chore: curate monthly assessment evidence and scripts

backfill_july_all_gps.py

@@ -0,0 +1,101 @@
+from ast import parse, walk, Call, Constant, keyword
+from collections import defaultdict
+from datetime import datetime, date
+from pathlib import Path
+import re
+import shutil
+
+import openpyxl
+import pymysql
+
+
+OUTPUT = Path('/Users/kkfluous/Downloads/7月核算结果/7月考核车辆多源里程汇总.xlsx')
+BACKUP = Path('/Users/kkfluous/Downloads/7月核算结果/7月考核车辆多源里程汇总_GPS全量回补前.xlsx')
+
+
+def norm(value):
+    return re.sub(r'\s+', '', str(value or '')).upper()
+
+
+def db_config():
+    tree = parse(Path('/Users/kkfluous/Downloads/replace_gps_jun.py').read_text())
+    for node in walk(tree):
+        if not isinstance(node, Call) or getattr(node.func, 'attr', None) != 'connect':
+            continue
+        values = {}
+        for item in node.keywords:
+            if isinstance(item, keyword) and isinstance(item.value, Constant):
+                values[item.arg] = item.value.value
+        if {'host', 'user', 'password', 'database'} <= values.keys():
+            return values
+    raise RuntimeError('未找到既有GPS数据库配置')
+
+
+wb = openpyxl.load_workbook(OUTPUT)
+ws = wb.active
+header_row = next(r for r in range(1, 8) if ws.cell(r, 12).value == 'GPS里程(km)')
+data_start = header_row + 1
+plates = sorted({norm(ws.cell(r, 2).value) for r in range(data_start, ws.max_row + 1) if ws.cell(r, 2).value})
+
+cfg = db_config()
+cfg.update(charset='utf8mb4', connect_timeout=8, read_timeout=30, write_timeout=30)
+conn = pymysql.connect(**cfg)
+try:
+    with conn.cursor() as cur:
+        placeholders = ','.join(['%s'] * len(plates))
+        cur.execute(
+            f'''SELECT plate_number, dates, total_mileage
+                FROM ln_vehicle_g7_mileage
+                WHERE dates >= %s AND dates <= %s
+                  AND plate_number IN ({placeholders})''',
+            [date(2026, 7, 1), date(2026, 7, 31), *plates],
+        )
+        db_rows = cur.fetchall()
+finally:
+    conn.close()
+
+daily = defaultdict(dict)
+for plate, day, raw_mileage in db_rows:
+    if isinstance(day, datetime):
+        day = day.date()
+    daily[norm(plate)][day] = float(raw_mileage or 0) / 100000.0
+
+if not BACKUP.exists():
+    shutil.copy2(OUTPUT, BACKUP)
+
+filled = []
+still_missing = []
+for row in range(data_start, ws.max_row + 1):
+    cell = ws.cell(row, 12)
+    if cell.value != '无来源':
+        continue
+    plate = norm(ws.cell(row, 2).value)
+    start = ws.cell(row, 4).value
+    end = ws.cell(row, 5).value
+    if isinstance(start, datetime):
+        start = start.date()
+    if isinstance(end, datetime):
+        end = end.date()
+    matched = [value for day, value in daily.get(plate, {}).items() if start <= day <= end]
+    if matched:
+        cell.value = round(sum(matched), 2)
+        cell.number_format = '#,##0.0'
+        filled.append((plate, start, end, cell.value))
+    else:
+        still_missing.append((plate, start, end))
+
+ws['A1'] = (
+    '1. 纯氢里程(km)、纯电里程(km)来源于广州平台统计，TBOX里程来源于车机上报的仪表盘里程统计，'
+    'GPS里程优先使用两份7月GPS日明细，缺失车辆由GPS数据库每日里程按考核区间回补；测试车辆以7月最终考核源表已确认GPS为准\n'
+    '2. 广州平台总里程(km)=纯氢里程(km)+纯电里程(km)\n'
+    '3. 已确认广州平台暂未接入86辆普货车数据；所有GPS来源均无对应记录时标记为“无来源”'
+)
+wb.save(OUTPUT)
+
+print({
+    'database_rows': len(db_rows),
+    'database_plates': len(daily),
+    'backfilled_records': len(filled),
+    'remaining_missing_records': len(still_missing),
+    'remaining_missing_plates': len({x[0] for x in still_missing}),
+})