Check links in CI with htmlcheck

Remove test_links.py

.gitignore

@@ -7,3 +7,6 @@ node_modules/
 public/
 resources/
 .deploy*/
+
+# htmltest
+tmp/

.gitlab-ci.yml

@@ -1,28 +1,43 @@
 variables:
   GIT_SUBMODULE_STRATEGY: recursive
 
-test-links:
-  image: python:latest
-  script:
-    - python3 ./test_links.py
+## GitLab defaults to:
+# stages:
+# - build
+# - test
+# - deploy
 
-test-trailing-spaces:
-  script:
-    - ./test_trailing_spaces.sh
-
-test-build:
-  image: registry.gitlab.com/pages/hugo/hugo_extended
-  script:
-  - hugo
-  except:
-  - main
-
-pages:
+build-site:
+  stage: build
   image: registry.gitlab.com/pages/hugo/hugo_extended
   script:
     - hugo
   artifacts:
     paths:
       - public
+
+test-html:
+  stage: test
+  image:
+    name: wjdp/htmltest:latest
+    entrypoint: [""]
+  cache:
+    key: htmltest-cache
+    paths:
+      - tmp/
+  script:
+    - htmltest
+
+test-trailing-spaces:
+  stage: test
+  needs: [] # ok to run before build-site
+  script:
+    - ./test_trailing_spaces.sh
+
+# The GitLab Pages job must always be named "pages", and result in a "public" artifact.
+# Our "public" artifact is created above, but GitLab jobs will fetch all artifacts from previous stages.
+pages:
+  stage: deploy
+  script: ls -ld public
   only:
     - main

.htmltest.yml

@@ -0,0 +1,11 @@
+# https://github.com/wjdp/htmltest?tab=readme-ov-file#wrench-configuration
+
+DirectoryPath: "public"
+EnforceHTTPS: true
+IgnoreURLs:
+# - "example.com"
+# - "^/misc/js/script.js$"
+- "https://gitlab.com/lvra/lvra.gitlab.io/-/edit"
+# IgnoreDirs:
+# - "lib"
+CacheExpires: "6h"

test_links.py

@@ -1,33 +0,0 @@
-#!/usr/bin/env python3
-from pathlib import Path
-import re
-from typing import List
-from os.path import exists
-
-
-def get_all_markdown_files():
-    return Path("content").rglob("*.md")
-
-
-def find_links(f: str) -> List[str]:
-    with open(f, "r") as fd:
-        content = fd.read()
-    return [
-        link.lstrip("]").strip("()") for link in re.findall(r"\]\(\/[\w\/]+\)", content)
-    ]
-
-
-def verify_link_exists(link: str) -> bool:
-    return exists(f"content{link}/_index.md") or exists(f'content{link.rstrip("/")}.md')
-
-
-md_files = get_all_markdown_files()
-
-res = 0
-for f in md_files:
-    for link in find_links(f):
-        if not verify_link_exists(link):
-            print(f"E: {f}: {link} does not exist")
-            res = 1
-
-exit(res)