#!/usr/bin/env python3 """Check every relative link and external URL in the gallery, pack, PDF, and deck.""" from __future__ import annotations import re import sys import tempfile import zipfile from pathlib import Path from urllib.error import HTTPError, URLError from urllib.request import Request, urlopen from xml.etree import ElementTree ROOT = Path(__file__).resolve().parents[1] REPORT = ROOT / "downloads" / "link-check.txt" NS = {"pr": "http://schemas.openxmlformats.org/package/2006/relationships"} BANNED = [ (r"linktree", "linktree"), (r"linktr\.ee", "linktr.ee"), (r"laxman", "laxman"), (r"kadiyala", "kadiyala"), (r"\blucky\b", "lucky"), (r"cypher", "cypher"), (r"cipher0x9", "cipher0x9"), (r"\bhermes\b", "hermes"), (r"h-1b", "h-1b"), (r"\bh1b\b", "h1b"), (r"grokbot", "grokbot"), (r"[\w.+\-]+@[\w\-]+\.[A-Za-z]{2,}", "email address"), ] SKIP_DIRS = {".git", "__pycache__", "fonts"} def fetch(url: str, cache: dict[str, str]) -> str: if url in cache: return cache[url] last = "no response" for agent in ("GemsLinkCheck/1.0", "Mozilla/5.0"): req = Request(url, headers={"User-Agent": agent}, method="GET") try: with urlopen(req, timeout=25) as res: code = getattr(res, "status", 200) cache[url] = "ok" if 200 <= code < 400 else f"HTTP {code}" return cache[url] except HTTPError as exc: last = f"HTTP {exc.code}" if exc.code in {301, 302, 303, 307, 308}: cache[url] = "ok" return "ok" if exc.code == 403: continue except URLError as exc: last = f"URL error {exc.reason}" except Exception as exc: # noqa: BLE001 last = str(exc) cache[url] = last return last def md_html_targets(text: str) -> list[str]: found = re.findall(r"\]\(([^)\s]+)\)", text) found += re.findall(r"""(?:href|src)\s*=\s*["']([^"']+)["']""", text, flags=re.I) found += [item.strip("\"'") for item in re.findall(r"""url\(([^)]+)\)""", text)] return found def has_anchor(text: str, anchor: str) -> bool: if not anchor: return True return f'id="{anchor}"' in text or f"id='{anchor}'" in text def check_tree(root: Path, label: str, failures: list[str], cache: dict[str, str]) -> int: checked = 0 files = [p for p in root.rglob("*") if p.is_file() and p.suffix.lower() in {".md", ".html", ".css"}] for path in files: if any(part in SKIP_DIRS for part in path.parts): continue text = path.read_text(encoding="utf-8", errors="replace") for raw in md_html_targets(text): if raw.startswith(("data:", "mailto:", "javascript:")): failures.append(f"{label} {path.relative_to(root)}: disallowed target {raw}") continue checked += 1 target, _, anchor = raw.partition("#") if target.startswith(("http://", "https://")): result = fetch(target, cache) if result != "ok": failures.append(f"{label} {path.relative_to(root)}: {raw} -> {result}") continue if not target: if not has_anchor(text, anchor): failures.append(f"{label} {path.relative_to(root)}: missing anchor #{anchor}") continue dest = (path.parent / target).resolve() try: dest.relative_to(root.resolve()) except ValueError: failures.append(f"{label} {path.relative_to(root)}: link escapes the tree {raw}") continue if not dest.is_file(): failures.append(f"{label} {path.relative_to(root)}: missing file {raw}") continue if anchor and not has_anchor(dest.read_text(encoding="utf-8", errors="replace"), anchor): failures.append(f"{label} {path.relative_to(root)}: missing anchor {raw}") return checked def check_pdf(path: Path, label: str, failures: list[str], cache: dict[str, str]) -> int: from pypdf import PdfReader reader = PdfReader(str(path)) page_refs = set() for page in reader.pages: ref = page.indirect_reference if ref is not None: page_refs.add((ref.idnum, ref.generation)) checked = 0 for index, page in enumerate(reader.pages, 1): annots = page.get("/Annots") or [] for annot in annots: obj = annot.get_object() if obj.get("/Subtype") != "/Link": continue checked += 1 action = obj.get("/A") dest = obj.get("/Dest") if action is not None and action.get("/S") == "/URI": url = str(action.get("/URI")) result = fetch(url, cache) if result != "ok": failures.append(f"{label} PDF page {index}: {url} -> {result}") continue if dest is not None: if isinstance(dest, list) and dest: ref = dest[0] key = (getattr(ref, "idnum", None), getattr(ref, "generation", None)) if key not in page_refs: failures.append(f"{label} PDF page {index}: destination is not a page") else: failures.append(f"{label} PDF page {index}: unsupported destination {dest!r}") continue failures.append(f"{label} PDF page {index}: link has no URI or destination") if checked == 0: failures.append(f"{label} PDF has no links") return checked def check_pptx(path: Path, base: Path, label: str, failures: list[str], cache: dict[str, str]) -> int: checked = 0 with zipfile.ZipFile(path) as archive: for name in archive.namelist(): if not name.startswith("ppt/slides/_rels/") or not name.endswith(".rels"): continue root = ElementTree.fromstring(archive.read(name)) for rel in root.findall("pr:Relationship", NS): if not str(rel.get("Type", "")).endswith("/hyperlink"): continue checked += 1 target = rel.get("Target") or "" mode = rel.get("TargetMode") if mode != "External": failures.append(f"{label} PPTX {name}: hyperlink is not external ({target})") continue if target.startswith(("http://", "https://")): result = fetch(target, cache) if result != "ok": failures.append(f"{label} PPTX: {target} -> {result}") continue dest = (base / target).resolve() tree = base.parent.resolve() try: dest.relative_to(tree) except ValueError: failures.append(f"{label} PPTX: link escapes the pack {target}") continue if not dest.is_file(): failures.append(f"{label} PPTX: missing {target} from {base}") if checked == 0: failures.append(f"{label} PPTX has no hyperlinks") return checked def check_privacy(root: Path, failures: list[str]) -> int: scanned = 0 files = [] for path in root.rglob("*"): if not path.is_file() or any(part in {".git", "__pycache__"} for part in path.parts): continue if path.name in {"check_links.py", "link-check.txt"}: continue if path.suffix.lower() in {".md", ".html", ".css", ".js", ".py", ".txt", ".svg"}: files.append(path) blob = [] for path in files: scanned += 1 blob.append(path.read_text(encoding="utf-8", errors="replace")) pdf = root / "downloads" / "Gemini-Gems-Guide.pdf" pptx = root / "downloads" / "Gemini-Gems-Overview.pptx" if pdf.is_file(): from pypdf import PdfReader blob.append("\n".join((page.extract_text() or "") for page in PdfReader(str(pdf)).pages)) scanned += 1 if pptx.is_file(): with zipfile.ZipFile(pptx) as archive: for name in archive.namelist(): if name.startswith("ppt/slides/slide") and name.endswith(".xml"): blob.append(archive.read(name).decode("utf-8", "replace")) scanned += 1 haystack = "\n".join(blob) for pattern, name in BANNED: match = re.search(pattern, haystack, flags=re.I) if match: failures.append(f"privacy: found {name} near '{match.group(0)[:80]}'") return scanned def main() -> int: failures: list[str] = [] cache: dict[str, str] = {} repo_links = check_tree(ROOT, "repo", failures, cache) pdf_links = check_pdf(ROOT / "downloads" / "Gemini-Gems-Guide.pdf", "repo", failures, cache) pptx_links = check_pptx( ROOT / "downloads" / "Gemini-Gems-Overview.pptx", ROOT / "downloads", "repo", failures, cache, ) pack = ROOT / "downloads" / "gemini-gems-pack.zip" with tempfile.TemporaryDirectory() as tmp: dest = Path(tmp) with zipfile.ZipFile(pack) as archive: names = set(archive.namelist()) archive.extractall(dest) required = [ "README.md", "LICENSE", "site/index.html", "site/print.html", "site/assets/gallery.css", "site/assets/gallery.js", "examples/README.md", "examples/gem_app.py", "examples/gem_app.mjs", "downloads/Gemini-Gems-Guide.pdf", "downloads/Gemini-Gems-Overview.pptx", "scripts/build_gallery.py", "scripts/check_links.py", ] required += [f"gems/{path.name}" for path in sorted((ROOT / "gems").glob("*.md"))] required += [f"site/gems/{path.name}" for path in sorted((ROOT / "site" / "gems").glob("*.html"))] for rel in required: if rel not in names: failures.append(f"zip missing {rel}") if "downloads/gemini-gems-pack.zip" not in names: failures.append("zip is missing its unpacked download link") zip_links = check_tree(dest, "zip", failures, cache) zip_pdf = check_pdf(dest / "downloads" / "Gemini-Gems-Guide.pdf", "zip", failures, cache) zip_pptx = check_pptx(dest / "downloads" / "Gemini-Gems-Overview.pptx", dest / "downloads", "zip", failures, cache) privacy_files = check_privacy(ROOT, failures) lines = [ "Gemini Gems Gallery link and privacy check", f"repo document links checked: {repo_links}", f"pdf links checked: {pdf_links}", f"pptx links checked: {pptx_links}", f"zip document links checked: {zip_links}", f"zip pdf links checked: {zip_pdf}", f"zip pptx links checked: {zip_pptx}", f"privacy files scanned: {privacy_files}", f"external urls checked: {len(cache)}", ] for url, result in sorted(cache.items()): lines.append(f" {result}: {url}") if failures: lines.append(f"FAILURES: {len(failures)}") lines.extend(f"- {item}" for item in failures) else: lines.append("FAILURES: 0") lines.append("All checked links resolved. No personal-detail patterns matched.") REPORT.write_text("\n".join(lines) + "\n", encoding="utf-8") print("\n".join(lines)) return 1 if failures else 0 if __name__ == "__main__": sys.exit(main())