def main(): # redis subscriber index:0-3
redisConnect = red.redis_connection('linode1', 'redis', db=0)
client = mon.mongo_connection('linode1', 'mongo')
coll_stockInfo = mon.mongo_collection(client, 'stocks', 'stockInfo')
while True:
# get keys and values from mongo
keys = red.redis_get_all_kv(redisConnect)
for key in keys:
amount = int(os.environ.get("amount")) # amount of subscriber
index = int(os.environ.get("index")) # subscriber num
num = int(key.split('No_')[-1]) # redis key
# 決定subscriber要取用哪筆資料
if num % int(amount) == int(index):
stock_id = red.redis_get_value(redisConnect, key)
print(f"get stock {stock_id}")
red.redis_delete_key(redisConnect, key) # 取出stock_id就從redis刪掉
coll_stockInfo.update({'_id': stock_id},
{'$set': {
'crawlerStatus': 2
}}) # 表示已經從redis刪掉
coll_stock = mon.mongo_collection(client, 'stocks',
f"stock{stock_id}")
for year in range(2010, 2022):
for month in range(1, 13):
if year == 2021 and month > 2:
break
url = f"""https://www.twse.com.tw/exchangeReport/STOCK_DAY?response=html&date={str(year)}{str(month).zfill(2)}01&stockNo={stock_id}"""
print(f"-- Crawler >>> {url}")
documents = crawler.crawler(url)
if documents:
for item in documents:
# 記錄爬取的股票資料並寫入mongo
mon.insert_document(coll_stock, item)
coll_stockInfo.update_one({'_id': stock_id}, {
'$set': {
'monthStatus':
str(year) + str(month).zfill(2)
}
}) # 當月爬完
print(
f'stock: {stock_id} in {year}{month} insert done.'
)
time.sleep(10)
print(
f'stock: {stock_id} in {year}{month} crawl done.')
coll_stockInfo.update_one({'_id': stock_id},
{'$set': {
'yearStatus': year
}}) # 當年爬完
coll_stockInfo.update({'_id': stock_id},
{'$set': {
'crawlerStatus': 3
}}) # 表示這支股票已經都爬完
def main(): # redis subscriber index:0-3
client = mon.mongo_connection('linode1', 'mongo')
with open('/Users/huangyiling/Github/stock/double_check_stock1.csv', newline='') as file:
rows = csv.reader(file)
for stock_id in rows:
stock_id = ''.join(stock_id)
print(f"get stock {stock_id}")
coll_stock = mon.mongo_collection(
client, 'stocks', f"stock{stock_id}")
# for year in range(2010, 2021):
# 測試當年度是否有資料
# test_month = 12
# test_url = f"""https://www.twse.com.tw/exchangeReport/STOCK_DAY?response=html&date={str(year)}{str(test_month).zfill(2)}01&stockNo={stock_id}"""
# print(f"test stock {stock_id} in {year} exist ?")
# test_docs = crawler.crawler(test_url)
# if test_docs:
# print("=> Yes, exist!")
for month in range(1, 13):
url = f"""https://www.twse.com.tw/exchangeReport/STOCK_DAY?response=html&date=2020{str(month).zfill(2)}01&stockNo={stock_id}"""
print(f"-- Crawler >>> {url}")
documents = crawler.crawler(url)
if documents:
# print(documents)
for item in documents:
# 記錄爬取的股票資料並寫入mongo
mon.insert_document(coll_stock, item)
print(
f'stock: {stock_id} in 2020{str(month).zfill(2)} insert done.')
time.sleep(10)
print(
f'stock: {stock_id} in 2020{str(month).zfill(2)} crawl done.')
