1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
import json
import os
import re
import ast
import sys
import time
import logging
import traceback
import threading
import zlib
import pandas as pd
import numpy as np
import redis
from datetime import datetime
sys.path.append("/opt/module/spark-3.2.0-bin-hadoop3.2/demo/py_demo/")
sys.path.append(os.path.dirname(sys.path[0])) # 上级目录
from sqlalchemy import create_engine
from sqlalchemy.exc import PendingRollbackError
from utils.templates import Templates
# from ..utils.templates import Templates
from utils.templates_mysql import TemplatesMysql
# from ..utils.templates_mysql import TemplatesMysql
from pyspark.sql.types import IntegerType
from pyspark.sql import functions as F
from pyspark.sql.types import *
from utils.mysql_db import sql_connect, sql_update_many, sql_delete, get_country_engine
from pyspark.sql import SparkSession
class SpiderAsinSearch(Templates):
def __init__(self, site_name='us', date_type='week', date_info='2023-11-16', consumer_type='lastest', batch_size=100000):
super(SpiderAsinSearch, self).__init__()
self.site_name = site_name
self.date_info = date_info
self.consumer_type = consumer_type # 消费实时还是消费历史
self.date_type = date_type
# 通过date_type 获取 topic
self.get_topic_name()
# 通过date_type 获取 schema
self.init_schema()
# self.topic_name = topic_name # 主题名字
self.batch_size = batch_size
self.batch_size_history = int(batch_size / 10)
self.db_save = f'spider_asin_search'
# self.spark = self.create_spark_object(
# app_name=f"{self.db_save}: {self.site_name},{self.date_type}, {self.date_info}, {self.consumer_type}")
self.app_name = self.get_app_name()
self.spark = self.create_spark_object(app_name=f"{self.app_name}")
# self.schema = self.init_schema()
# 连接mysql
self.engine = get_country_engine(self.site_name)
self.pg14_engine = self.get_14pg_country_engine(self.site_name)
sql_connect(self.site_name)
logging.basicConfig(format='%(asctime)s %(name)s %(levelname)s %(message)s',
level=logging.INFO)
# 爬虫类型--流程表
self.spider_type = "反查搜索词"
def judge_spider_asin_detail_is_finished(self):
while True:
try:
sql = f'SELECT * from selection.workflow_progress WHERE page="反查搜索词" and site_name="{self.site_name}" and date_type="{self.date_type}" and date_info="{self.date_info}" and status_val=3'
df = pd.read_sql(sql, con=self.engine)
if df.shape[0] == 1:
print(f"ASIN详情状态为3, 抓取完成并终止程序, site_name:{self.site_name}, date_type:{self.date_type}, date_info:{self.date_info}")
self.spark.stop()
quit() # 退出程序
break
except Exception as e:
print(e, traceback.format_exc())
time.sleep(10)
self.engine = self.get_connection()
def init_schema(self):
self.schema = StructType([
StructField("cate_type", StringType(), True),
StructField("data_list", StringType(), True),
StructField("date_info", StringType(), True),
StructField("spider_time", StringType(), True),
])
# self.col = ['search_term', 'asin', 'page', 'buy_data', 'label']
@staticmethod
def clean_kafka_df(df):
df = df.withColumnRenamed("seller_id", "account_id")
# cols_python = ["asin", "parentAsin", "variat_num", "best_sellers_rank", "best_sellers_herf", "price", "rating",
# "brand", "brand", "account_id", "account_name", "account_url", "buy_box_seller_type",
# "volume", "weight", "weight_str", "launchTime", "total_comments", "page_inventory"]
# oneCategoryRank, aoVal, bsrOrders, bsrOrdersSale
# siteName volumeFormat weightFormat asinUpdateTime
# java那边插件的字段名称
cols_java = ['asin', 'parentAsin', 'asinVarNum', 'oneCategoryRank', 'bestSellersRank', 'lastHerf', 'aoVal', 'price', 'rating',
