vault-cli: extract to standalone repo, add flake.nix, move one-off scripts to scripts/one-off/

2026-09-16 13:12:01 -04:00
parent 8e8ad466a8
commit 9f3a99a42e
6 changed files with 33 additions and 0 deletions
@@ -0,0 +1,80 @@
+#!/usr/bin/env python3
+"""Scan /vault for [[wikilinks]] and report which targets don't resolve
+to an actual note. Obsidian resolves links by basename match anywhere
+in the vault (unless the link includes a path), so we build an index
+of all note basenames (and their full relative paths) first.
+"""
+import os, re, json
+from collections import defaultdict
+
+VAULT = "/vault"
+LINK_RE = re.compile(r'\[\[([^\]|#]+)(?:#[^\]|]*)?(?:\|[^\]]*)?\]\]')
+
+def all_notes():
+    idx = defaultdict(list)  # basename (no ext, lowercase) -> [full paths]
+    all_paths = []
+    for dp, dirnames, fns in os.walk(VAULT):
+        if ".git" in dp.split(os.sep):
+            continue
+        for fn in fns:
+            if fn.lower().endswith(".md"):
+                p = os.path.join(dp, fn)
+                base = fn[:-3]
+                idx[base.lower()].append(p)
+                all_paths.append(p)
+    return idx, all_paths
+
+def resolve(link_target, index, current_dir):
+    # Obsidian: link may be a bare name, or a relative/vault path (with /)
+    target = link_target.strip()
+    if "/" in target:
+        # path-style link; try resolving relative to vault root or current dir
+        candidates = [
+            os.path.join(VAULT, target + ".md"),
+            os.path.join(VAULT, target),
+            os.path.join(current_dir, target + ".md"),
+            os.path.join(current_dir, target),
+        ]
+        for c in candidates:
+            if os.path.isfile(c):
+                return True
+        # also try basename match as fallback
+        base = os.path.basename(target)
+        return base.lower() in index
+    else:
+        return target.lower() in index
+
+def main():
+    index, all_paths = all_notes()
+    unresolved = defaultdict(list)  # target -> [source files]
+    total_links = 0
+    for p in all_paths:
+        try:
+            with open(p, "r", encoding="utf-8", errors="ignore") as f:
+                text = f.read()
+        except Exception:
+            continue
+        for m in LINK_RE.finditer(text):
+            target = m.group(1)
+            total_links += 1
+            if not resolve(target, index, os.path.dirname(p)):
+                unresolved[target].append(os.path.relpath(p, VAULT))
+
+    print(f"Total wikilinks scanned: {total_links}")
+    print(f"Unique unresolved targets: {len(unresolved)}")
+    print(f"Total unresolved link occurrences: {sum(len(v) for v in unresolved.values())}")
+
+    # dump full detail to json for further processing
+    out = {t: srcs for t, srcs in unresolved.items()}
+    with open("/tmp/unresolved_links.json", "w") as f:
+        json.dump(out, f, indent=2)
+    print("Full detail written to /tmp/unresolved_links.json")
+
+    # sorted by occurrence count desc
+    by_count = sorted(unresolved.items(), key=lambda kv: -len(kv[1]))
+    print("\n=== TOP 40 unresolved targets by occurrence count ===")
+    for target, srcs in by_count[:40]:
+        print(f"  {len(srcs):4d}x  [[{target}]]")
+
+if __name__ == "__main__":
+    main()