Refuse image fetches that are not http or https

src/docgen/ai_client.py

@@ -35,6 +35,7 @@
 import random
 import time
 import urllib.error
+import urllib.parse
 import urllib.request
 import uuid
 from dataclasses import dataclass
@@ -609,8 +610,17 @@ def _call() -> Any:
     return _openai_stt_payload(result)
 
 
+def _require_http_url(url: str) -> None:
+    """Image fetches are http(s) only. ``file:`` and other schemes are refused."""
+    scheme = urllib.parse.urlparse(url).scheme.lower()
+    if scheme not in ("http", "https"):
+        shown = scheme or "an empty scheme"
+        raise AIError(f"image URL must use http or https, not {shown!r}")
+
+
 def fetch_url_bytes(url: str, *, timeout: int = 120) -> bytes:
     """GET ``url`` and return the response body (image CDN / signed URLs)."""
+    _require_http_url(url)
     last_exc: BaseException | None = None
     for attempt in range(_MAX_HTTP_ATTEMPTS):
         req = urllib.request.Request(url, method="GET")

tests/test_ai_client.py

@@ -931,6 +931,15 @@ def _urlopen(req: object, timeout: int = 120) -> _Resp:
     assert calls["n"] == 2
 
 
+def test_fetch_url_bytes_refuses_file_scheme() -> None:
+    from docgen.ai_client import fetch_url_bytes
+
+    with patch("docgen.ai_client.urllib.request.urlopen") as urlopen:
+        with pytest.raises(AIError, match="http or https"):
+            fetch_url_bytes("file:///etc/passwd")
+    urlopen.assert_not_called()
+
+
 def test_chat_retries_rate_limit(
     tmp_path: Path, monkeypatch: pytest.MonkeyPatch, clear_ai_env: None
 ) -> None: