feat: timezone="auto" resolves from any egress + weekly geoip auto-update

Refine timezone="auto" so it ALWAYS resolves (drop the "host" sentinel): - ""/"auto" resolve from the proxy egress when a proxy is set, else from the host own public IP (direct lookup); an explicit zone is the only opt-out. - on failure: with a proxy raise; without a proxy fall back to the host TZ. GeoIP DB now auto-updates against daijro/geoip-all-in-one weekly rebuild: cache the latest, re-check after GEOIP_REFRESH_DAYS (7), prune old tags, reuse a stale cache offline; GEOIP_MMDB_VERSION is only the cold fallback. tests: test_geo.py (37) + test_geoip_update.py; full unit suite 429 green plus 8 live combinations (proxy / no-proxy / explicit / failing / freshness).
2026-06-06 05:16:20 +02:00
parent 369f3f7fdb
commit f2664f96e1
8 changed files with 361 additions and 113 deletions
@@ -136,6 +136,20 @@ def test_discover_egress_ip_all_fail_raises(monkeypatch):
        discover_egress_ip(SOCKS)


+@pytest.mark.unit
+def test_discover_egress_ip_no_proxy_is_direct(monkeypatch):
+    # proxy=None → direct request, requests.get must get proxies=None.
+    seen = {}
+
+    def fake_get(url, **kw):
+        seen["proxies"] = kw.get("proxies", "MISSING")
+        return _FakeResp("192.0.2.55")
+
+    monkeypatch.setattr(_geo.requests, "get", fake_get)
+    assert discover_egress_ip(None) == "192.0.2.55"
+    assert seen["proxies"] is None
+
+
 # ──────────────────────────────────────────────────────────────────────
 #  ip_to_timezone — mocked mmdb reader
 # ──────────────────────────────────────────────────────────────────────
@@ -194,8 +208,9 @@ def stub_egress(monkeypatch):
    """Make egress resolution deterministic + offline; record if it ran."""
    state = {"called": False}

-    def fake_discover(proxy, **kw):
+    def fake_discover(proxy=None, **kw):
        state["called"] = True
+        state["proxy_arg"] = proxy
        return "203.0.113.7"

    monkeypatch.setattr(_geo, "discover_egress_ip", fake_discover)
@@ -208,56 +223,66 @@ def stub_egress(monkeypatch):


@pytest.mark.unit
-@pytest.mark.parametrize("sentinel", ["host", "local", "HOST", "Local"])
-def test_resolve_host_sentinel_forces_host_tz(sentinel, stub_egress):
-    # Even with a proxy set, "host"/"local" force the host TZ and never resolve.
-    assert resolve_session_timezone(sentinel, SOCKS) == ""
-    assert stub_egress["called"] is False
-
-
-@pytest.mark.unit
-def test_resolve_explicit_iana_wins_over_proxy(stub_egress):
+def test_resolve_explicit_iana_wins(stub_egress):
+    # An explicit zone wins and never triggers resolution (proxy or not).
    assert resolve_session_timezone("Asia/Tokyo", SOCKS) == "Asia/Tokyo"
-    assert stub_egress["called"] is False  # no resolution when explicit
-
-
-@pytest.mark.unit
-def test_resolve_empty_no_proxy_is_host(stub_egress):
-    assert resolve_session_timezone("", None) == ""
+    assert resolve_session_timezone("Asia/Tokyo", None) == "Asia/Tokyo"
    assert stub_egress["called"] is False


@pytest.mark.unit
-def test_resolve_auto_no_proxy_is_host(stub_egress):
-    assert resolve_session_timezone("auto", None) == ""
-    assert stub_egress["called"] is False
-
-
-@pytest.mark.unit
-def test_resolve_empty_with_proxy_defaults_to_auto(stub_egress):
-    # NEW default: a proxy with no timezone auto-resolves from the egress.
+def test_resolve_empty_with_proxy_resolves_from_proxy(stub_egress):
    assert resolve_session_timezone("", SOCKS) == "America/New_York"
    assert stub_egress["called"] is True
+    assert stub_egress["proxy_arg"] == SOCKS  # routed through the proxy


