Files
enduro-trails/tests/api/test_gps_tracks_sources_osm.py
claude-bot edbe9a3044
Some checks failed
CI / lint (push) Failing after 4s
CI / test (push) Failing after 4s
CI / build (push) Has been skipped
CI / lint (pull_request) Failing after 3s
CI / test (pull_request) Failing after 4s
CI / build (pull_request) Has been skipped
fix(gps-tracks): normalise GeoJSON props, add health fields, OSM meta fetch, z-order fix
Refs: ET-008

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-06-01 12:42:36 +00:00

249 lines
9.3 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
"""Unit тесты для OSM GPS-источника (ET-008).
U-42: split_bbox_for_osm разбивает правильно
U-43: длина через Haversine
U-44: защита от XXE через defusedxml
"""
import os
import pytest
from src.api.gps_tracks.sources.osm import (
OsmParser,
split_bbox_for_osm,
_haversine_m,
_parse_gpx_trackpoints,
)
# ─── U-42: split_bbox_for_osm ────────────────────────────────────────────────
def test_u42_split_bbox_basic():
"""U-42: корректное разбиение на ячейки."""
bbox = (37.0, 55.0, 38.0, 56.0) # 1° x 1°
cells = split_bbox_for_osm(bbox, cell_size=0.25)
# 1° / 0.25° = 4 ячейки по каждой оси = 16 ячеек
assert len(cells) == 16
def test_u42_split_bbox_cell_size():
"""U-42: каждая ячейка не больше cell_size по размеру."""
bbox = (29.0, 49.5, 47.5, 60.0) # ЦФО
cells = split_bbox_for_osm(bbox, cell_size=0.25)
for cell in cells:
west, south, east, north = cell
assert east - west <= 0.25 + 1e-9
assert north - south <= 0.25 + 1e-9
def test_u42_split_bbox_covers_region():
"""U-42: все ячейки вместе покрывают весь регион."""
bbox = (37.0, 55.0, 38.0, 56.0)
cells = split_bbox_for_osm(bbox, cell_size=0.25)
min_lon = min(c[0] for c in cells)
min_lat = min(c[1] for c in cells)
max_lon = max(c[2] for c in cells)
max_lat = max(c[3] for c in cells)
assert abs(min_lon - 37.0) < 1e-9
assert abs(min_lat - 55.0) < 1e-9
assert abs(max_lon - 38.0) < 0.25 + 1e-9 # последняя ячейка обрезается
assert abs(max_lat - 56.0) < 0.25 + 1e-9
def test_u42_split_small_bbox():
"""U-42: bbox меньше cell_size даёт одну ячейку."""
bbox = (37.0, 55.0, 37.1, 55.1)
cells = split_bbox_for_osm(bbox, cell_size=0.25)
assert len(cells) == 1
def test_u42_split_bbox_no_overlap():
"""U-42: ячейки не перекрываются (west следующей = east предыдущей)."""
bbox = (37.0, 55.0, 37.5, 55.25)
cells = split_bbox_for_osm(bbox, cell_size=0.25)
# При bbox шириной 0.5° и cell_size=0.25 должно быть 2 ячейки по оси lon
assert len(cells) == 2
# Восток первой ячейки = запад второй
cells_sorted = sorted(cells, key=lambda c: c[0])
assert abs(cells_sorted[0][2] - cells_sorted[1][0]) < 1e-9
# ─── U-43: Haversine длина ───────────────────────────────────────────────────
def test_u43_haversine_known_distance():
"""U-43: проверка haversine на известном расстоянии."""
# Москва (37.617, 55.755) → Химки (37.425, 55.889) ≈ 20 км
dist = _haversine_m(37.617, 55.755, 37.425, 55.889)
assert 18000 < dist < 22000
def test_u43_haversine_zero_distance():
"""U-43: одна точка → расстояние 0."""
dist = _haversine_m(37.617, 55.755, 37.617, 55.755)
assert dist == pytest.approx(0.0, abs=1e-6)
def test_u43_haversine_symmetry():
"""U-43: расстояние A→B = B→A."""
d1 = _haversine_m(37.617, 55.755, 37.425, 55.889)
d2 = _haversine_m(37.425, 55.889, 37.617, 55.755)
assert abs(d1 - d2) < 1e-6
def test_u43_haversine_short_distance():
"""U-43: короткое расстояние (~111 м на экваторе при 0.001° по lon)."""
dist = _haversine_m(0.0, 0.0, 0.001, 0.0)
assert 100 < dist < 120
# ─── U-44: защита от XXE ─────────────────────────────────────────────────────
def test_u44_xxe_protection():
"""U-44: defusedxml блокирует XXE атаку."""
fixture_path = os.path.join(
os.path.dirname(__file__),
"../../tests/fixtures/gps-tracks/xxe-payload.gpx",
)
with open(fixture_path, "rb") as f:
content = f.read()
# Должен либо выбросить исключение, либо вернуть пустой список без чтения /etc/passwd
try:
tracks = _parse_gpx_trackpoints(content, "osm", "")
# Если парсинг прошёл без ошибки — проверяем что /etc/passwd не попал в данные
for track in tracks:
assert "root:" not in str(track)
assert "/bin/" not in str(track)
except Exception:
# defusedxml выбросил исключение — это ожидаемое поведение
pass
def test_u44_valid_gpx_parsed_correctly():
"""U-44: корректный GPX с gpx_id парсится правильно."""
fixture_path = os.path.join(
os.path.dirname(__file__),
"../../tests/fixtures/gps-tracks/osm-trackpoints-bbox-moscow.gpx",
)
with open(fixture_path, "rb") as f:
content = f.read()
tracks = _parse_gpx_trackpoints(content, "osm", "")
assert len(tracks) == 1
track = tracks[0]
assert track.points_count == 3
assert abs(track.min_lat - 55.751) < 0.001
assert abs(track.max_lat - 55.753) < 0.001
assert track.source_id == "osm"
def test_u44_anonymous_trackpoints_skipped():
"""U-44: анонимные точки без gpx_id пропускаются."""
gpx_without_ids = b"""<?xml version="1.0"?>
<gpx version="1.0" creator="OpenStreetMap.org" xmlns="http://www.topografix.com/GPX/1/0">
<trk>
<trkseg>
<trkpt lat="55.751" lon="37.618"><time>2024-05-12T10:00:00Z</time></trkpt>
<trkpt lat="55.752" lon="37.619"><time>2024-05-12T10:01:00Z</time></trkpt>
</trkseg>
</trk>
</gpx>"""
tracks = _parse_gpx_trackpoints(gpx_without_ids, "osm", "")
assert len(tracks) == 0
def test_u44_multiple_tracks_in_gpx():
"""U-44: несколько gpx_id в одном ответе парсятся как разные треки."""
gpx_multi = b"""<?xml version="1.0"?>
<gpx version="1.0" creator="OpenStreetMap.org" xmlns="http://www.topografix.com/GPX/1/0">
<trk>
<trkseg>
<trkpt lat="55.751" lon="37.618" gpx_id="111"><time>2024-05-12T10:00:00Z</time></trkpt>
<trkpt lat="55.752" lon="37.619" gpx_id="111"><time>2024-05-12T10:01:00Z</time></trkpt>
<trkpt lat="55.760" lon="37.700" gpx_id="222"><time>2024-05-13T08:00:00Z</time></trkpt>
<trkpt lat="55.765" lon="37.710" gpx_id="222"><time>2024-05-13T08:05:00Z</time></trkpt>
</trkseg>
</trk>
</gpx>"""
tracks = _parse_gpx_trackpoints(gpx_multi, "osm", "")
assert len(tracks) == 2
ids = {t.external_id for t in tracks}
assert "111" in ids
assert "222" in ids
# ─── U-45: _parse_gpx_meta_response ──────────────────────────────────────────
def test_u45_meta_response_with_known_tag():
"""U-45: _parse_gpx_meta_response extracts activity via MAPPING."""
from src.api.gps_tracks.sources.osm import _parse_gpx_meta_response
content = b"""<?xml version="1.0"?>
<osm version="0.6">
<gpx_file id="123" name="my_ride.gpx" user="alice">
<description>Weekend ride</description>
<tag>enduro</tag>
<tag>motorcycle</tag>
</gpx_file>
</osm>"""
meta = _parse_gpx_meta_response(content)
assert meta is not None
assert meta["activity_type"] == "enduro"
assert meta["name"] == "my_ride.gpx"
assert meta["user"] == "alice"
assert meta["description"] == "Weekend ride"
def test_u45_meta_response_unknown_tag_returns_none_activity():
"""U-45: unknown tag → activity_type is None."""
from src.api.gps_tracks.sources.osm import _parse_gpx_meta_response
content = b"""<?xml version="1.0"?>
<osm version="0.6">
<gpx_file id="99" name="trip.gpx" user="bob">
<tag>unknown-sport</tag>
</gpx_file>
</osm>"""
meta = _parse_gpx_meta_response(content)
assert meta is not None
assert meta["activity_type"] is None
def test_u45_meta_response_motorcycle_maps_to_moto():
"""U-45: 'motorcycle' tag maps to 'moto' via MAPPING."""
from src.api.gps_tracks.sources.osm import _parse_gpx_meta_response
content = b"""<?xml version="1.0"?>
<osm version="0.6">
<gpx_file id="77" name="ride.gpx" user="carl">
<tag>motorcycle</tag>
</gpx_file>
</osm>"""
meta = _parse_gpx_meta_response(content)
assert meta["activity_type"] == "moto"
def test_u45_meta_response_invalid_xml_returns_none():
"""U-45: malformed XML returns None."""
from src.api.gps_tracks.sources.osm import _parse_gpx_meta_response
meta = _parse_gpx_meta_response(b"not xml at all <<<")
assert meta is None
def test_u45_meta_response_no_gpx_file_element():
"""U-45: valid XML but no gpx_file element → result has all None values."""
from src.api.gps_tracks.sources.osm import _parse_gpx_meta_response
content = b"""<?xml version="1.0"?><osm version="0.6"></osm>"""
meta = _parse_gpx_meta_response(content)
# Function should return the dict with None values, not None itself
assert meta is not None
assert meta["activity_type"] is None
assert meta["name"] is None