How to restart tweepy script in case of error?

Question

I have a python script that continuously stores tweets related to tracked keywords to a file. However, the script tends to crash repeatedly due to an error appended below. How do I edit the script so that it automatically restarts? I've seen numerous solutions including this (Restarting a program after exception) but I'm not sure how to implement it in my script.

import sys
import tweepy
import json
import os

consumer_key=""
consumer_secret=""
access_key = ""
access_secret = ""

auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
auth.set_access_token(access_key, access_secret)
api = tweepy.API(auth)
# directory that you want to save the json file
os.chdir("C:\Users\json_files")
# name of json file you want to create/open and append json to
save_file = open("12may.json", 'a')

class CustomStreamListener(tweepy.StreamListener):
    def __init__(self, api):
        self.api = api
        super(tweepy.StreamListener, self).__init__()

        # self.list_of_tweets = []

    def on_data(self, tweet):
        print tweet
        save_file.write(str(tweet))

    def on_error(self, status_code):
        print >> sys.stderr, 'Encountered error with status code:', status_code
        return True # Don't kill the stream
        print "Stream restarted"

    def on_timeout(self):
        print >> sys.stderr, 'Timeout...'
        return True # Don't kill the stream
        print "Stream restarted"

sapi = tweepy.streaming.Stream(auth, CustomStreamListener(api))
sapi.filter(track=["test"])

===========================================================================

Traceback (most recent call last):
  File "C:\Users\tweets_to_json.py", line 41, in <module>
    sapi.filter(track=["test"])
  File "C:\Python27\lib\site-packages\tweepy-2.3-py2.7.egg\tweepy\streaming.py", line 316, in filter
    self._start(async)
  File "C:\Python27\lib\site-packages\tweepy-2.3-py2.7.egg\tweepy\streaming.py", line 235, in _start
    self._run()
  File "C:\Python27\lib\site-packages\tweepy-2.3-py2.7.egg\tweepy\streaming.py", line 165, in _run
    self._read_loop(resp)
  File "C:\Python27\lib\site-packages\tweepy-2.3-py2.7.egg\tweepy\streaming.py", line 206, in _read_loop
    for c in resp.iter_content():
  File "C:\Python27\lib\site-packages\requests-1.2.3-py2.7.egg\requests\models.py", line 541, in generate
    chunk = self.raw.read(chunk_size, decode_content=True)
  File "C:\Python27\lib\site-packages\requests-1.2.3-py2.7.egg\requests\packages\urllib3\response.py", line 171, in read
    data = self._fp.read(amt)
  File "C:\Python27\lib\httplib.py", line 543, in read
    return self._read_chunked(amt)
  File "C:\Python27\lib\httplib.py", line 603, in _read_chunked
    value.append(self._safe_read(amt))
  File "C:\Python27\lib\httplib.py", line 660, in _safe_read
    raise IncompleteRead(''.join(s), amt)
IncompleteRead: IncompleteRead(0 bytes read, 1 more expected)

Solution 1

Figured out how to incorporate the while/try loop by writing a new function for the stream:

def start_stream():
    while True:
        try:
            sapi = tweepy.streaming.Stream(auth, CustomStreamListener(api))
            sapi.filter(track=["Samsung", "s4", "s5", "note" "3", "HTC", "Sony", "Xperia", "Blackberry", "q5", "q10", "z10", "Nokia", "Lumia", "Nexus", "LG", "Huawei", "Motorola"])
        except: 
            continue

start_stream()

I tested the auto restart by manually interrupting the program with CMD + C. Nonetheless, happy to hear of better ways to test such functionality.

Solution 2

I had this problem occurring recently and wanted to share more detailed information about it.

The error that's causing it is because the streaming filter that's chosen is too broad test. Therefore you receive streams at a faster rate than you can accept which causes an IncompleRead error.

This can be fixed by either refining the search or by using a more specific exception:

from http.client import IncompleteRead
...
try:
    sapi = tweepy.streaming.Stream(auth, CustomStreamListener(api))
    sapi.filter(track=["test"])
except IncompleRead:
    pass

Solution 3

It's better to use recursive call instead of infinite while loop. Take a look at filter function below. e.g.

from tweepy import Stream
from service.twitter.listener.tweety_listener import TweetyStreamDataListener
from settings import twitter_config

class Tweety(object):
    def __init__(self, listener=TweetyStreamDataListener()):
        self.listener = listener
        self.__auth__ = None

    def __authenticate__(self):
        from tweepy import OAuthHandler
        if self.__auth__ is None:
            self.__auth__ = OAuthHandler(twitter_config['consumer_key'], twitter_config['consumer_secret'])
            self.__auth__.set_access_token(twitter_config['access_token'], twitter_config['access_token_secret'])
        return self.__auth__ is not None

    def __streamer__(self):
        is_authenticated = self.__authenticate__()
        if is_authenticated:
            return Stream(self.__auth__, self.listener)
        return None

    def filter(self, keywords=None, async=True):
        streamer = self.__streamer__()
        try:
            print "[STREAM] Started steam"
            streamer.filter(track=keywords, async=async)
        except Exception as ex:
            print "[STREAM] Stream stopped! Reconnecting to twitter stream"
            print ex.message, ex.args
            self.filter(keywords=keywords, async=async)