data_cleaning/task/summary_func.py
2020-12-23 13:35:42 +08:00

51 lines
1.9 KiB
Python

from pymongo import UpdateOne
from pydantic import Field
import pandas as pd
from .task import Task
from utils import *
from db import GBaseModel
class SummaryFunc(Task):
"""
功能分析
"""
class Model(GBaseModel):
prize: list = Field(None, title='奖励')
need: list = Field(None, title='消耗')
ftype: str = Field(..., title='功能')
data: dict = Field(None, title='功能数据')
cdate: int = Field(..., title='当天0点')
def cleaning(self, cursor_list):
for cursor in cursor_list: # type:dict
for event_coll, ts in cursor.items(): # type:str,dict
if ts['cursor_st'] == ts['cursor_et']:
continue
logger.info(f'开始处理{self.game_name} 处理 {event_coll} 游标 {ts}')
where = {
'_event_name': 'Func',
'_event_time': {
'$gte': ts['cursor_st'],
'$lt': ts['cursor_et'],
}
}
projection = self.Model.get_fields()
bulk_data = []
for item in self.local_db[event_coll].find(where, projection):
try:
item['cdate'] = int(pd.Timestamp(item['_event_time'], unit='s', tz=self.timezone) \
.normalize().timestamp())
model = self.Model(**item)
data = model.dict(by_alias=True)
bulk_data.append(UpdateOne({'_id': data['_id']}, {'$set': data}, upsert=True))
except Exception as e:
logger.error(f'ftype {item["ftype"]} msg:{e}')
# pass
if bulk_data:
self.remote_db[self.dest_coll].bulk_write(bulk_data, ordered=False)
self.set_cursor(cursor_st=ts['cursor_st'], cursor_et=ts['cursor_et'])