'bsrOrders', 'bsrOrdersSale', 'brandName', 'accountId', 'accountName', 'accountUrl', 'siteName', 'buyBoxSellerType',
'volume', 'volumeFormat', 'weight', 'weightFormat', 'launchTime', 'totalComments', 'pageInventory', 'asinUpdateTime']
df = df.select("asin", "parentAsin", "variat_num", "best_sellers_rank", "best_sellers_herf", "price", "rating",
"brand", "account_id", "account_name", "account_url", "buy_box_seller_type",
"volume", "weight", "weight_str", "launch_time", "total_comments", "page_inventory", "asinUpdateTime", "site_name", "node_id")
return df
def get_topic_name(self):
if self.date_type == "week":
self.topic_name = f"{self.site_name}_search_term_rank_type"
elif self.date_type == "month":
self.topic_name = f"{self.site_name}_search_term_month_{self.date_info.replace('-', '_')}"
else:
logging.info("self.date_type error -----")
quit()
def get_14pg_country_engine(self, site_name="us"):
h14_pg_us = {
"user": "postgres",
"password": "fazAqRRVV9vDmwDNRNb593ht5TxYVrfTyHJSJ3BS",
# "host": "61.145.136.61",
"host": "192.168.10.223",
"port": "5432",
# "port": 54328,
"database": "selection",
}
if site_name == 'us' or site_name == 'mx' or site_name == 'ca':
h14_pg_us["database"] = f"selection"
db_ = 'postgresql+psycopg2://{}:{}@{}:{}/{}'.format(*h14_pg_us.values())
else:
h14_pg_us["database"] = f"selection_{site_name}"
db_ = 'postgresql+psycopg2://{}:{}@{}:{}/{}'.format(*h14_pg_us.values())
engine = create_engine(db_, encoding='utf-8') # , pool_recycle=3600
return engine
def cate_type(self, cate_type, df):
if cate_type in ["sp", "zr"]:
df = df.loc[:, ['search_term', 'asin', 'page', 'page_row','date_info', 'created_time']]
df.drop_duplicates(['search_term', 'asin', 'page', 'page_row'], inplace=True)
return df
elif cate_type in ['buy']:
df = df.loc[:, ['search_term', 'asin', 'page', 'buy_data', 'date_info', 'label', 'created_time']]
df.drop_duplicates(['search_term', 'asin', 'page', 'buy_data', 'label'], inplace=True)
return df
else:
if cate_type in ['sb', 'tr']:
df = df.loc[:, ['search_term', 'asin', 'page', 'cate_type', 'date_info', 'created_time']]
df.rename(columns={"cate_type": "data_type"}, inplace=True)
df.drop_duplicates(['search_term', 'asin', 'page', 'data_type'], inplace=True)
return df
elif cate_type in ['buy']:
df = df.loc[:, ['search_term', 'asin', 'page', 'buy_data', 'date_info', 'label', 'created_time']]
df.drop_duplicates(['search_term', 'asin', 'page', 'buy_data', 'label'], inplace=True)
return df
else:
df = df.loc[:, ['search_term', 'asin', 'page', 'date_info', 'created_time']]
df.drop_duplicates(['search_term', 'asin', 'page'], inplace=True)
return df
def add_column_to_list(self, row):
l = []
for sub_list in json.loads(row['data_list']):
l.append(sub_list + [row["spider_time"]])
return l
def save_data_common(self, name, group):
logging.info(f"name: {name}")
search_exploded_list = group['data_list'].explode()
# 展开后转换为一个大列表
search_list = [i for i in search_exploded_list.tolist() if not isinstance(i, float)]
if search_list:
logging.info(f"搜索词处理{search_list[0:5]}")
# 列表等分
# self.list_svg(search_list, chunk_size=100)
# 转换为df对象
if name[0] in ['buy']:
# 将数据转换为df对象
df_1 = pd.DataFrame(data=search_list, columns=['search_term', 'asin', 'page', 'buy_data', 'label', 'created_time'])
else:
# if self.date_type == "week":
# # 将数据转换为df对象
df_1 = pd.DataFrame(data=search_list, columns=['search_term', 'asin', 'page', 'page_row',
'cate_type', 'title', 'img_url', 'price', 'rating',
'total_comments', 'created_time'])
