feat: 增加制式和高干扰筛选

main.py

@@ -46,6 +46,22 @@ EXPECTED_TYPES = (
     "反开RD干扰监控",
 )
 
+NETWORK_TYPE_BY_SOURCE = {
+    "5G下FDD干扰监控": "2.6G",
+    "5G干扰监控": "2.6G",
+    "700M下FDD干扰监控": "700M",
+    "700M干扰监控": "700M",
+    "SDR_FDD干扰监控": "4G",
+    "SDR_TDD干扰监控": "4G",
+    "反开RD干扰监控": "4G",
+}
+
+MIN_INTERFERENCE_BY_NETWORK_TYPE = {
+    "2.6G": Decimal("-107"),
+    "700M": Decimal("-110"),
+    "4G": Decimal("-110"),
+}
+
 DEFAULT_STORAGE_ID = "stg_4d9a910d72"
 DEFAULT_ROOT = "/网优日常优化数据文档/(勿删)干扰定时小时指标"
 CELL_DATA_DIRECTORIES = (
@@ -155,6 +171,7 @@ EXPECTED_HEADERS = {
 SUMMARY_HEADER = (
     "hour_start",
     "hour_end",
+    "network_type",
     "cgi",
     "cell_name",
     "interference_dbm",
@@ -266,6 +283,7 @@ class ApiSummaryStore:
             f"""
                 CREATE TABLE IF NOT EXISTS `{self.table}` (
                     metric_time DATETIME NOT NULL COMMENT '指标开始时间',
+                    network_type VARCHAR(16) NOT NULL DEFAULT '',
                     cgi VARCHAR(128) NOT NULL,
                     cell_name VARCHAR(255) NOT NULL,
                     interference_dbm DECIMAL(10,3) NOT NULL,
@@ -285,6 +303,7 @@ class ApiSummaryStore:
             raise ProcessingError("Metrix Database API returned invalid column metadata")
         existing_columns = {str(item.get("name")) for item in columns if isinstance(item, dict)}
         migrations = {
+            "network_type": "VARCHAR(16) NOT NULL DEFAULT ''",
             "longitude": "DECIMAL(10,6) NULL",
             "latitude": "DECIMAL(10,6) NULL",
             "azimuth": "DECIMAL(6,2) NOT NULL DEFAULT 0",
@@ -318,7 +337,7 @@ class ApiSummaryStore:
                     START TRANSACTION;
                     DELETE FROM `{self.table}` WHERE metric_time = {metric_literal};
                     INSERT INTO `{self.table}`
-                        (metric_time, cgi, cell_name, interference_dbm, longitude, latitude, azimuth)
+                        (metric_time, network_type, cgi, cell_name, interference_dbm, longitude, latitude, azimuth)
                     VALUES
                         {values};
                     DELETE FROM `{self.table}` WHERE metric_time <> {metric_literal};
@@ -366,6 +385,7 @@ class ApiSummaryStore:
         return "(" + ", ".join(
             (
                 metric_literal,
+                sql_text_literal(row["network_type"]),
                 sql_text_literal(row["cgi"]),
                 sql_text_literal(row["cell_name"]),
                 sql_decimal_literal(row["interference_dbm"], "interference_dbm"),
@@ -623,6 +643,7 @@ def process(
     converted_dir.mkdir(parents=True)
 
     summary_rows: list[dict[str, str]] = []
+    threshold_filtered_rows = 0
     manifest_files: list[dict[str, object]] = []
     database_result: dict[str, object] = {"enabled": False}
     try:
@@ -634,9 +655,11 @@ def process(
             write_csv(converted_dir / csv_name, header, rows)
             for row_number, row in enumerate(rows, start=2):
                 record = dict(zip(header, row, strict=True))
-                summary_rows.append(
-                    summary_record(record, source_type, candidate.path, window, row_number, cell_metadata)
-                )
+                summary = summary_record(record, source_type, candidate.path, window, row_number, cell_metadata)
+                if passes_interference_threshold(summary):
+                    summary_rows.append(summary)
+                else:
+                    threshold_filtered_rows += 1
             manifest_files.append(
                 {
                     "source_size": candidate.size,
@@ -661,6 +684,7 @@ def process(
             "source_file_count": len(manifest_files),
             "cell_data_file_count": len(cell_data_workbooks),
             "summary_rows": len(summary_rows),
+            "threshold_filtered_rows": threshold_filtered_rows,
             "coordinate_matched_rows": matched_coordinates,
             "coordinate_unmatched_rows": len(summary_rows) - matched_coordinates,
             "warnings": warnings_out,
@@ -682,6 +706,7 @@ def process(
     print(f"source_files={len(manifest_files)}")
     print(f"cell_data_files={len(cell_data_workbooks)}")
     print(f"summary_rows={len(summary_rows)}")
+    print(f"threshold_filtered_rows={threshold_filtered_rows}")
     print(f"coordinate_matched_rows={matched_coordinates}")
     print(f"coordinate_unmatched_rows={len(summary_rows) - matched_coordinates}")
     if database_result["enabled"]:
@@ -728,6 +753,7 @@ def summary_record(
     return {
         "hour_start": hour_start,
         "hour_end": hour_end,
+        "network_type": NETWORK_TYPE_BY_SOURCE[source_type],
         "cgi": cgi,
         "cell_name": cell_name,
         "interference_dbm": interference,
@@ -737,6 +763,17 @@ def summary_record(
     }
 
 
+def passes_interference_threshold(record: dict[str, str]) -> bool:
+    network_type = record["network_type"]
+    try:
+        interference = Decimal(record["interference_dbm"])
+    except InvalidOperation as exc:
+        raise ProcessingError(f"Invalid interference value: {record['interference_dbm']}") from exc
+    if not interference.is_finite():
+        raise ProcessingError(f"Invalid interference value: {record['interference_dbm']}")
+    return interference >= MIN_INTERFERENCE_BY_NETWORK_TYPE[network_type]
+
+
 def build_cgi(prefix: str, record: dict[str, object], columns: tuple[str, ...], path: str, row: int) -> str:
     parts = [required(record, column, path, row) for column in columns]
     return "-".join(([prefix] if prefix else []) + parts)