Add CI to validate docs links and table structure

.github/workflows/docs-check.yml

@@ -0,0 +1,18 @@
+name: Docs check
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+
+jobs:
+  validate:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.x"
+      - name: Validate markdown docs (links, tables)
+        run: python3 scripts/check_docs.py

scripts/check_docs.py

@@ -0,0 +1,100 @@
+#!/usr/bin/env python3
+"""Validate the project's markdown docs: link integrity and table shape.
+
+Checks every README.md / docs/*.md file for:
+- balanced [] and () (catches broken/truncated markdown links)
+- relative links that resolve to an existing file
+- markdown tables where every row has the same column count as the header
+
+Exits non-zero if any file has issues, printing a report to stdout.
+"""
+import os
+import re
+import sys
+
+ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+
+
+def find_docs():
+    docs = [os.path.join(ROOT, "README.md")]
+    docs_dir = os.path.join(ROOT, "docs")
+    if os.path.isdir(docs_dir):
+        for name in sorted(os.listdir(docs_dir)):
+            if name.endswith(".md"):
+                docs.append(os.path.join(docs_dir, name))
+    return [d for d in docs if os.path.isfile(d)]
+
+
+def check_brackets(text):
+    issues = []
+    if text.count("[") != text.count("]"):
+        issues.append(f"bracket mismatch: {text.count('[')} [ vs {text.count(']')} ]")
+    if text.count("(") != text.count(")"):
+        issues.append(f"paren mismatch: {text.count('(')} ( vs {text.count(')')} )")
+    return issues
+
+
+def check_links(path, text):
+    issues = []
+    base = os.path.dirname(path)
+    for m in re.finditer(r"\[[^\]]*\]\(([^)]+)\)", text):
+        target = m.group(1).strip()
+        if target.startswith("http://") or target.startswith("https://") or target.startswith("#") or target.startswith("mailto:"):
+            continue
+        target_path = target.split("#", 1)[0]
+        resolved = os.path.normpath(os.path.join(base, target_path))
+        if not os.path.exists(resolved):
+            issues.append(f"broken link: {target} -> {os.path.relpath(resolved, ROOT)}")
+    return issues
+
+
+def check_tables(text):
+    issues = []
+    lines = text.split("\n")
+    i = 0
+    while i < len(lines):
+        line = lines[i]
+        if line.strip().startswith("|"):
+            header_cols = line.count("|")
+            block_start = i
+            j = i + 1
+            row_num = 1
+            while j < len(lines) and lines[j].strip().startswith("|"):
+                cols = lines[j].count("|")
+                if cols != header_cols:
+                    issues.append(
+                        f"table column mismatch at line {j + 1} (row {row_num} of block starting line {block_start + 1}): "
+                        f"expected {header_cols} pipes, got {cols}"
+                    )
+                row_num += 1
+                j += 1
+            i = j
+        else:
+            i += 1
+    return issues
+
+
+def main():
+    docs = find_docs()
+    if not docs:
+        print("No markdown docs found.")
+        return 0
+
+    had_issues = False
+    for path in docs:
+        rel = os.path.relpath(path, ROOT)
+        text = open(path, encoding="utf-8").read()
+        issues = check_brackets(text) + check_links(path, text) + check_tables(text)
+        if issues:
+            had_issues = True
+            print(f"FAIL {rel}")
+            for issue in issues:
+                print(f"  - {issue}")
+        else:
+            print(f"OK   {rel}")
+
+    return 1 if had_issues else 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())