# else:
# # 将数据转换为df对象
# df_1 = pd.DataFrame(data=search_list, columns=['search_term', 'asin', 'page', 'page_row',
# 'data_type', 'title', 'img_url', 'price', 'rating',
# 'total_comments', 'created_time'])
# 获取date_info
df_1['date_info'] = name[1]
# 通过类别对对应数据字段进行清洗
new_df = self.cate_type(name[0], df_1)
logging.info(f"{name} {new_df.shape} \n {new_df.keys()} {new_df.head()}")
# 通过站点 类别 和date_info 拼接表名
if name[0] == "buy":
if self.date_type == 'month':
table_name = f"{self.site_name}_other_search_term_month_{name[1].replace('-', '_')}"
else:
table_name = f"{self.site_name}_other_search_term_{name[1].replace('-', '_')}"
else:
if self.date_type == 'month':
table_name = f"{self.site_name}_search_term_rank_{name[0]}_month_{name[1].replace('-', '_')}"
else:
table_name = f"{self.site_name}_search_term_rank_{name[0]}_{name[1].replace('-', '_')}"
while True:
try:
start_time = time.time()
new_df.to_sql(name=f'{table_name}', con=self.pg14_engine, if_exists='append', index=False)
end_time = time.time()
logging.info(f"入库 {table_name} 表 {new_df.shape} {new_df.head(10)} 成功, 耗时：{end_time - start_time}s")
break
except PendingRollbackError as e:
time.sleep(3)
logging.info(f"error {e} sleep 3")
continue
def save_data(self, df):
threads = []
# 将字符串类型改为 python list
# df['data_list'] = df['data_list'].apply(json.loads)
df["data_list"] = df.apply(self.add_column_to_list, axis=1)
for name, group in df.groupby(['cate_type', 'date_info']):
thread = threading.Thread(target=self.save_data_common, args=(name, group))
threads.append(thread)
thread.start()
for thread in threads:
thread.join()
logging.info("线程处理完成")
def data_save(self, df):
if not isinstance(df, pd.DataFrame):
logging.info("df 不是一个 DataFrame 对象 ")
df = df.toPandas()
if self.consumer_type != "latest":
logging.info(f"过滤前： {df.shape}")
df = df.loc[(~df.date_info.isna()) & (df.date_info == self.date_info)]
logging.info(f"过滤后： {df.shape}")
if self.date_type == "week":
logging.info(f"周数据处理")
if df.shape[0]:
self.save_data(df)
else:
logging.info("处理month数据")
if df.shape[0]:
self.save_data(df)
def handle_kafka_history(self, kafka_df):
self.data_save(kafka_df)
def handle_kafka_stream(self, kafka_df, epoch_id):
print(type(kafka_df), kafka_df)
count = kafka_df.count()
print("当前批次传输的数据量为df.count():", count)
if count == 0:
print(f"当前批次{epoch_id}没有数据进来")
self.judge_spider_asin_detail_is_finished()
else:
self.data_save(kafka_df)
if __name__ == '__main__':
site_name = sys.argv[1] # 参数1：站点
date_type = sys.argv[2] # 参数2：类型：week/4_week/month/quarter/day
date_info = sys.argv[3] # 参数3：年-周/年-月/年-季/年-月-日, 比如: 2022-1
consumer_type = sys.argv[4] # 参数3：实时 lastest 历史 history
handle_obj = SpiderAsinSearch(site_name=site_name, date_type=date_type, date_info=date_info, consumer_type=consumer_type, batch_size=10000)
handle_obj.run_kafka()
# for i in `ps -ef|grep "spider_asin_search" |awk '{print $2}' `; do kill -9 $i ; done;
# 历史
# /opt/module/spark/bin/spark-submit --packages org.apache.spark:spark-sql-kafka-0-10_2.12:3.1.3 --master yarn --driver-memory 2g --executor-memory 20g --executor-cores 4 --num-executors 2 --queue spark /opt/module/spark/demo/py_demo/my_kafka/spider_asin_search.py it week 2023-49 history > amazon_history_search_it.log 2>&1 &
# 实时
# /opt/module/spark/bin/spark-submit --packages org.apache.spark:spark-sql-kafka-0-10_2.12:3.1.3 --master yarn --driver-memory 2g --executor-memory 2g --executor-cores 4 --num-executors 2 --queue spark /opt/module/spark/demo/py_demo/my_kafka/spider_asin_search.py it week 2023-49 latest > amazon_latest_search_it.log 2>&1 &