From 71912b8358da4647ca15e306843e2eb9eee3d1cc Mon Sep 17 00:00:00 2001
From: Gahow Wang <gahow.wang@gmail.com>
Date: Tue, 7 Apr 2026 16:06:54 +0800
Subject: [PATCH] Wrap additional network errors in factor download

---
 factor_attribution.py            | 12 ++++++++-
 tests/test_factor_attribution.py | 44 +++++++++++++++++++++++++++++---
 2 files changed, 51 insertions(+), 5 deletions(-)

diff --git a/factor_attribution.py b/factor_attribution.py
index 7aa6c92..a5439a4 100644
--- a/factor_attribution.py
+++ b/factor_attribution.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import http.client
 import io
 import socket
 import ssl
@@ -35,7 +36,16 @@ def _download_kf_zip_bytes() -> bytes:
     try:
         with urlopen(request, timeout=30) as response:
             return response.read()
-    except (URLError, TimeoutError, ConnectionError, socket.timeout, ssl.SSLError) as exc:
+    except (
+        URLError,
+        TimeoutError,
+        ConnectionError,
+        socket.timeout,
+        socket.gaierror,
+        ssl.SSLError,
+        http.client.IncompleteRead,
+        http.client.RemoteDisconnected,
+    ) as exc:
         raise ExternalFactorDownloadError(f"Failed to download external factor data: {exc}") from exc
 
 
diff --git a/tests/test_factor_attribution.py b/tests/test_factor_attribution.py
index 4c3a80b..dc0fc07 100644
--- a/tests/test_factor_attribution.py
+++ b/tests/test_factor_attribution.py
@@ -1,3 +1,4 @@
+import http.client
 import io
 import socket
 import ssl
@@ -41,6 +42,7 @@ class ExternalFactorLoaderTests(unittest.TestCase):
             TimeoutError("timed out"),
             ConnectionError("conn reset"),
             socket.timeout("socket timed out"),
+            socket.gaierror("dns failed"),
             ssl.SSLError("tls failed"),
         ):
             with self.subTest(error_type=type(error).__name__):
@@ -48,6 +50,16 @@ class ExternalFactorLoaderTests(unittest.TestCase):
                     with self.assertRaises(ExternalFactorDownloadError):
                         _download_kf_zip_bytes()
 
+    def test_download_kf_zip_bytes_wraps_incomplete_read_errors(self):
+        response = mock.MagicMock()
+        response.read.side_effect = http.client.IncompleteRead(b"partial", 10)
+        response.__enter__.return_value = response
+        response.__exit__.return_value = False
+
+        with mock.patch("factor_attribution.urlopen", return_value=response):
+            with self.assertRaises(ExternalFactorDownloadError):
+                _download_kf_zip_bytes()
+
     def test_load_external_us_factors_parses_percent_values_and_dates_from_zip_payload(self):
         csv_text = (
             "This line is ignored\n"
@@ -92,10 +104,34 @@ class ExternalFactorLoaderTests(unittest.TestCase):
         with tempfile.TemporaryDirectory() as tmpdir:
             cache_dir = Path(tmpdir)
             cached.to_csv(cache_dir / "ff5_us_daily.csv")
-            with mock.patch(
-                "factor_attribution._download_kf_zip_bytes",
-                side_effect=ExternalFactorDownloadError("boom"),
-            ):
+            with mock.patch("factor_attribution.urlopen", side_effect=socket.gaierror("dns failed")):
+                with self.assertWarnsRegex(UserWarning, "cached data"):
+                    factors = load_external_us_factors(cache_dir=cache_dir)
+
+        self.assertEqual(len(factors), 1)
+        self.assertAlmostEqual(factors.iloc[0]["MKT_RF"], 0.01)
+
+    def test_load_external_us_factors_falls_back_to_cache_when_download_read_is_incomplete(self):
+        cached = pd.DataFrame(
+            {
+                "MKT_RF": [0.01],
+                "SMB": [0.0],
+                "HML": [0.0],
+                "RMW": [0.0],
+                "CMA": [0.0],
+                "RF": [0.0001],
+            },
+            index=pd.to_datetime(["2026-01-02"]),
+        )
+        response = mock.MagicMock()
+        response.read.side_effect = http.client.IncompleteRead(b"partial", 10)
+        response.__enter__.return_value = response
+        response.__exit__.return_value = False
+
+        with tempfile.TemporaryDirectory() as tmpdir:
+            cache_dir = Path(tmpdir)
+            cached.to_csv(cache_dir / "ff5_us_daily.csv")
+            with mock.patch("factor_attribution.urlopen", return_value=response):
                 with self.assertWarnsRegex(UserWarning, "cached data"):
                     factors = load_external_us_factors(cache_dir=cache_dir)