Metrix/InterferenceETL
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: 增加小区经纬度匹配

Browse Source
This commit is contained in:
nixevol
2026-08-03 10:42:45 +08:00
parent 1a6658dd03
commit 470763a818
4 changed files with 222 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/test_pipeline.py
+70 -1
View File
@@ -27,6 +27,7 @@ class PipelineTest(unittest.TestCase):
for source_type in main.EXPECTED_TYPES:
create_archive(root, source_type, COMPLETE_WINDOW)
create_archive(root, main.EXPECTED_TYPES[0], INCOMPLETE_WINDOW)
create_cell_data_sources(root)
store = RecordingStore()
result = main.process(main.LocalSource(root), output, lookback_days=3, store=store)
@@ -37,17 +38,26 @@ class PipelineTest(unittest.TestCase):
with (result / f"interference_summary_{COMPLETE_WINDOW}.csv").open(encoding="utf-8-sig", newline="") as file:
rows = list(csv.DictReader(file))
self.assertEqual(len(rows), 7)
self.assertEqual(list(rows[0]), ["hour_start", "hour_end", "cgi", "cell_name", "interference_dbm"])
self.assertEqual(
list(rows[0]),
["hour_start", "hour_end", "cgi", "cell_name", "interference_dbm", "longitude", "latitude"],
)
rows_by_type = {row["cell_name"].removesuffix("-小区"): row for row in rows}
self.assertEqual(rows_by_type["5G干扰监控"]["cgi"], "460-00-200-1")
self.assertEqual(rows_by_type["700M干扰监控"]["cgi"], "460-00-200-1")
for source_type in set(main.EXPECTED_TYPES) - {"5G干扰监控", "700M干扰监控"}:
self.assertEqual(rows_by_type[source_type]["cgi"], "460-00-100-1")
self.assertTrue(all(row["interference_dbm"] == "-100.5" for row in rows))
self.assertEqual(rows_by_type["5G干扰监控"]["longitude"], "113.123456")
self.assertEqual(rows_by_type["5G干扰监控"]["latitude"], "22.654321")
self.assertTrue(all(not row["longitude"] and not row["latitude"] for row in rows if row["cgi"] == "460-00-100-1"))
manifest = json.loads((result / "manifest.json").read_text(encoding="utf-8"))
self.assertEqual(manifest["source_file_count"], 7)
self.assertEqual(manifest["cell_data_file_count"], 5)
self.assertEqual(manifest["summary_rows"], 7)
self.assertEqual(manifest["coordinate_matched_rows"], 2)
self.assertEqual(manifest["coordinate_unmatched_rows"], 5)
self.assertEqual(len(manifest["warnings"]), 1)
self.assertIn(INCOMPLETE_WINDOW, manifest["warnings"][0])
self.assertNotIn("source_types", manifest)
@@ -78,6 +88,30 @@ class PipelineTest(unittest.TestCase):
("2026-07-30 23:00:00", "2026-07-31 00:00:00"),
)
def test_latest_dated_cell_data_file_is_selected(self) -> None:
entries = [
{"name": "江门5G小区信息表20260727.xlsx", "path": "/old.xlsx", "is_dir": False},
{"name": "江门5G小区信息表20260728.xlsx", "path": "/latest.xlsx", "is_dir": False},
{"name": "说明.txt", "path": "/说明.txt", "is_dir": False},
]
selected = main.select_latest_cell_data_file(entries, "/cell-data")
self.assertEqual(selected["path"], "/latest.xlsx")
def test_cell_data_schema_change_is_rejected(self) -> None:
raw = create_cell_data_workbook(include_latitude=False)
with self.assertRaisesRegex(main.ProcessingError, "missing columns.*纬度"):
main.parse_cell_data_workbook(raw, "bad.xlsx")
def test_conflicting_cell_data_coordinates_are_rejected(self) -> None:
first = create_cell_data_workbook(longitude=113.1)
second = create_cell_data_workbook(longitude=113.2)
with self.assertRaisesRegex(main.ProcessingError, "Conflicting CellData coordinates"):
main.load_cell_coordinates([("first.xlsx", first), ("second.xlsx", second)])
def test_mysql_store_replaces_same_hour_and_deletes_other_hours(self) -> None:
connection = FakeConnection(datetime(2026, 7, 31, 10), same_hour_rows=7, old_rows=20)
store = main.MySQLSummaryStore(connect_factory=lambda: connection)
@@ -89,11 +123,15 @@ class PipelineTest(unittest.TestCase):
self.assertEqual(result["refreshed_rows"], 7)
self.assertEqual(result["old_rows_deleted"], 20)
self.assertEqual(connection.cursor_instance.inserted[0][0], datetime(2026, 7, 31, 10))
self.assertIsNone(connection.cursor_instance.inserted[0][4])
self.assertIsNone(connection.cursor_instance.inserted[0][5])
self.assertTrue(connection.committed)
self.assertFalse(connection.rolled_back)
self.assertTrue(connection.closed)
self.assertNotIn("source_type", " ".join(connection.cursor_instance.queries))
self.assertNotIn("source_path", " ".join(connection.cursor_instance.queries))
self.assertTrue(any("ADD COLUMN `longitude`" in query for query in connection.cursor_instance.queries))
self.assertTrue(any("ADD COLUMN `latitude`" in query for query in connection.cursor_instance.queries))
def test_mysql_store_refuses_to_replace_a_newer_hour(self) -> None:
connection = FakeConnection(datetime(2026, 7, 31, 11))
@@ -169,6 +207,9 @@ class FakeCursor:
def fetchone(self) -> tuple[datetime | None]:
return (self.latest_time,)
def fetchall(self) -> list[tuple[str]]:
return [(name,) for name in ("metric_time", "cgi", "cell_name", "interference_dbm")]
def close(self) -> None:
self.closed = True
@@ -224,6 +265,8 @@ def database_row() -> dict[str, str]:
"cgi": "460-00-200-1",
"cell_name": "测试小区",
"interference_dbm": "-100.5",
"longitude": "",
"latitude": "",
}
@@ -273,5 +316,31 @@ def mock_row(source_type: str, window: str) -> list[object]:
return [values[column] for column in header]
def create_cell_data_sources(root: Path) -> None:
raw = create_cell_data_workbook()
for index, configured_directory in enumerate(main.CELL_DATA_DIRECTORIES, start=1):
directory = root / Path(configured_directory.lstrip("/"))
directory.mkdir(parents=True, exist_ok=True)
(directory / f"江门小区信息表{index}-20260728.xlsx").write_bytes(raw)
def create_cell_data_workbook(longitude: float = 113.123456, include_latitude: bool = True) -> bytes:
workbook = Workbook()
sheet = workbook.active
sheet.title = "小区信息表"
header = ["小区名称", "eNB/gNB", "CI", "经度"]
if include_latitude:
header.append("纬度")
sheet.append(header)
row: list[object] = ["测试小区", 200, 1, longitude]
if include_latitude:
row.append(22.654321)
sheet.append(row)
content = io.BytesIO()
workbook.save(content)
workbook.close()
return content.getvalue()
if __name__ == "__main__":
unittest.main()