From 10227b959fea9db95e75169edaced82cc46ec847 Mon Sep 17 00:00:00 2001
From: GabMus <gabmus@disroot.org>
Date: Fri, 22 Dec 2023 09:16:24 +0000
Subject: [PATCH] feat: test if local links exist in ci

---
 .gitlab-ci.yml | 11 ++++++++---
 test_links.py  | 33 +++++++++++++++++++++++++++++++++
 2 files changed, 41 insertions(+), 3 deletions(-)
 create mode 100644 test_links.py

diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml
index 04772ba..fe610f5 100644
--- a/.gitlab-ci.yml
+++ b/.gitlab-ci.yml
@@ -1,15 +1,20 @@
-image: registry.gitlab.com/pages/hugo/hugo_extended
-
 variables:
   GIT_SUBMODULE_STRATEGY: recursive
 
-test:
+test-links:
+  image: python:latest
+  script:
+    - python3 ./test_links.py
+
+test-build:
+  image: registry.gitlab.com/pages/hugo/hugo_extended
   script:
   - hugo
   except:
   - main
 
 pages:
+  image: registry.gitlab.com/pages/hugo/hugo_extended
   script:
     - hugo
   artifacts:
diff --git a/test_links.py b/test_links.py
new file mode 100644
index 0000000..07e0a9a
--- /dev/null
+++ b/test_links.py
@@ -0,0 +1,33 @@
+#!/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)