Python - multiprocessing while writing to a single result file - python

I am really new to the multiprocessing package and I am failing to get the task done.
I have lots of calculations to do on a list of objects.
The results I need to write down are saved in those objects, too.
The results should be written in a single file as soon as the process finished the calculations (the way I got it at least working, waits until all calculations are done).
import multiprocessing
import time
import csv
class simpl():
def __init__(self, name, val):
self.name = name
self.val = val
def pot_val(inpt):
print("Process %s\t ..." % (inpt.name))
old_v = inpt.val
inpt.val *= inpt.val
if old_v != 8:
time.sleep(old_v)
print("Process %s\t ... Done" % (inpt.name))
def mp_worker(inpt):
pot_val(inpt)
return inpt
def mp_handler(data_list):
p = multiprocessing.Pool(4)
with open('results.csv', 'a') as f:
res = p.map_async(mp_worker, data_list)
results = (res.get())
for result in results:
print("Writing result for ",result.name)
writer= csv.writer(f, lineterminator = '\n', delimiter=";")
writer.writerow((result.name, result.val))
if __name__=='__main__':
data = []
counter=0
for i in range(10):
data.append(simpl("name"+str(counter),counter))
counter += 1
for d in data:
print(d.name, d.val)
mp_handler(data)
How to write the results from the calculations simultaneously to one single file, without having to wait for all processes to finish?

You can use imap_unordered
def mp_handler(data_list):
p = multiprocessing.Pool(4)
with open('results.csv', 'a') as f:
writer= csv.writer(f, lineterminator = '\n', delimiter=";")
for result in p.imap_unordered(mp_worker, data_list):
print("Writing result for ",result.name)
writer.writerow((result.name, result.val))
With Python 3.3+ better do
def mp_handler(data_list):
with multiprocessing.Pool(4) as p:
with open('results.csv', 'a') as f:
writer= csv.writer(f, lineterminator = '\n', delimiter=";")
for result in p.imap_unordered(mp_worker, data_list):
print("Writing result for ",result.name)
writer.writerow((result.name, result.val))

