ci: replace htmlproofer with wjdp/htmltest

Signed-off-by: David Karlsson <david.karlsson@docker.com>

.github/workflows/build.yml

@@ -35,7 +35,7 @@ jobs:
       fail-fast: false
       matrix:
         target:
-          - htmlproofer
+          - htmltest
           - mdl
     steps:
       -
@@ -70,7 +70,7 @@ jobs:
 
             // construct annotations by parsing output
             switch ("${{ matrix.target }}") {
-            case "htmlproofer":
+            case "htmltest":
               const re = /^- (.+)\n  \*  (.+) \(line (\d+)\)\n(.+)$/gm;
               while (true) {
                 const result = re.exec(results);
@@ -79,7 +79,7 @@ jobs:
                 }
 
                 core.error(`${result[2]}\n${result[4]}`, {
-                  title: 'Link check failed',
+                  title: 'HTML test failed',
                   // file: result[1],
                   // startLine: result[3],
                 });

.gitignore

@@ -10,3 +10,4 @@ CNAME
 _kbase/**
 /vendor
 /lint
+tmp/.htmltest/**

.htmltest.yml

@@ -0,0 +1,16 @@
+DirectoryPath: "_site"
+EnforceHTTPS: false
+CheckDoctype: false
+CheckExternal: false
+IgnoreAltMissing: true
+IgnoreAltEmpty: true
+IgnoreEmptyHref: true
+IgnoreDirectoryMissingTrailingSlash: true
+IgnoreURLs:
+- "^/docker-hub/api/latest/.*$"
+- "^/engine/api/v.+/#.*$"
+- "^/glossary/.*$"
+IgnoreDirs:
+- "engine/api"
+- "registry/configuration"
+CacheExpires: "6h"

Dockerfile

@@ -63,28 +63,27 @@ RUN --mount=type=bind,target=.,rw \
   bundle exec jekyll build --profile -d ${TARGET} --config ${CONFIG_FILES}
 EOT
 
-# htmlproofer checks for broken links
-FROM gem AS htmlproofer-base
-RUN --mount=type=bind,from=generate,source=/out,target=_site <<EOF
-  htmlproofer ./_site \
-    --disable-external \
-    --internal-domains="docs.docker.com,docs-stage.docker.com,localhost:4000" \
-    --file-ignore="/^./_site/engine/api/.*$/,./_site/registry/configuration/index.html" \
-    --url-ignore="/^/docker-hub/api/latest/.*$/,/^/engine/api/v.+/#.*$/,/^/glossary/.*$/" > /results 2>&1
-  rc=$?
-  if [[ $rc -eq 0 ]]; then
-    echo -n > /results
-  fi
+# htmltest checks for broken links
+FROM wjdp/htmltest:v0.17.0 as htmltest-base
+RUN --mount=type=bind,from=generate,source=/out,target=_site \
+    --mount=type=bind,source=.htmltest.yml,target=.htmltest.yml \
+    <<EOF
+    htmltest > /results 2>&1
+    rc=$?
+    if [[ $rc -eq 0 ]]; then
+      echo -n > /results
+    fi
 EOF
 
-FROM htmlproofer-base as htmlproofer
+FROM base as htmltest
+COPY --from=htmltest-base /results /results
 RUN <<EOF
   cat /results
   [ ! -s /results ] || exit 1
 EOF
 
-FROM scratch as htmlproofer-output
-COPY --from=htmlproofer-base /results /results
+FROM scratch as htmltest-output
+COPY --from=htmltest-base /results /results
 
 # mdl is a lint tool for markdown files
 FROM gem AS mdl-base

Gemfile

@@ -17,7 +17,6 @@ gem 'rouge', '3.27.0'
 
 gem 'front_matter_parser', '1.0.1'
 gem 'git', '1.13.0'
-gem 'html-proofer', '3.19.4'
 gem 'mdl', '0.11.0'
 gem 'octopress-hooks', '2.6.2'
 gem 'rake', '13.0.6'

Gemfile.lock

@@ -10,8 +10,6 @@ GEM
     em-websocket (0.5.3)
       eventmachine (>= 0.12.9)
       http_parser.rb (~> 0)
-    ethon (0.15.0)
-      ffi (>= 1.15.0)
     eventmachine (1.2.7)
     ffi (1.15.5)
     forwardable-extended (2.6.0)
@@ -19,14 +17,6 @@ GEM
     git (1.13.0)
       addressable (~> 2.8)
       rchardet (~> 1.8)
-    html-proofer (3.19.4)
-      addressable (~> 2.3)
-      mercenary (~> 0.3)
-      nokogiri (~> 1.13)
-      parallel (~> 1.10)
-      rainbow (~> 3.0)
-      typhoeus (~> 1.3)
-      yell (~> 2.0)
     http_parser.rb (0.8.0)
     i18n (1.12.0)
       concurrent-ruby (~> 1.0)
@@ -75,20 +65,11 @@ GEM
       tomlrb
     mixlib-shellout (3.2.7)
       chef-utils
-    nokogiri (1.14.3-aarch64-linux)
-      racc (~> 1.4)
-    nokogiri (1.14.3-arm-linux)
-      racc (~> 1.4)
-    nokogiri (1.14.3-x86_64-linux)
-      racc (~> 1.4)
     octopress-hooks (2.6.2)
       jekyll (>= 2.0)
-    parallel (1.22.1)
     pathutil (0.16.2)
       forwardable-extended (~> 2.6)
     public_suffix (5.0.1)
-    racc (1.6.2)
-    rainbow (3.1.1)
     rake (13.0.6)
     rb-fsevent (0.11.2)
     rb-inotify (0.10.1)
@@ -102,10 +83,7 @@ GEM
     terminal-table (2.0.0)
       unicode-display_width (~> 1.1, >= 1.1.1)
     tomlrb (2.0.3)
-    typhoeus (1.4.0)
-      ethon (>= 0.9.0)
     unicode-display_width (1.8.0)
-    yell (2.2.2)
 
 PLATFORMS
   aarch64-linux
@@ -115,7 +93,6 @@ PLATFORMS
 DEPENDENCIES
   front_matter_parser (= 1.0.1)
   git (= 1.13.0)
-  html-proofer (= 3.19.4)
   jekyll (= 4.2.2)
   jekyll-redirect-from
   jekyll-relative-links

docker-bake.hcl

@@ -36,18 +36,18 @@ target "vendor" {
 }
 
 group "validate" {
-  targets = ["htmlproofer", "mdl"]
+  targets = ["htmltest", "mdl"]
 }
 
-target "htmlproofer" {
+target "htmltest" {
   inherits = ["_common"]
-  target = "htmlproofer"
+  target = "htmltest"
   output = ["type=cacheonly"]
 }
 
-target "htmlproofer-output" {
+target "htmltest-output" {
   inherits = ["_common"]
-  target = "htmlproofer-output"
+  target = "htmltest-output"
   output = ["./lint"]
 }