai>
Date: Thu, 8 Aug 2024 15:24:23 +0200
Subject: [PATCH] feat: add workflow to run htmlproofer

---
 .github/workflows/jekyll-link-checker.yml | 78 +++++++++++++++++++++++
 1 file changed, 78 insertions(+)
 create mode 100644 .github/workflows/jekyll-link-checker.yml

diff --git a/.github/workflows/jekyll-link-checker.yml b/.github/workflows/jekyll-link-checker.yml
new file mode 100644
index 0000000..6cc3356
--- /dev/null
+++ b/.github/workflows/jekyll-link-checker.yml
@@ -0,0 +1,78 @@
+---
+name: Link Checker
+
+on:
+  workflow_call:
+    inputs:
+      ignore-urls:
+        description: "Comma separated list of URLs to ignore"
+        required: false
+        default: ""
+        type: string
+      ignore-files:
+        description: "Comma separated list of files to ignore"
+        required: false
+        default: ""
+        type: string
+      swap-urls:
+        description: "Comma separated list of URLs to swap"
+        required: false
+        type: string
+        default: >-
+          (https\://github.com.*/master/.*)#.*:\1,
+          (https\://github.com.*/main/.*)#.*:\1,
+          (https\://github.com.*/blob/.*)#.*:\1
+
+defaults:
+  run:
+    # Specify to ensure "pipefail and errexit" are set.
+    # Ref: https://docs.github.com/en/actions/writing-workflows/workflow-syntax-for-github-actions#defaultsrunshell
+    shell: bash
+
+jobs:
+  htmlproofer:
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Setup Ruby
+        uses: ruby/setup-ruby@v1
+        with:
+          ruby-version: 3.1
+          bundler-cache: true
+
+      - name: Build site
+        run: |
+          bunbdle exec jekyll build
+
+      - name: Clean redirections
+        run: |
+          # Remove the redirect-files before link-check
+          find _site/ -name \*.html | \
+            xargs grep -l "Click here if you are not redirected." | xargs rm
+
+      - name: Clean code elements from html
+        run: |
+          # htmlproofer does not check links inside "<code>" and "<pre>" elements
+          find _site -name \*.html | xargs sed -i.orig 's/<code[^>]*>//g; s/<\/code>//g; s/<pre[^>]*>//g; s/<\/pre>//g;'
+          find _site -name \*.orig | xargs rm
+
+      - name: check links
+        env:
+          LANG: "C.UTF-8"
+          IGNORE_URLS: "${{ inputs.ignore-urls }}"
+          IGNORE_FILES: "${{ inputs.ignore-files }}"
+          SWAP_URLS: "${{ inputs.swap-urls }}"
+        run: |
+          bundle exec htmlproofer \
+            --assume-extension .html \
+            --no-enforce-https \
+            --no-check-external-hash \
+            --allow-missing-href \
+            --typhoeus '{"connecttimeout": 10, "timeout": 30, "accept_encoding": "zstd,br,gzip,deflate"}' \
+            --hydra '{"max_concurrency": 1}' \
+            --ignore-files "${IGNORE_FILES}" \
+            --ignore-urls "${IGNORE_URLS}" \
+            --swap-urls "${SWAP_URLS}" \
+            _site