Related

Storing result of a thread or process with concurrent.futures

I'm writing a utility I can use to check ports on many subnets. Currently I'm adding my results to a csv file and then sorting the file. I would like to instead add my results to a single list and then output the list so I'm doing fewer file open/close operations. I cannot seem to figure out how to make my results persist between threads. Below is my code:
import csv
import test_ports
import pandas
import ipaddress
import concurrent.futures
import time
import os
class check_subnets(object):
def __init__(self):
self.tested_list = []
def setup(self, l_subnets):
with concurrent.futures.ProcessPoolExecutor(max_workers=4) as executor:
executor.map(self.subnet_search, l_subnets)
return self.tested_list
def subnet_search(self, sub):
print("Testing the " + sub + " subnet.")
with concurrent.futures.ThreadPoolExecutor(max_workers=16) as executor2:
executor2.map(self.ip_search, ipaddress.IPv4Network(sub))
def ip_search(self, ip):
test = test_ports.TestPort()
s_ip_addr = str(ip)
print("Tested " + s_ip_addr)
test_ssh = test.test_ssh(s_ip_addr)
test_rdp = test.test_rdp(s_ip_addr)
this_list = [s_ip_addr, test_ssh, test_rdp]
self.tested_list.append(this_list)
with open('tested.csv', 'a') as file:
writer = csv.writer(file)
writer.writerow(this_list)
file.close()
if __name__ == '__main__':
subnets = pandas.read_csv('hosts.csv')
list_subnets = subnets['Subnet'].values.tolist()
fields = ['IP_Addr', "SSH(22)", "RDP(443)"]
with open('tested.csv', 'w') as f:
write = csv.writer(f)
write.writerow(fields)
f.close()
t0 = time.time()
checker = check_subnets()
results = checker.setup(list_subnets)
print(results)
t1 = time.time()
print(t1-t0)
with open("tested.csv", 'r',newline='') as f_input:
csv_input = csv.DictReader(f_input)
data = sorted(csv_input, key=lambda row: (row['IP_Addr']))
f_input.close()
with open("sorted.csv", 'w', newline='') as f_output:
csv_output = csv.DictWriter(f_output, fieldnames=csv_input.fieldnames)
csv_output.writeheader()
csv_output.writerows(data)
f_output.close()
if os.path.exists("tested.csv"):
os.remove("tested.csv")
else:
print("The file does not exist")
I'm using the class to try and create some kind of location each method would see. I have a feeling the class-specific tested_list is not available to each thread, rather each thread is seeing one instance of tested_list and not a shared list.
The test_ports module is just a wrapper for some socket operations.
I figured out that there is a small difference in concurrent.futures.ProcessPoolExecutor
and
concurrent.futures.ThreadPoolExecutor
ThreadPoolExecutor is doing exactly what I wanted, preserving data between threads. New code looks like this:
import csv
import test_ports
import pandas
import ipaddress
import concurrent.futures
import time
class check_subnets(object):
def __init__(self):
self.tested_list = []
def setup(self, l_subnets):
with concurrent.futures.ThreadPoolExecutor(max_workers=4) as executor:
executor.map(self.subnet_search, l_subnets)
return self.tested_list
def subnet_search(self, sub):
with concurrent.futures.ThreadPoolExecutor(max_workers=16) as executor2:
executor2.map(self.ip_search, ipaddress.IPv4Network(sub))
def ip_search(self, ip):
test = test_ports.TestPort()
s_ip_addr = str(ip)
test_ssh = test.test_ssh(s_ip_addr)
test_rdp = test.test_rdp(s_ip_addr)
this_list = [s_ip_addr, test_ssh, test_rdp]
self.tested_list.append(this_list)
if __name__ == '__main__':
subnets = pandas.read_csv('hosts.csv')
list_subnets = subnets['Subnet'].values.tolist()
t0 = time.time()
checker = check_subnets()
results = checker.setup(list_subnets)
t1 = time.time()
print(t1-t0)
sorted_list = (sorted(results, key=lambda x: x[0]))
fields = ['IP_Addr', "SSH(22)", "RDP(443)"]
with open('tested.csv', 'w') as f:
write = csv.writer(f)
write.writerow(fields)
write.writerows(sorted_list)
f.close()
The end result is a sorted list of opened and closed ssh and rdp ports.

How to extract one column to another file from a 300GB file

Problem was the huge data number, and I have to do it with my personal laptop with 12GB RAM. I tried a loop with 1M. lines every round, and used csv.writer. But csv.writer wrote like 1M. lines every two hours. So, any other ways worth to try?
lines = 10000000
for i in range(0, 330):
list_str = []
with open(file, 'r') as f:
line_flag = 0
for _ in range(i*lines):
next(f)
for line in f:
line_flag = line_flag + 1
data = json.loads(line)['name']
if data != former_str:
list_str.append(data)
former_str = data
if line_flag == lines:
break
with open(self.path + 'data_range\\names.csv', 'a', newline='') as writeFile:
writer = csv.writer(writeFile, delimiter='\n')
writer.writerow(list_str)
writeFile.close()
another version
def read_large_file(f):
block_size = 200000000
block = []
for line in f:
block.append(line[:-1])
if len(block) == block_size:
yield block
block = []
if block:
yield block
def split_files():
with open(write_file, 'r') as f:
i = 0
for block in read_large_file(f):
print(i)
file_name = write_name + str(i) + '.csv'
with open(file_name, 'w', newline='') as f_:
writer = csv.writer(f_, delimiter='\n')
writer.writerow(block)
i += 1
This was after it read a block and writing ... I wonder how come the rate of data trasmission was keeping about 0.
It should be as simple as this:
import json
import csv
with open(read_file, 'rt') as r, open(write_file, 'wt', newline='') as w:
writer = csv.writer(w)
for line in r:
writer.writerow([json.loads(line)['name']])
I tried the loop inside the file, but I always get me a Error, I guessed we cannot write the data into another file while opening the file?
You totally can write data in one file while reading another. I can't tell you more about your error until you post what it said, though.
There was a bit in your code about former_str which is not covered under "extract one column", so I did not write anything about it.
Would something like this work?
Essentially using a generator to avoid reading the entire file in memory, and writing the data one line at a time.
import jsonlines # pip install jsonlines
from typing import Generator
def gen_lines(file_path: str, col_name: str) -> Generator[str]:
with jsonline.open(file_path) as f:
for obj in f:
yield obj[col_name]
# Here you can also change to writing a jsonline again
with open(output_file, "w") as out:
for item in gen_lines(your_file_path, col_name_to_extract):
out.write(f"{item}\n")