@pytest.mark.unit
-def test_resolve_auto_with_proxy_resolves(stub_egress):
+def test_resolve_auto_with_proxy_resolves_from_proxy(stub_egress):
    assert resolve_session_timezone("auto", HTTP) == "America/New_York"
+    assert stub_egress["proxy_arg"] == HTTP
+
+
+@pytest.mark.unit
+def test_resolve_empty_no_proxy_resolves_from_host(stub_egress):
+    # auto ALWAYS resolves — without a proxy, from the host's own public IP.
+    assert resolve_session_timezone("", None) == "America/New_York"
    assert stub_egress["called"] is True
+    assert stub_egress["proxy_arg"] is None  # direct request, no proxy


@pytest.mark.unit
-def test_resolve_direct_proxy_treated_as_no_proxy(stub_egress):
-    assert resolve_session_timezone("auto", {"server": "direct://"}) == ""
-    assert stub_egress["called"] is False
+def test_resolve_auto_no_proxy_resolves_from_host(stub_egress):
+    assert resolve_session_timezone("auto", None) == "America/New_York"
+    assert stub_egress["proxy_arg"] is None


@pytest.mark.unit
-def test_resolve_fail_early_propagates(monkeypatch):
-    # With a proxy set, a discovery failure must raise — never silent host TZ.
-    def boom(proxy, **kw):
+def test_resolve_direct_proxy_resolves_via_host(stub_egress):
+    # direct:// counts as "no proxy" → resolve from the host IP, don't skip.
+    assert resolve_session_timezone("auto", {"server": "direct://"}) == "America/New_York"
+    assert stub_egress["proxy_arg"] is None
+
+
+@pytest.mark.unit
+def test_resolve_no_proxy_failure_falls_back_to_host(monkeypatch):
+    # Without a proxy, a lookup failure must NOT break the launch → host TZ ("").
+    def boom(proxy=None, **kw):
+        raise GeoTimezoneError("offline")
+
+    monkeypatch.setattr(_geo, "discover_egress_ip", boom)
+    assert resolve_session_timezone("auto", None) == ""
+    assert resolve_session_timezone("", None) == ""
+
+
+@pytest.mark.unit
+def test_resolve_proxy_failure_raises(monkeypatch):
+    # With a proxy set, a failure must raise — never a silent host-TZ fallback.
+    def boom(proxy=None, **kw):
        raise GeoTimezoneError("no egress")