def count_url_check():
expected_count = 938 * 11 * 12
collection = mon.mongo_collection(mon.mongo_connection('linode1', 'mongo'),
'stocks', 'crawlerURL')
# reality_count = collection.find({}).count()
reality_count = collection.count_documents
print("Expected :", expected_count)
print("Reality :", reality_count)
def count_stocks():
stock_count = 0
client = mon.mongo_connection('linode1', 'mongo')
collection = mon.mongo_collection(client, 'stocks', 'stockIndustry')
contents = collection.find({}, {'stocks_count': 1})
for item in contents:
id_list_count = item['stocks_count']
stock_count += id_list_count
print('stocks_count :', stock_count)
def all_stock_id():
client = mon.mongo_connection('linode1', 'mongo')
collection_stock = mon.mongo_collection(client, 'stocks', "stockInfo")
contents = list(collection_stock.find({}, {'_id': 1}))
for item in contents:
stock_id = item['_id']
# print(stock_id)
print("amount of stocks:", len(contents))
return contents
def create_urls(stock_ids):
mongoClient = mon.mongo_connection('linode1', 'mongo')
mongoCollection = mon.mongo_collection(mongoClient, 'stocks', 'crawlerURL')
for stock_id in stock_ids:
for year in year_list:
for month in month_list:
url = f'https://www.twse.com.tw/exchangeReport/STOCK_DAY?response=html&date={year + month.zfill(2)}01&stockNo={stock_id}'
doc = {
'_id': stock_id + year + month.zfill(2),
'url': url,
'crawlerStatus': 0
}
mon.insert_document(mongoCollection, doc)
def industry_crawler():
url = 'https://www.cnyes.com/twstock/stock_astock.aspx?ga=nav'
headers = {
'user-agent':
'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/84.0.4147.105 Safari/537.36'
}
res = requests.get(url, headers=headers)
# print(res.status_code)
soup = BeautifulSoup(res.text, 'html.parser')
industries = soup.select('div[id="kinditem_0"]>ul[class="kdlist"]>li')
# get all industries
for industry in industries:
industry_name = industry.a.text
print(industry_name)
industry_url = 'https://www.cnyes.com/twstock/' + industry.a["href"]
print(industry_url)
industry_id = industry_url.split('groupId=')[-1].split('&stitle')[0]
# get all stocks from the industry
res_stock = requests.get(industry_url, headers=headers)
# print(res_stock.status_code)
soup_stock = BeautifulSoup(res_stock.text, 'html.parser')
stocks = soup_stock.select('div[class="TableBox"]>table>tr')
stock_list = []
stock_dict = dict()
for stock in stocks[1:]:
stock_info = stock.find_all('td')
stock_id = stock_info[1].text
# print(stock_id)
stock_name = stock_info[2].text
# print(stock_name)
stock_list.append(stock_id)
stock_dict[stock_id] = stock_name
industry_key_id = 'industry_' + industry_id
doc = {
'_id': industry_key_id,
'industry': industry_kv[industry_name],
'industry_name': industry_name,
'stocks_list': stock_list,
'stocks_count': len(stock_list),
'stocks': stock_dict
}
# print(doc)
mongo_client = mon.mongo_connection('linode1', 'mongo')
mongo_collection = mon.mongo_collection(mongo_client, 'stocks',
'stockIndustry')
mon.insert_document(mongo_collection, doc)
time.sleep(20)
def get_month_record_mongo(stock_id, year, month):
client = mon.mongo_connection('linode1', 'mongo')
collection = mon.mongo_collection(client, 'stocks', f"stock{stock_id}")
# 檢查當是否有存資料
if collection.find({
"trade_date": {
"$regex": f"{str(year)+str(month)}"
}
}, {
"trade_date": 1
}).count() != 0:
contents = list(
collection.find({'_id': re.compile(f"{stock_id}{year}{month}")}))
return contents
else:
return
def pick_proxy(amount=10):
client = mon.mongo_connection('linode1', 'mongo')
collection = mon.mongo_collection(client, 'proxy', 'proxyPool_1')
contents = collection.find({}, {'ip': 1, 'port': 1}).limit(amount)
redisConnect = red.redis_connection('linode1', 'redis',
db=1) # proxy use db=1
for item in contents:
proxy = item['ip'] + ':' + item['port']
try:
validate_proxy(proxy)
for index in range(1, 21):
key = f'proxy{index}'
if not redisConnect.exists(key):
red.redis_set_key_value(redisConnect, key, proxy)
except Exception:
pass
def check_records_exist():
client = mon.mongo_connection('linode1', 'mongo')
for content in allStockID.all_stock_id():
stock_id = content['_id']
# print(stock_id)
coll_stock = mon.mongo_collection(client, 'stocks', f"stock{stock_id}")
# stocks_con = list(coll_stock.find(
# {"trade_date": {"$regex": "2020"}}, {"trade_date": 1}))
records_count = coll_stock.find({
"trade_date": {
"$regex": "202102"
}
}, {
"trade_date": 1
}).count()
if records_count < 5:
print(stock_id, records_count)
wcsv.writeToCsv("double_check_stock", [stock_id])
def stockInfo():
client = mon.mongo_connection('linode1', 'mongo')
coll_stockIndustry = mon.mongo_collection(s
client, 'stocks', 'stockIndustry')
coll_stockInfo = mon.mongo_collection(client, 'stocks', 'stockInfo')
for item in mon.find_all_mongo(coll_stockIndustry):
# print(item)
ids = item['stocks_list']
for stock_id in ids:
doc = {
'_id': stock_id,
'industry': item['_id'],
'name': 'name',
'abbreviation': item['stocks'][stock_id],