Python Multithreading CSV Read Write Sort

I have a use case to read from a csv file (inputs.csv) and call an API endpoint for each row of the csv. And write the output into a different csv (outputs.csv) and sort the output csv by a particular column. I am able to achieve all that by the below code. Need to find out if I can do it more efficiently in a multi-threaded way.
def main():
start = time.time()
print "read from csv file "
input_file = 'inputs.csv'
output_file = 'outputs.csv'
read_write_csv(input_file, output_file)
print("after output-->",time.time() - start)
sort_csv(output_file)
print("after sort -->",time.time() - start)
def read_write_csv(input_file, output_file):
with open(input_file, 'r') as csv_file:
reader = csv.reader(csv_file)
# Reading row by row
count = 0
for row in reader:
# Opening csv result file in append mode
with open(output_file, "a+") as csv_save:
writer = csv.writer(csv_save)
print "mac address ", row[0]
writer.writerow([row[0], callExternalAPI(row[0])]),(row,))
print "{0} devices processed so far".format(count+1)
csv_save.close()
def sort_csv(output_file):
with open('sorted.csv', 'w') as csv_final:
r = csv.reader(open(output_file), delimiter=",")
writer_final = csv.writer(csv_final)
sortedResponse = sorted(r, key=operator.itemgetter(1), reverse=True)
for row in sortedResponse:
writer_final.writerow(row)
main()
As you can see, I am a python newbie here so any suggestion to improve the code is most welcome.