    monkeypatch.setattr(_geo, "discover_egress_ip", boom)
    with pytest.raises(GeoTimezoneError):
        resolve_session_timezone("auto", SOCKS)
+    with pytest.raises(GeoTimezoneError):
+        resolve_session_timezone("", SOCKS)
@@ -0,0 +1,131 @@
+"""Unit tests for the intelligent geoip mmdb auto-update in `download.py`.
+
+daijro/geoip-all-in-one rebuilds weekly; `ensure_geoip_mmdb` keeps the cache
+fresh without a download (or API call) on every launch. These tests mock the
+cache root, the latest-tag API, and the per-tag download so nothing touches the
+network.
+"""
+import os
+import time
+
+import pytest
+
+import invisible_playwright.download as dl
+
+
+@pytest.fixture
+def cache(tmp_path, monkeypatch):
+    """Point the cache at tmp_path and clear the env override."""
+    monkeypatch.setattr(dl, "cache_root", lambda: tmp_path)
+    monkeypatch.delenv("STEALTHFOX_GEOIP_MMDB", raising=False)
+    return tmp_path
+
+
+def _make_cached(root, tag, name=dl.GEOIP_MMDB_NAME):
+    d = root / "geoip" / tag
+    d.mkdir(parents=True, exist_ok=True)
+    f = d / name
+    f.write_bytes(b"FAKE-MMDB")
+    return f
+
+
+def _set_marker_age(root, days):
+    m = root / "geoip" / ".last_check"
+    m.parent.mkdir(parents=True, exist_ok=True)
+    m.touch()
+    old = time.time() - days * 86400
+    os.utime(m, (old, old))
+
+
+# ──────────────────────────────────────────────────────────────────────
+#  env override
+# ──────────────────────────────────────────────────────────────────────
+@pytest.mark.unit
+def test_env_override_returns_file(tmp_path, monkeypatch):
+    f = tmp_path / "mine.mmdb"
+    f.write_bytes(b"X")
+    monkeypatch.setenv("STEALTHFOX_GEOIP_MMDB", str(f))
+    assert dl.ensure_geoip_mmdb() == f
+
+
+@pytest.mark.unit
+def test_env_override_missing_raises(tmp_path, monkeypatch):
+    monkeypatch.setenv("STEALTHFOX_GEOIP_MMDB", str(tmp_path / "nope.mmdb"))
+    with pytest.raises(RuntimeError):
+        dl.ensure_geoip_mmdb()
+
+
+# ──────────────────────────────────────────────────────────────────────
+#  freshness window
+# ──────────────────────────────────────────────────────────────────────
+@pytest.mark.unit
+def test_fresh_cache_no_network(cache, monkeypatch):
+    f = _make_cached(cache, "2026.06.03")
+    _set_marker_age(cache, 0)  # just checked
+
+    def boom():
+        raise AssertionError("latest-tag API must NOT be called within the window")
+
+    monkeypatch.setattr(dl, "_latest_geoip_tag", boom)
+    assert dl.ensure_geoip_mmdb(max_age_days=7) == f
+
+
+@pytest.mark.unit
+def test_stale_same_tag_no_download(cache, monkeypatch):
+    f = _make_cached(cache, "2026.06.03")
+    _set_marker_age(cache, 30)  # stale → will re-check
+    monkeypatch.setattr(dl, "_latest_geoip_tag", lambda: "2026.06.03")
+    # real _download_geoip_tag runs but target exists, so no actual download:
+    monkeypatch.setattr(dl, "_download_file", lambda *a, **k: (_ for _ in ()).throw(
+        AssertionError("must not download when tag already cached")))
+    assert dl.ensure_geoip_mmdb(max_age_days=7) == f
+
+
+@pytest.mark.unit
+def test_stale_new_tag_downloads_and_prunes(cache, monkeypatch):
+    old = _make_cached(cache, "2026.06.03")
+    _set_marker_age(cache, 30)
+    monkeypatch.setattr(dl, "_latest_geoip_tag", lambda: "2026.06.10")
+
+    def fake_download(tag):
+        return _make_cached(cache, tag)  # simulate fetch+extract of the new tag
+
+    monkeypatch.setattr(dl, "_download_geoip_tag", fake_download)
+    got = dl.ensure_geoip_mmdb(max_age_days=7)
+    assert got.parent.name == "2026.06.10"
+    assert not old.parent.exists()  # old tag pruned
+    assert got.exists()
+
+
+# ──────────────────────────────────────────────────────────────────────
+#  offline resilience
+# ──────────────────────────────────────────────────────────────────────
+@pytest.mark.unit
+def test_api_down_with_cache_uses_cache(cache, monkeypatch):
+    f = _make_cached(cache, "2026.06.03")
+    _set_marker_age(cache, 30)
+
+    def boom():
+        raise OSError("offline")
+
+    monkeypatch.setattr(dl, "_latest_geoip_tag", boom)
+    assert dl.ensure_geoip_mmdb(max_age_days=7) == f  # stale cache reused, no raise
+
+
+@pytest.mark.unit
+def test_cold_cache_api_down_falls_back_to_pinned(cache, monkeypatch):
+    # no cache at all + API unreachable → pinned GEOIP_MMDB_VERSION fallback.
+    def boom():
+        raise OSError("offline")
+
+    monkeypatch.setattr(dl, "_latest_geoip_tag", boom)
+    captured = {}
+
+    def fake_download(tag):
+        captured["tag"] = tag
+        return _make_cached(cache, tag)
+
+    monkeypatch.setattr(dl, "_download_geoip_tag", fake_download)
+    got = dl.ensure_geoip_mmdb(max_age_days=7)
+    assert captured["tag"] == dl.GEOIP_MMDB_VERSION
+    assert got.exists()