'dailyStatus': 0,
'monthStatus': 0,
'yearStatus': 0}
print(doc)
mon.insert_document(coll_stockInfo, doc)
def get_season_record_mongo(stock_id, year, season):
client = mon.mongo_connection('linode1', 'mongo')
collection = mon.mongo_collection(client, 'stocks', f"stock{stock_id}")
print('count_documents:', collection.count_documents({}))
if season == 1:
contents = get_3_month_records(collection, stock_id, year,
['01', '02', '03'])
return contents
elif season == 2:
contents = get_3_month_records(collection, stock_id, year,
['04', '05', '06'])
return contents
elif season == 3:
contents = get_3_month_records(collection, stock_id, year,
['07', '08', '09'])
return contents
elif season == 4:
contents = get_3_month_records(collection, stock_id, year,
['10', '11', '12'])
return contents
else:
return
def get_stock_ids():
mongo_client = mon.mongo_connection('linode1', 'mongo')
coll_stockIndustry = mon.mongo_collection(mongo_client, 'stocks',
'stockIndustry')
contents = mon.find_some_fields_mongo(coll_stockIndustry, ['stocks_list'])
return contents
def get_year_record_mongo(stock_id, year):
client = mon.mongo_connection('linode1', 'mongo')
collection = mon.mongo_collection(client, 'stocks', f"stock{stock_id}")
contents = list(collection.find({'_id': re.compile(f"{stock_id}{year}")}))
return contents
def main():
client = mon.mongo_connection('linode1', 'mongo')
coll_stockInfo = mon.mongo_collection(client, 'stocks', 'stockInfo')
redisConnect = red.redis_connection('linode1', 'redis', db=0)
while True:
try:
# check if key in redis doesn't exist
for num in range(1, 9):
key = f'stock_No_{num}'
if not redisConnect.exists(key):
# 表示還有未丟到redis的stock_id
if coll_stockInfo.find({
'crawlerStatus': 0
}, {
'_id': 1
}).count() != 0:
content = coll_stockInfo.find({
'crawlerStatus': 0
}, {
'_id': 1
}).limit(1)
stock_id = content[0]['_id']
print(f"{key} disapear >>> set {stock_id}")
# 放到redis
red.redis_set_key_value(redisConnect, key, stock_id)
# 表示已經放到redis
coll_stockInfo.update({'_id': stock_id},
{'$set': {
'crawlerStatus': 1
}})
# 表示已經從redis刪掉但還沒爬蟲好的stock_id
# elif coll_stockInfo.find({'crawlerStatus': 2}, {'_id': 1}).count() != 0:
# coll_stockInfo.update({'crawlerStatus': {'$ne': 3}}, {
# '$set': {'crawlerStatus': 0}})
# main()
# content = coll_stockInfo.find(
# {'crawlerStatus': {'$ne': 1}}, {'_id': 1}).limit(1)
# stock_id = content[0]['_id']
# print(f"{key} disapear >>> set {stock_id}")
# # 再丟上去一次redis
# red.redis_set_key_value(redisConnect, key, stock_id)
else:
wcsv.writeToCsv('./data/publisherStatus', [
'all stock ids have published on redis',
datetime.datetime.now()
])
break
print(f"{key} still exist.")
time.sleep(100)
# 全部爬蟲完成就中止丟資料到redis
if coll_stockInfo.find({
'crawlerStatus': {
'$ne': 3
}
}, {
'_id': 1
}).count() == 0:
print("== All stock crawlering done ==")
break
except Exception as e:
wcsv.writeToCsv("./dataStore/redisException", [e])
print(e)
def get_stock_ids():
client = mon.mongo_connection('linode1', 'mongo')
collection = mon.mongo_collection(client, 'stocks', 'stockIndustry')
contents = collection.find({}, {'stocks_list': 1})
return contents
def crawler_daily():
counts = 0
# notify daily updation starts
goo.main('stock_crawler', 'Stocks Daily Updation Starts!')
# start time
t1 = datetime.datetime.now()
# set daily status zero for default
client = mon.mongo_connection('linode1', 'mongo')
coll_stockInfo = mon.mongo_collection(client, 'stocks', 'stockInfo')
coll_stockInfo.update_many({}, {'$set': {'dailyStatus': 0}})
# today
today = datetime.date.today() #-datetime.timedelta(1)
year = today.strftime("%Y")
month = today.strftime("%m")
day = today.strftime("%d")
# get all stocks' id
for content in allStockID.all_stock_id():
stock_id = content['_id']
print(stock_id)
retry = 0
url = f"""https://www.twse.com.tw/exchangeReport/STOCK_DAY?response=html&date={year}{month}01&stockNo={stock_id}"""
coll_stock = mon.mongo_collection(client, 'stocks', f"stock{stock_id}")
while retry < 3:
try:
contents = crawler.crawler(url)
# print(contents)
for item in contents:
# daily record to mongo
mon.insert_document(coll_stock, item)
# crawlering and writing to mongo done, set daily status as datetime
coll_stockInfo.update_one(
{'_id': stock_id},
{'$set': {
'dailyStatus': f"{year+month+day}"
}})
counts += 1
time.sleep(10)
break
except Exception as e:
print(e)
time.sleep(10)
retry += 1
if retry == 3:
# sent notify with googlebot
goo.main('stock_crawler',
f"{stock_id}, {year,month,day} Wrong: {e}")
wcsv.writeToCsv(
f'./dataStore/DailyCrawlerException_{today}',
[stock_id, year, month, day])
continue
# check daily update done
if coll_stockInfo.find({
'dailyStatus': {
'$ne': f"{year+month+day}"
}
}, {
'_id': 1
}).count() != 0:
crawler_daily()
# notify daily updation done
cost_time = datetime.datetime.now() - t1
goo.main(
'stock_crawler',
f"{datetime.date.today()}: Daily Updation Finished!\nCheck amount of stock: {counts}, except: {938-counts}\nCost_time: {cost_time}"
)
return