gevent pool getting stuck

I am a gevent newbie, but I think I got it working — in a limited sense. Basically, for pools of 1, the code proceeds, while for larger pools the code gets stuck, usually within the first pool (e.g. with a pool of 5, I see 3 greenlet finishing, but not more). What is going wrong? Spawn? Join?
I cannot verify whether the remote server gets confused by multiple queries, but it has no problem with a rapid sequence of serial requests, so probably not…
(I share the code in its entirety as I am not sure where the bug is. Thanks for bearing with me.)
from urllib2 import urlopen
from lxml.etree import parse
import os, csv, cStringIO, codecs, pickle
from selenium import webdriver
from time import sleep
import gevent
from gevent import socket
from gevent import monkey, pool
# patches stdlib (including socket and ssl modules) to cooperate with other greenlets
monkey.patch_all()
class UnicodeWriter:
"""
A CSV writer which will write rows to CSV file "f",
which is encoded in the given encoding.
"""
def __init__(self, f, dialect=csv.excel, encoding="utf-8", **kwds):
# Redirect output to a queue
self.queue = cStringIO.StringIO()
self.writer = csv.writer(self.queue, dialect=dialect, **kwds)
self.stream = f
self.encoder = codecs.getincrementalencoder(encoding)()
def writerow(self, row):
self.writer.writerow([unicode(s).encode("utf-8") for s in row])
# Fetch UTF-8 output from the queue ...
data = self.queue.getvalue()
data = data.decode("utf-8")
# ... and reencode it into the target encoding
data = self.encoder.encode(data)
# write to the target stream
self.stream.write(data)
# empty queue
self.queue.truncate(0)
def writerows(self, rows):
for row in rows:
self.writerow(row)
os.chdir('/Users/laszlosandor/Downloads/kozbeszerzes')
HOSTNAME = 'http://kozbeszerzes.ceu.hu'
driver = webdriver.Chrome()
results = set()
for y in xrange(1998,2015):
for p in xrange(0,9999):
driver.get('http://kozbeszerzes.ceu.hu/searchresults.xhtml?q={}&page={}'.format(y,p))
sleep(1)
if len(driver.find_elements_by_class_name('result'))==0:
break
for e in driver.find_elements_by_class_name('result'):
link = e.find_element_by_tag_name('a')
r = link.get_attribute('href').encode('ascii', 'ignore')
if r[:34]== 'http://kozbeszerzes.ceu.hu/tender/':
results.add(r)
driver.quit()
with open('list_of_urls', 'wb') as f:
pickle.dump(results, f)
#with open('list_of_urls', 'r') as f:
# results = pickle.load(f)
entities = set()
header = ('TenderID','RequestorName','URL','Year','RequestorID','Subject','SourceURL','EstValue','Currency','DecisionDate','Value','VAT')
# """Spawn multiple workers and wait for them to complete"""
# # limit ourselves to max 10 simultaneous outstanding requests
p = pool.Pool(10)
f = open('tenders.csv', 'w')
f.write(codecs.BOM_UTF8)
writer = UnicodeWriter(f)
writer.writerow(header)
def workres(res):
try:
tender = parse(urlopen(res)).getroot()
print ('%s succeeded' % res)
for requestor in tender.findall('requestor'):
entities.add(HOSTNAME + requestor.get('url'))
id = tender.get('id')
reqname = tender.get('requestor')
url = tender.get('url')
year = tender.get('year')
reqid = tender.get('requestor_id')
subject = tender.get('subject')
source = tender.get('source_url')
estval = tender.get('estimated_value')
for part in tender.findall('./parts/part'):
winner = part.find('winner')
entities.add(HOSTNAME + winner.get('url'))
curr = part.find('currency').text
date = part.find('decisionDate').text
value = part.find('value').text
vat = part.find('vat').text
row = id, reqname, url, year, reqid, subject, source, estval, curr, date, value, vat
writer.writerow(row)
except socket.gaierror:
ex = sys.exc_info()[1]
print ('%s failed with %s' % (res, ex))
jobs = [p.spawn(workres, res) for res in results]
p.join()
f.close()
with open('entities', 'wb') as f:
pickle.dump(entities, f)
header = ['ID','URL','Name','NominalCity','City', 'ZIP', 'Address']
f = open('entities.csv', 'w')
f.write(codecs.BOM_UTF8)
writer = UnicodeWriter(f)
writer.writerow(header)
def workent(ent):
try:
ent = parse(urlopen(ent)).getroot()
print ('%s succeeded' % ent)
id = ent.get('id')
url = ent.get('url')
name = ent.get('name')
nominalcity = ent.get('city')
cities = ent.findall('./resolved_addresses/whitelistAddress/city')
zips = ent.findall('./resolved_addresses/whitelistAddress/postalCode')
streets = ent.findall('./resolved_addresses/whitelistAddress/street')
for a in xrange(0,len(cities)):
city = cities[a].text
zip = zips[a].text
street = streets[a].text
row = id, url, name, nominalcity, city, zip, street
writer.writerow(row)
except socket.gaierror:
ex = sys.exc_info()[1]
print ('%s failed with %s' % (ent, ex))
jobs = [p.spawn(workent, ent) for ent in entities]
p.join()
f.close()
I see many mistakes here.
There is not used gevent.sleep() and not time.sleep which is
blocking.
Your variables names are too short. Your could add
descriptions on what each part of code is supposed to do. for example the variable 'p'
is used twice..
There are multiple urls gets using urlopen and the driver module? confusing..
I would use queues between different workers and have just one worker do
write_row calls and deal with the file access now you have multiple green lets accessing
the same file..
use less list compehensions just write out the loops.
I would suggest putting the try except in 'workres' only around the 'parse(urlopen())'
code maybe there are more exceptions happening, which you now don't see.
more tips for gevent

Removing Duplicate CSV Entries with python

I have just completed a script that (sigh) finally works. It searches twitter for keywords. The results are written to a csv with 4 columns of keyword, Tweet, Lat, Lon (location). The code that I'm using is:
import tweepy
import csv
keywordList = ['McDonalds', 'Taco Bell', 'Burger King',]
for keyword in keywordList:
result = tweepy.api.search(q=keyword,rpp=1000,page=2, geocode= "34.085422,-117.900879,500mi" )
with open(r'C:\Temp\results.csv', 'a') as acsv:
w = csv.writer(acsv)
for tweet in result:
lat, lon = tweet.geo if tweet.geo else ('', '')
try:
a = tweet.geo['coordinates']
print a[0] , a[1]
print tweet.text
w.writerow((keyword, tweet.text, a[0] , a[1]))
except:
pass
I want to use task manager or python to run this search every 5 minutes but It will rewrite duplicates. I was going to use the following code to remove duplicates but two things happen. The resutls2.csv is blank and when I go to open the csv, it is locked and I have to view it in a read only. I tried f1.close(), writer.close() etc but it says 'csv.reader' object has no attribute close.
My biggest concern is getting no duplicates either by writing to the new csv or somehow removing and writing to the same table on each search. Any suggestions are much appreciated!!
import csv
f1 = csv.reader(open(r'C:\Temp\results.csv', 'rb'))
writer = csv.writer(open(r'C:\Temp\results2.csv', 'wb'))
tweet = set()
for row in f1:
if row[1] not in tweet:
writer.writerow(row)
tweet.add( row[1] )
f1.close()
writer.close()
Here's a refactored version:
Edit: unicode, what fun - I've added a .decode() call in read_csv() and an .encode() call in append_csv(); this should solve your problem (I think - you might need to decide on a string codec).
import tweepy
import csv
from collections import defaultdict
import time
FILE = 'c:/temp/results.csv'
KEYWORDS = ['McDonalds', 'Taco Bell', 'Burger King']
WHERE = "34.085422,-117.900879,500mi"
DELAY = 300 # seconds
def _float(s, err=None):
try:
return float(s)
except ValueError:
return err
def _str(f, err=""):
return err if f is None else str(f)
def read_csv(fname=FILE):
data = defaultdict(dict)
with open(fname, 'rb') as inf:
incsv = csv.reader(inf)
for kw,tw,lat,lon in incsv:
# added .decode() call to handle saved unicode chars
data[kw][tw.decode()] = (_float(lat), _float(lon))
return data
def append_csv(data, fname=FILE):
with open(fname, "ab") as outf:
outcsv = csv.writer(outf)
# added .encode() call to handle saved unicode chars
outcsv.writerows((kw,tw.encode(),_str(lat),_str(lon)) for kw,dat in data.iteritems() for tw,(lat,lon) in dat.iteritems())
def search_twitter(keywords=KEYWORDS, loc=WHERE):
data = defaultdict(dict)
for kw in keywords:
for tweet in tweepy.api.search(q=kw, rpp=1000, page=2, geocode=loc):
data[kw][tweet.text] = tweet.geo if tweet.geo else (None,None)
return data
def calc_update(old_data, new_data):
diff = defaultdict(dict)
for kw,dat in new_data.iteritems():
for tw,loc in dat.iteritems():
if tw not in old_data[kw]:
diff[kw][tw] = old_data[kw][tw] = loc
return old_data, diff
def show_data(data):
for kw,dat in data.iteritems():
for tw,(lat,lon) in dat.iteritems():
print("<{},{}> {} [{}]".format(_str(lat,"???"), _str(lon,"???"), tw, kw))
def main():
data = read_csv()
while True:
new_data = search_twitter()
data,diff = calc_update(data, new_data)
append_csv(diff)
show_data(diff)
time.sleep(DELAY)
if __name__=="__main__":
main()

Categories